{"id": "f1e9a3cd0534fb97e149ba30818ae576fc64bd509e69b7c64fad1fda368ee3de", "sources": ["arxiv"], "title": "Exploration Structure in LLM Agents for Multi-File Change Localization", "abstract": "Software engineering tools increasingly rely on LLM based agents to localize files to change to resolve a software issue. Most AI agents explore repositories linearly, that is, visiting one directory or file per step. We postulate that this is a structural mismatch for changes that span several subsystems. We compare linear sequential exploration against non-linear, domain-scoped parallel agentic exploration. Using SWE Bench Pro as initial benchmark, we focus on ansible as an exemplar. We construct an approach for persistent-session evaluation of GitHub issues anchored at a single base commit. We compare our non-linear domain-agent file traversal system against a base LLM without direct repository access, a single agent Recursive Language Model (RLM) baseline with a persistent Python REPL and an external CLI baseline using Codex 5.5 High. Domain scoped parallel agent spawning with a small Haiku-class model achieves the highest micro F1 among Haiku class models by a large margin. Domain-agents is the second highest behind only the much larger Codex 5.5 High on our own expanded benchmark including over more recent PRs from 2025 and 2026. On the original, curated, 2020 SWE-bench Pro benchmark, a larger Sonnet plain LLM baseline attains higher micro F1 by predicting few files, leading to higher precision, but at significantly lower all gold recall. We also present three additional findings. First, documentation evolution is a latent dependency unresolved by any approach. Second, naive file system access can degrade localization driven by test-file over prediction. Lastly, forced multi-agent consultation does not measurably help and raises token cost substantially.", "authors": ["Akeela Darryl Fattha", "Kia Ying Chua", "Lingxiao Jiang", "Laura Wynter"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": [], "published_date": "2026-06-10", "url": "https://arxiv.org/abs/2606.11976", "pdf_url": "https://arxiv.org/pdf/2606.11976v1", "arxiv_id": "2606.11976", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "ae195c5a11ce9cd1c2f521b98fc8e6bf9ec4fadb23bbeed5453b2cc19ae1f643", "sources": ["arxiv", "semantic_scholar"], "title": "Early-Token Confidence Predicts Reasoning Quality in Multi-Agent LLM Debate", "abstract": "Evaluating reasoning quality in multi-agent LLM systems is challenging, especially for open-ended tasks without reference answers. We investigate whether intrinsic confidence signals, token-level log-probabilities from decoding, can predict reasoning quality as assessed by LLM-as-judge evaluation. Using a debate-based essay scoring framework, we compare confidence proxies against rubric-based judge scores across two ASAP essay sets. We find that early-token confidence, particularly within the first few generated tokens, is consistently the strongest predictor of reasoning quality, outperforming full-sequence statistics. Analysis of log-probability trajectories shows that the opening phase of generation is the most heterogeneous and therefore most informative. We also observe a systematic asymmetry between agent roles, with stronger alignment between confidence and quality for supportive reasoning than for adversarial critique. These results suggest that early decoding dynamics provide a lightweight and effective signal for estimating reasoning reliability in multi-agent LLM systems.", "authors": ["Ali Keramati", "Justin Cheok", "Jacob Horne", "Mark Warschauer"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10307", "pdf_url": "https://arxiv.org/pdf/2606.10307v1", "arxiv_id": "2606.10307", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2093031762620b5e6328a41fd02c5910fd3b0fac9aaae4e1bba4fe5c01f52017", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation", "abstract": "Large language model (LLM) agents are rapidly moving from conversational interfaces to software components that plan, invoke tools, maintain memory, and act on external environments. This transition changes the nature of security risk. In agentic settings, failures are no longer limited to unsafe text generation. Untrusted content may redirect control flow, misuse tool privileges, corrupt persistent state, leak sensitive information, or trigger harmful external actions. At the same time, research on LLM agent security is expanding quickly but remains fragmented across attack families, defense layers, application domains, and evaluation settings. This paper synthesizes 247 papers through a lifecycle-based, systems-oriented framework that models agent security around the interaction of information flow, delegated authority, and persistent state. We organize the literature around four questions: how LLM agent security should be modeled, which threat surfaces and attack families dominate, what defenses have been proposed and with what tradeoffs, and how security claims are evaluated. We find that prompt injection and tool-mediated control-flow hijacking still dominate the field, while persistent state corruption and multi-agent propagation are becoming central emerging concerns. We further find that current defenses provide useful building blocks but remain weakly compositional, and that existing benchmarks still underrepresent long-horizon, stateful, and deployment-sensitive risks. We argue that secure LLM agents require explicit trust boundaries, principled privilege control, provenance-aware state management, and evaluation practices aligned with realistic operational settings.", "authors": ["Yuchen Ling", "Shengcheng Yu", "Zhenyu Chen", "Chunrong Fang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10749", "pdf_url": "https://arxiv.org/pdf/2606.10749v1", "arxiv_id": "2606.10749", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1175f28a811a7734352b0aa219624996906b3d4783a4d49d48c4cff8f15b7271", "sources": ["arxiv", "semantic_scholar"], "title": "Game-Theoretic Multi-Agent Control for Robust Contextual Reasoning in LLMs", "abstract": "Large Language Models (LLMs) in multi-turn interactions maintain evolving context rather than generating isolated responses, making them vulnerable to prompt-injection and context-poisoning attacks in which locally plausible adversarial fragments gradually distort reasoning trajectories. Existing defenses mainly filter individual outputs and often ignore context evolution across turns, leaving long-horizon reasoning exposed. Although the Model Context Protocol (MCP) standardizes context exchange and tool invocation, it functions as a passive routing layer and does not enforce contextual stability. To address these limitations, we introduce the Game-Theoretic Secure Model Context Protocol (GT-MCP), a controller-driven multi-agent method that treats context management as a closed-loop dynamical process. GT-MCP coordinates three heterogeneous LLM agents and selects outputs through a trust function that jointly evaluates causal consistency against a validated context graph, semantic agreement among agents, and distributional drift over time. When instability is detected, a rollback-based self-healing mechanism restores the validated context and prevents unsupported fragments from propagating. Empirical evaluation over 500 interaction turns under an adaptive adversarial threat model shows that contextual drift remains bounded in 99.6% of turns, with recovery required in only 0.4%. Per-turn utility remains tightly concentrated, with median = -0.19, P05 = -0.72, and P95 = 0.30; severe degradation below -1 occurs in only 0.4% of cases, and no injection attempt succeeds at the controller level. Selected outputs maintain stable win rates above 98%, and computational overhead remains predictable, with latency per token = 1.63e-3 s.", "authors": ["Saeid Jamshidi", "Amin Nikanjam", "Arghavan Moradi Dakhel", "Kawser Wazed Nafi", "Foutse Khomh"], "categories": ["cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10322", "pdf_url": "https://arxiv.org/pdf/2606.10322v1", "arxiv_id": "2606.10322", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "19304d81686c86a0104aa49b91a6852644e0447736571e9f2f6bc8f0f65422a4", "sources": ["arxiv", "semantic_scholar"], "title": "The Confident Liar: Diagnosing Multi-Agent Debate with Log-Probabilities and LLM-as-Judge", "abstract": "Multi-agent debate systems are typically evaluated only on whether the final answer is correct, overlooking the quality of the intermediate reasoning that debate is designed to produce. This paper studies the relationship between three signals in multi-agent debate: token-level log-probability distributions over reasoning tokens, LLM-as-judge rubric scores assigned to those tokens, and final task accuracy. We examine whether internal confidence signals predict externally evaluated reasoning quality, and whether either signal aligns with task correctness, across three domains: rubric-based scoring, mathematical reasoning, and factual question answering. Our framework pairs a two-agent debate architecture -- a Constructor and an Auditor -- with an LLM-as-judge that scores each agent's reasoning along instruction following, justification quality, and evidence grounding, together with a critical-failure flag. Experiments in the rubric-scoring domain reveal a consistent four-phase confidence trajectory and a substantial role asymmetry: confidence aligns with judged reasoning quality roughly twice as strongly for the Constructor as for the Auditor, and confidence-based detection of critical reasoning failures is markedly more reliable for the Constructor (AUROC 0.804) than for the Auditor (0.634). These findings motivate the broader cross-domain investigation proposed in this paper.", "authors": ["Ali Keramati", "Justin Cheok", "Jacob Horne", "Mark Warschauer"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10296", "pdf_url": "https://arxiv.org/pdf/2606.10296v1", "arxiv_id": "2606.10296", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "8e1008b62ae6b0d13f4e09bddf01d2b8ea9e59a46c80e66193e106378e554d22", "sources": ["arxiv", "semantic_scholar"], "title": "The Arbiter Agent: Continually Monitoring Multi-Agent Conversations to Detect Emergent Misalignment", "abstract": "As AI systems built from multiple language-model agents become more common, they are increasingly used to make decisions together: discussing, negotiating, and acting on shared tasks. While individual agents may appear well-aligned when tested on their own, problems can arise from how they interact with one another. We introduce the Arbiter, an agent designed to monitor multi-agent conversations in real time and identify which participants may be behaving in misaligned ways. The Arbiter operates under a limited \"inspection budget\", meaning it must decide carefully how to use its resources. As it observes a conversation step by step, it can choose to wait, question a participant, examine internal information such as system prompts or reasoning traces, or log concerning behavior. At the end, it produces a report identifying the likely source of misalignment. We evaluate the Arbiter across five conversation conditions, ranging from risky financial advice model organisms to evaluation-aware and colluding agents, we test five tool configurations of increasing capability and two backbone models. We find that the Arbiter reliably detects misaligned agents well before the end of the conversation, with active inspection tools improving both detection accuracy and speed. Weight-induced misalignment proves hardest to detect, while instruction-induced misalignment is identified reliably even under passive observation. The logging tool exhibits a dual effect, improving recall at the cost of precision. These results suggest that continual, budget-aware monitoring can effectively catch misalignment, and that overseeing multi-agent systems may require treating the auditor as an active participant in the process. The code is available at https://github.com/aisilab/arbiter.", "authors": ["Filippo Tonini", "Federico Torrielli", "Anton Danholt Lautrup", "Peter Schneider-Kamp", "Mustafa Mert Çelikok", "Lukas Galke Poech"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10747", "pdf_url": "https://arxiv.org/pdf/2606.10747v1", "arxiv_id": "2606.10747", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/aisilab/arbiter", "venue": null, "quality_score": 0.65}
{"id": "b9d5e324ad7e57f2dbcf0e7ea938d961dffbee982adb1e73e75a2d87bb286c54", "sources": ["arxiv"], "title": "INFRAMIND: Infrastructure-Aware Multi-Agent Orchestration", "abstract": "Existing multi-agent LLM orchestration methods, ranging from brute-force ensembles to learned routers, select models and topologies based on task and model features. However, these methods do not consider the runtime state of the serving infrastructure. On shared GPU clusters under concurrent load, this infrastructure blindness causes systematic resource underutilization: preferred models accumulate deep request queues while equally capable alternatives sit idle. In multi-agent pipelines, where each query triggers multiple sequential model calls, these delays then compound across every downstream step. Closing this gap is challenging because the relevant infrastructure signals (queue depths, KV-cache pressure, latencies) are dynamic and noisy, and they must drive three different decisions: planning, per-step routing, and scheduling. We introduce INFRAMIND, a framework that makes the entire multi-agent stack infrastructure-aware. An infra-aware planner conditions topology and role selection on real-time system load and remaining budget, biasing toward simpler graphs under congestion and richer ones at low load. An infra-aware executor then observes per-model queue depths, cache utilization, and response latencies at each agent step to decide which model to call and how deeply to reason; a budget-aware scheduler further reorders each model's queue so that urgent requests are served first. Cast as a hierarchical constrained MDP and solved end-to-end via reinforcement learning, the system learns to balance quality against latency automatically. Across five benchmarks, INFRAMIND delivers up to +7.6 pp accuracy over the prior baseline at low load with up to 7x lower latency, and sustains up to 99.9% SLO compliance under high load where every baseline drops below 50%.", "authors": ["Ahasan Kabir", "Jiaqi Xue", "Mengxin Zheng", "Qian Lou"], "categories": ["cs.AI"], "fields_of_study": [], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.11440", "pdf_url": "https://arxiv.org/pdf/2606.11440v1", "arxiv_id": "2606.11440", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "70c2f0d8aded166fe1a6201c70f204817780c4547c6900c10d25b731dd1bd260", "sources": ["arxiv", "semantic_scholar"], "title": "Divide and Cooperate: Role-Decomposed Multi-Agent LLM Training with Cross-Agent Learning Signals", "abstract": "Modern language agents which perform multi-step reasoning have shown strong performance in knowledge-intensive question answering. However, existing approaches typically couple evidence acquisition and answer generation within a single policy. This forces a single model to play multiple potentially conflicting roles, inducing a combinatorial explosion in the policy space and hindering efficient exploration. It also introduces a credit assignment problem during training: a search action that retrieves sufficient evidence may still be penalized when generation fails, and vice versa. We propose DAC (Divide and Cooperate), a role-decomposed multi-agent training framework that divides agentic search into two cooperative subtasks, each handled by a dedicated agent trained with role-specific learning signals. The generator serves a dual role as both an answer producer and an evidence sufficiency verifier, abstaining when retrieved evidence is insufficient. This abstention signal is incorporated into the search agent's reward, providing structured cross-agent learning signals that improve credit assignment. Conversely, the searcher exposes the generator to diverse and challenging evidence environments by hard-positive evidence augmentation, improving its robustness. Experiments on general and multi-hop QA benchmarks show that DAC, implemented via parameter-efficient LoRA modules over a shared backbone, achieves strong performance against prior baselines that rely on full fine-tuning of monolithic models.", "authors": ["Jaewan Park", "Solbee Cho", "Jay-Yoon Lee"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-09", "url": "https://arxiv.org/abs/2606.10684", "pdf_url": "https://arxiv.org/pdf/2606.10684v1", "arxiv_id": "2606.10684", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c6a5773175161ea0251d226cb68e6e144ec0a651216ccbcb4cd0a6b832307b5d", "sources": ["arxiv", "semantic_scholar"], "title": "Context-Fractured Decomposition Attacks on Tool-Using LLM Agents: Exploiting Artifact Provenance Gaps", "abstract": "Tool-using LLM agents interact with the world through actions that persist state in artifacts (e.g., workspace files or logs). Consequently, jailbreak defenses must reason about cross-step composition rather than isolated text. Yet most existing attacks and defenses, including ``multi-turn'' jailbreaks such as Crescendo and Tree of Attacks,still assume a single contiguous conversation visible to the defender. This assumption breaks down in real agent pipelines, where enforcement is fragmented across tools, modules, and time, and where artifact provenance is often not tracked. We operationalize a deployment failure mode for tool-using LLM agents, the \\emph{provenance gap}, and study reproducible triggers for it: \\emph{Context-Fractured Decomposition} (CFD), a family of cross-context multi-step jailbreaks that preserve benign-looking intermediate artifacts from an early interaction and elicit harmful behavior much later, potentially in a different agent instance or workflow stage, via individually innocuous tool actions whose risk emerges only under delayed artifact-mediated composition. We instrument the failure mode with trace-level diagnostics and outline a verifiable mitigation direction (provenance lineage tagging). Across agent-system jailbreak benchmarks, CFD improves success rates by up to 28.3 percentage points over state-of-the-art baselines, even against strong single-turn judges. Disclaimer: This paper contains examples of harmful or offensive language.", "authors": ["Xiaofeng Lin", "Yukai Yang", "Daniel Guo", "Sahil Arun Nale", "Charles Fleming", "Guang Cheng"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-08", "url": "https://arxiv.org/abs/2606.09084", "pdf_url": "https://arxiv.org/pdf/2606.09084v1", "arxiv_id": "2606.09084", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1024a29f8c6411aa5434ac1b840eaf327f14b2912c1789702afcce0562bff4da", "sources": ["arxiv", "semantic_scholar"], "title": "AGENTSERVESIM: A Hardware-aware Simulator for Multi-Turn LLM Agent Serving", "abstract": "Multi-turn LLM agents interleave model calls with external tool invocations, shifting serving from stateless request processing to stateful program execution. Serving these workloads requires scheduling, KV-cache management, and routing policies that use program-level context, including turn dependencies, tool-induced gaps, and reusable KV state. Evaluating such policies directly on real systems is costly, since each design point may require dedicated accelerator time across arrival rates, model scales, serving-instance counts, and memory hierarchies. Simulation offers a scalable alternative, but existing LLM serving simulators target stateless request-level workloads and therefore omit the core dynamics of agent serving: multi-turn program execution, cross-turn cache locality, and KV-cache residency during tool gaps. We present AGENTSERVESIM, a hardware-aware simulator for multi-turn LLM agent serving. AGENTSERVESIM evaluates serving policies at program granularity through composable modules: a Program Orchestrator preserves program identity and turn order, a Tool Simulator materializes tool-induced gaps, a Session-Aware Router maintains program-to-instance affinity for cache-aware dispatch, and a KV Residency Model tracks policy-defined KV placement across HBM, host DRAM/CXL, and eviction. Across real serving deployments and hardware configurations, AGENTSERVESIM reproduces real-system behavior within 6% error across key performance metrics while running entirely on commodity CPUs. These results show that AGENTSERVESIM enables controlled, repeatable exploration of agent-serving policies without requiring exhaustive deployment on costly accelerators.", "authors": ["Rakibul Hasan Rajib", "Mengxin Zheng", "Qian Lou"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-08", "url": "https://arxiv.org/abs/2606.09613", "pdf_url": "https://arxiv.org/pdf/2606.09613v1", "arxiv_id": "2606.09613", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9f832e79793f1899478661987639471a4c7fe28bbe918d4dfc029dc0af44c369", "sources": ["arxiv", "semantic_scholar"], "title": "Prisma-World: Camera-Controllable Multi-Agent Video World Model", "abstract": "Video world models have made rapid progress in generating controllable visual experiences, but most of them still simulate the world from a single observer. Extending such models to multiple agents raises a central challenge: if each agent's future state is generated independently, overlapping views may instantiate different versions of the same scene, leading to inconsistent objects, layouts, and appearances across agents. Conventional camera conditioning controls individual trajectories, but it does not explicitly couple the generation of views that should agree under shared scene geometry. We introduce Prisma-World, a camera-controllable multi-agent world model that formulates multi-agent generation as a joint geometry-aware denoising process for cross-view consistency. Prisma-World processes all agent videos within one full-attention sequence, uses a multi-agent RoPE design to distinguish agent identities while preserving synchronized temporal coordinates, and injects relative camera geometry into attention to bias overlapping viewpoints toward shared scene evidence. To further strengthen multi-view consistency and enhance global spatial perception, we augment our framework with an overlap-decaying curriculum training paradigm alongside minimap-conditioned structural guidance. To facilitate the training and evaluation of multi-agent models, we introduce PrismaDataset, a large-scale UE5 dataset with panoramic acquisition across diverse scenes, composable multi-agent view groups with flexible agent counts and complex camera trajectories, and precise camera/action annotations for consistency training and evaluation. Experiments show that a single Prisma-World model can generate high-fidelity multi-agent videos with flexible agent numbers, camera controllability, improved cross-view consistency, and spatial grounding under minimap guidance.", "authors": ["Huiqiang Sun", "Zhan Peng", "Size Wu", "Kun Wang", "Kang Liao", "Dianyi Wang", "Xingyu Zeng", "Sheng Jin", "Yangguang Li", "Zhiguo Cao", "Ziwei Liu", "Wei Li"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-08", "url": "https://arxiv.org/abs/2606.09507", "pdf_url": "https://arxiv.org/pdf/2606.09507v1", "arxiv_id": "2606.09507", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "88f5e95d2ddcad1d198398171bdc862c63afe907e41fe0157736e2e54cb9fa02", "sources": ["arxiv", "semantic_scholar"], "title": "PerspectiveGap: A Benchmark for Multi-Agent Orchestration Prompting", "abstract": "Real-world LLM applications are moving beyond single-agent workflows toward orchestrated multi-agent systems, yet current models still struggle to determine what each sub-agent needs to know. To measure this, we introduce PerspectiveGap, a benchmark for evaluating LLMs' ability to compose orchestration prompts for multi-agent systems. PerspectiveGap contains 110 scenarios, each evaluated through two distractor-mixed task formats: role-fragment assignment and free-form prompt writing. These scenarios are organized into 10 topologies, which are distilled from the authors' real-world engineering practice and framed by the Prompt Economy principle: building loop-centered orchestrations that maximize utility with minimal role and engineering overhead. In experiments with 27 commercial models from 10 companies, GPT-5.5 substantially outperforms all competitors, whereas Opus 4.7 shows a notable weakness in orchestration prompting despite its strong coding performance. Nevertheless, PerspectiveGap remains challenging: the evaluated models achieve an average combined pass rate of only 14.9\\% (GPT-5.5 62.0\\%) and an average overall leakage rate of 246.5\\% (a per-scenario information leak-event count, not a proportion; GPT-5.5 49.1\\%). These findings suggest that multi-agent orchestration prompting is a distinct and under-evaluated capability, and PerspectiveGap provides a foundation for measuring and improving it systematically.", "authors": ["Youran Sun", "Xingyu Ren", "Kejia Zhang", "Xinpeng Liu", "Jiaxuan Guo"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-07", "url": "https://arxiv.org/abs/2606.08878", "pdf_url": "https://arxiv.org/pdf/2606.08878v1", "arxiv_id": "2606.08878", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4422ba6bd5afbbaeee25db98661efd55f326124dd1123c3660245ec40556ab64", "sources": ["arxiv", "semantic_scholar"], "title": "SGTO-MAS: Secure Gorilla Troops Optimization for Multi-Agent LLM Systems", "abstract": "Multi-agent large language model (LLM) systems offer strong capabilities for complex reasoning and decision-making, yet coordination across agents introduces error propagation, security risks, and inefficient use of resources. Existing methods often rely on heuristic, static strategies and lack a principled mechanism for balancing performance, security, and computational cost. This paper formulates multi-agent LLM coordination as a constrained optimization problem and proposes a security-aware method for adaptive agent selection. The method integrates trust modeling, risk-aware evaluation, and collective intelligence within a unified optimization objective. To solve the problem efficiently, we use a swarm-intelligence strategy inspired by Gorilla Troops Optimization (GTO), enabling adaptive coordination under varying threat conditions. Controlled experiments across 500 independent runs demonstrate the effectiveness of the proposed method. The system achieves a stable average performance score of 0.5281, with high consensus (0.8764), controlled risk (0.3000), and compact agent subsets averaging 4.04 selected agents. The optimization process converges efficiently, with an average runtime of 24.09 seconds per run and low score variability (standard deviation = 0.0173). Robustness analysis indicates graceful degradation under perturbations, with performance drops limited to 2.5% under agent removal and 5.3% under consensus disruption. These results show that effective multi-agent coordination can be achieved through structured optimization that jointly manages performance, security, and efficiency. The proposed method provides a practical security-aware solution for coordinating multi-agent LLM systems in complex adversarial settings.", "authors": ["Saeid Jamshidi"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.07940", "pdf_url": "https://arxiv.org/pdf/2606.07940v1", "arxiv_id": "2606.07940", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a67b489174115d59302461692a661b33a73183bda3dc6ee38b4d8c55136bf0d4", "sources": ["arxiv", "semantic_scholar"], "title": "Causal Agent Replay: Counterfactual Attribution for LLM-Agent Failures", "abstract": "When an LLM agent fails -- issues a refund it should not have, calls the wrong tool, leaks data -- existing tooling answers what happened (observability) or whether it passed (evaluation), but not which step caused the failure. The obvious heuristics are wrong: the step that executes the harmful action is usually not the step that decided on it, and LLM-judge attribution is correlational and unreliable (state-of-the-art step-level accuracy on the Who&When benchmark is about 14%). We present Causal Agent Replay (CAR), which answers the question by intervention: it models an agent run as a structural causal model, applies a do-operation to a step, and re-executes the trajectory forward under the same stochastic policy, measuring the shift in the outcome distribution. We define an intervention algebra over agent steps, a single-step contrastive estimator whose point-of-commitment rule resolves a confound specific to stochastic run-forward, and a budget-bounded Monte-Carlo Shapley estimator that splits credit across interacting steps. Every effect is reported with confidence intervals. We validate against synthetic structural causal models with planted ground truth: the contrastive estimator recovers the pivotal step, and Shapley recovers a two-step interaction (0.44, 0.45, ~0; efficiency sum 0.909 versus the analytic 0.91). CAR is open source and runs on hosted or free local models.", "authors": ["Jaineet Shah"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.08275", "pdf_url": "https://arxiv.org/pdf/2606.08275v1", "arxiv_id": "2606.08275", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/jaineet17/causal-agent-replay", "venue": null, "quality_score": 0.65}
{"id": "8fd6fafdc2599ecc0966046ca9214783e084f4345d20359bf74155d5fc602f0c", "sources": ["arxiv", "semantic_scholar"], "title": "Hallucination Cascade: Analyzing Error Propagation in Multi-Agent LLM Systems", "abstract": "Large Language Models (LLMs) generate fluent text but remain vulnerable to hallucinations, producing unsupported, inconsistent, and factually incorrect claims. Most prior work treats hallucination as a static property of isolated outputs. In multi-agent LLM systems, however, responses are exchanged across agents, revised through sequential stages, and reused as context for later reasoning. Hallucination, therefore, becomes a dynamic process shaped by interaction history, cascade depth, and model heterogeneity. This paper analyzes hallucination dynamics in multi-agent LLM cascades by tracking claim-level factual inconsistencies across sequential agent interactions. We conduct 500 cascade experiments across 10 knowledge domains using GPT-5.3, DeepSeek-V3, and LLaMA-3-70B-Instruct, yielding 1,250 evaluated responses. Results show that deeper cascades reduce the normalized hallucination score from 0.422 at the first agent to 0.272 at the final agent in 3-agent chains, with an amplification factor of 0.644, indicating net attenuation. This reduction is accompanied by a decline in factual accuracy from 0.789 to 0.769, revealing a trade-off between hallucination suppression and factual preservation. Transition-level analysis shows that each agent-to-agent refinement reduces hallucination by an average of 0.072, with small but consistent losses in factual consistency and response quality. Model-level results reveal reliability-efficiency trade-offs: LLaMA-3-70B-Instruct achieves the lowest hallucination score, whereas GPT-5.3 provides faster generation with a higher hallucination rate. Domain-level analysis shows that hallucination varies with topic complexity, with lower scores in well-grounded scientific domains and higher scores in more abstract domains.", "authors": ["Saeid Jamshidi", "Arghavan Moradi Dakhel", "Kawser Wazed Nafi", "Foutse Khomh"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.07937", "pdf_url": "https://arxiv.org/pdf/2606.07937v1", "arxiv_id": "2606.07937", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a1ff74635280f78bb7a28de1c484cf0245d0ee85ca50f73e05833318704c31ef", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Human-Centered Multi-Agent Systems: Integrating Cognition, Culture, Values, and Cooperation in AI Agents", "abstract": "The emergence of large language model (LLM)-based agents and multi-agent systems has enabled a shift from narrow task automation to more autonomous decision-making. Despite progress in language generation, planning, tool use, and coordination, most agents still treat intelligence as prediction, optimization, and task completion. Human environments are social and normative, where people reason under bounded rationality, communicate in culturally situated language, and make decisions guided by values, beliefs, trust, and social norms. This survey argues that future AI agents, especially those acting on behalf of humans, must move beyond task competence toward human-centered capabilities. We review research across six areas: (1) evolution of intelligent agents, (2) human cognition and decision-making, (3) language, culture, and social context, (4) human values and belief systems, (5) human-agent collaboration, and (6) multi-agent coordination and modeling of human characteristics. We synthesize work from cognitive science, sociolinguistics, computational social science, and AI alignment, along with recent advances in LLM agents, cultural alignment benchmarks, preference learning, explainability, and agent societies. We identify a key gap: existing systems do not provide a unified framework integrating cognition, culture, values, and social behavior into autonomous agents. We conclude with directions for building culturally aware, value-aligned, cognitively grounded, and cooperative multi-agent systems.", "authors": ["Safia Baloch", "Rahemeen Khan"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.08274", "pdf_url": "https://arxiv.org/pdf/2606.08274v1", "arxiv_id": "2606.08274", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3e9b0be011efef4291109bf370698c6756c5b0c499c1adfd06004cb7e245dd2e", "sources": ["arxiv", "semantic_scholar"], "title": "Collective Hallucination in Multi-Agent LLMs:Modeling and Defense", "abstract": "Hallucinations in large language models (LLMs) create heightened risks in multi-agent settings, where recursive agent interactions can propagate, reinforce, and amplify unsupported claims. This paper models hallucination as a system-level, time-evolving process across a network of interacting LLM agents, where nodes represent agents and edges encode information exchange. The proposed formulation captures how hallucinated claims diffuse through communication topologies, intensify under adversarial perturbations, and affect collective reliability across reasoning rounds. To suppress error propagation, we introduce an interaction-aware control method that combines confidence-weighted aggregation, adaptive impact regulation, external claim verification, and selective isolation of unreliable agents. Experiments on TruthfulQA and TriviaQA show that the proposed method reduces hallucination by up to 39.0% relative to undefended multi-agent reasoning, improves factual accuracy from 0.79 to 0.87, and increases semantic consistency from 0.75 to 0.84. Under adversarial conditions, the method limits hallucination amplification to 1.08, compared with 1.45 without adaptive control, maintaining stable collective behavior across recursive interaction rounds. These results indicate that hallucination in multi-agent LLM systems is governed by both individual model reliability and system-level interaction dynamics, including communication topology, confidence coupling, and recursive information flow.", "authors": ["Saeid Jamshidi"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.07941", "pdf_url": "https://arxiv.org/pdf/2606.07941v1", "arxiv_id": "2606.07941", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0474cd39b1aa7342093d0e6062f33a01f5c7dfee901afcd1a6609e30052d07e3", "sources": ["arxiv", "semantic_scholar"], "title": "Benchmarking Open-Ended Multi-Agent Coordination in Language Agents", "abstract": "As language models are increasingly deployed as autonomous agents, they must coordinate with others over long horizons in open-ended interactive tasks. Yet existing evaluations rarely test these demands together, instead emphasising single-agent tasks, short interactions, or highly structured multi-agent settings. We introduce $alem$, a JAX-based benchmark for open-ended multi-agent coordination built on Craftax-like dynamics. Alem embeds procedurally generated coordination tasks, soft specialisation, communication, and controllable coordination difficulty into a long-horizon survival world with exploration, crafting, trading, and combat. We evaluate $13$ modern LLMs zero-shot within homogeneous teams, with trained MARL agents as reference points. Current LLM agents remain far from solving alem, averaging only ~6% normalised return, but their failures are not uniform. On the hardest coordination setting, zero-shot Gemini-3.1-Pro-High approaches MARL agents trained for one billion steps, while GPT-5.4-High achieves strong base-task reward but much lower coordination reward. This contrast shows that individual task competence does not imply coordination competence. Ablations show that communication is the largest contributor to coordination, while memory and reasoning help when used to maintain multi-step plans. Overall, our results identify coordination as a distinct bottleneck for frontier LLM agents, separate from single-agent capabilities. Alem makes this bottleneck measurable and provides a controlled testbed for developing agents that communicate, allocate roles, and execute shared plans. Code is available at https://github.com/alem-world/alem-env.", "authors": ["Kale-ab Abebe Tessera", "Andras Szecsenyi", "Cameron Barker", "Alexander Rutherford", "Davide Paglieri", "Aidan Scannell", "Henry Gouk", "Elliot J. Crowley", "Tim Rocktäschel", "Amos Storkey"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.08340", "pdf_url": "https://arxiv.org/pdf/2606.08340v1", "arxiv_id": "2606.08340", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/alem-world/alem-env", "venue": null, "quality_score": 0.65}
{"id": "0f34a99d9b4d1f3761b22d5da920010273bb0d32ea3375cb6cd6f2ec6095fd5b", "sources": ["arxiv", "semantic_scholar"], "title": "Emergence World: A Platform for Evaluating Long-Horizon Multi-Agent Autonomy", "abstract": "Most evaluations of LLM agents look like exams: a discrete task, a clean environment, a score in minutes or hours. We argue that this approach is mismatched with the deployment conditions of autonomous systems, where the relevant timescale can be weeks to months, and where the dynamics that matter most, such as behavioral drift, governance in diverse environmental contexts, and cross-influence between agents from different model families, only emerge over time. We introduce Emergence World, a continuously running multi-agent simulation platform designed to make those dynamics measurable. The platform hosts populations of LLM-driven agents in a shared spatial world grounded in live external data (e.g. real-time weather, news APIs, internet access), equips each agent with 120+ specialized tools and three persistent memory systems, and lets them govern themselves through democratic mechanisms with consequential outcomes. The platform is model-agnostic at the reasoning layer and supports heterogeneous populations in which agents from different vendors share the same world. To illustrate the kinds of questions the platform makes tractable, we present a 15-day cross-vendor study with five parallel worlds powered by Claude Sonnet 4.6, Grok 4.1 Fast, Gemini 3 Flash, GPT-5-mini, and a mixed population. Identical roles and starting conditions produced radically different outcomes, ranging from stable deliberative governance to total population collapse. We release the prompts, log data and configurations to support further research on long-horizon multi-agent autonomy.", "authors": ["Deepak Akkil", "Ravi Kokku", "Karthik Vikram", "Tamer Abuelsaad", "Aditya Vempaty", "Satya Nitta"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.08367", "pdf_url": "https://arxiv.org/pdf/2606.08367v1", "arxiv_id": "2606.08367", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c285e2cada1775bd86d02c7c4def51cdbb8a1e608e41c3d06054141b215edc5b", "sources": ["arxiv", "semantic_scholar"], "title": "\"So There's a Catch-22 Here\": How Early Adopters Who Build Multi-Agent LLM Systems Conceptualize Transparency", "abstract": "Multi-agent large language model (LLM) systems are rapidly emerging, yet transparency, a cornerstone of responsible AI, remains under-defined in these distributed architectures, which have complexities of inter-agent coordination and orchestration. In this paper, we present one of the first empirical study of how early adopters of multi-agent LLM systems, who are both the builders and users, understand and practice transparency. We conducted semi-structured interviews with 13 early adopters in [Large Technology Organization] and applied thematic analysis to identify recurring patterns. Participants articulated divergent yet complementary framings of transparency, including reproducibility, debugging, boundary-setting, visualization, and auditing. These perspectives spanned questions of what transparency entails, why it matters, and how it is achieved. We synthesize these into a multidimensional framework, which is developer, user, and governance-focused positioning transparency as a situated socio-technical practice that informs future HCI and AI design and research around aligning expectations and capacities of their intended audiences.", "authors": ["Suchismita Naik", "Samir Passi", "Mihaela Vorvoreanu", "Scott Saponas", "Amanda Hall"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-06", "url": "https://arxiv.org/abs/2606.08323", "pdf_url": "https://arxiv.org/pdf/2606.08323v1", "arxiv_id": "2606.08323", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "78ffc6d5ce47ae083029340c81d5498b49d5f086d452d546711feb85dfe4dc77", "sources": ["arxiv", "semantic_scholar"], "title": "Contract2Tool: Learning Preconditions and Effects for Reliable Tool-Augmented LLM Agents", "abstract": "Tool-augmented large language model agents increasingly rely on external APIs, but standard tool schemas describe how to call a tool, not when the tool is causally appropriate or what task state it produces. Causal tool filtering addresses this gap by using lightweight contracts that specify each tool's preconditions, effects, risk level, and cost. However, manually writing and maintaining such contracts does not scale to large or changing tool ecosystems. We introduce Contract2Tool, a framework for inferring tool contracts from metadata, schemas, documentation, and execution traces. Contract2Tool converts observable tool evidence into normalized symbolic contracts that can be evaluated intrinsically and deployed inside downstream causal tool filtering. We evaluate learned contracts against gold preconditions, effects, and risk labels, and measure their downstream utility on multi-step agent tasks. Our results show that hybrid documentation-and-trace evidence produces contracts accurate enough to preserve most of the reliability and efficiency benefits of gold contracts. Learned-contract CMTF achieves 0.980 downstream success, close to 0.990 for gold-contract CMTF, while reducing visible tools from 100 to 1 and reducing average token usage from 26,172 to 2,528 relative to all-tools exposure. These results suggest that learned contracts can provide a scalable contract layer between tool schemas and reliable agent execution.", "authors": ["Rahul Suresh Babu", "Laxmipriya Ganesh Iyer"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-05", "url": "https://arxiv.org/abs/2606.07904", "pdf_url": "https://arxiv.org/pdf/2606.07904v1", "arxiv_id": "2606.07904", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "024194d091c59bda910a8e4920c8bbfef34f2cd1739e70826f21985b0b57c2d9", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Agentic Tool-Calling Decisions via Uncertainty-Aligned Reinforcement Learning", "abstract": "Large language model (LLM)-based agents often make suboptimal tool-use decisions, including unsupported tool invocation and hallucinated direct responses, which may accumulate errors throughout multi-step interactions. Existing approaches mainly improve these behaviors through inference-time correction or coarse-grained reward signals based on decision outcomes and structured checklists, leaving the uncertainty characteristics of agent decisions underexplored. We observe that decision-oriented reinforcement learning tends to weaken the uncertainty separation between correct and incorrect actions, resulting in overconfident mistakes and weaker exploration signals. Therefore, we propose TRUST, which incorporates uncertainty quantification into reward design as a repulsive force for maintaining uncertainty separation, and labels lightweight key-turn annotations for unified post-training of multi-turn trajectories. Experimental results across diverse tool-use benchmarks show that TRUST consistently enhances both decision quality and agent performance while maintaining more reliable uncertainty estimates during optimization.", "authors": ["Yijin Zhou", "Linqian Zeng", "Xiaoya Lu", "Wenyuan Xie", "Dongrui Liu", "Junchi Yan", "Jing Shao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-05", "url": "https://arxiv.org/abs/2606.06976", "pdf_url": "https://arxiv.org/pdf/2606.06976v1", "arxiv_id": "2606.06976", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "043e1592190460ae552cab9ee16ee4ecc9b5e2d347d349ecb8f55290f1cc5413", "sources": ["arxiv", "semantic_scholar"], "title": "Does Persona Make LLMs K-pop Fans? A Pilot Study of LLM-Based Online Concert Audience Agents", "abstract": "A concert is a collective experience, but recorded performance videos are typically watched alone, stripping away the shared audience presence that makes concerts feel eventful. We investigate whether persona-based LLM audience agents can recreate aspects of this collective experience by generating real-time fan chat alongside a K-pop performance video. We present a multi-agent system in which ten LLM agents react through live-chat messages, comparing a persona-conditioned audience (each agent assigned a distinct fan identity, bias, and chat style) with a no-persona baseline. In a within-subjects pilot with K-pop fans (N=11), persona conditioning substantially improved model-level chat quality and perceived naturalness, but did not translate into differences in social connectedness, engagement, or affective response. Interviews suggest that online K-pop concert chat may operate as collective monologue rather than interpersonal dialogue, and that meaningful participation depends on shared identification with the specific artist and fandom. Persona conditioning can make LLM audiences appear more natural, but culturally meaningful collective experience may require deeper alignment between persona, crowd behavior, fandom identity, and user expectations.", "authors": ["Kirak Kim", "Hyojin Kim", "Yejin Son", "Sungyoung Kim", "Kyung Myun Lee"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-05", "url": "https://arxiv.org/abs/2606.07837", "pdf_url": "https://arxiv.org/pdf/2606.07837v1", "arxiv_id": "2606.07837", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d663a17dc06a3db96ec0f80fe501f6ab78deb97b0add589d55c70f0b4f4914d7", "sources": ["arxiv", "semantic_scholar"], "title": "PDE-Agents: An LLM-Orchestrated Multi-Agent Framework for Automated Finite Element Simulations with Knowledge Graph-Augmented Reasoning", "abstract": "We present PDE-Agents, a multi-agent ecosystem that automates the full lifecycle of partial differential equation (PDE) / finite element method (FEM) simulations through natural-language interaction. Three specialist large language model (LLM) agents (Simulation, Analytics, Database) are orchestrated via a LangGraph supervisor, with a local open-source LLM stack (Qwen3-Coder-Next, Llama 4 Scout) on dual NVIDIA RTX PRO 6000 GPUs. The architecture is model-agnostic, validated across two LLM generations. A GraphRAG knowledge base (Neo4j, 768-d vector embeddings) encodes curated material properties, known failure patterns, and prior run lineage. We report seven contributions: (i) a verification and validation (V&V) study confirming second-order spatial convergence (O(h^2)) on the heat-equation solver; (ii) a three-way ablation over 50 tasks with a frozen KG (KG On, KG Off, KG Smart), where KG Smart reaches 100% success and the highest output quality (physics 0.933 vs. 0.853 for KG Off; MPF 0.926 vs. 0.796); (iii) a novel-material experiment with three fictional materials known only to the KG, where KG Smart attains near-perfect material property fidelity (MPF = 1.00) versus 0.34 for the KG-free baseline; (iv) a failure analysis tracing KG On's three failures to budget exhaustion and timeout, establishing warm-start injection as the dominant reliability factor; (v) an adaptive framework selecting the optimal retrieval mode per task; (vi) production metrics from 1,369 runs (97.8% success, 57.6% first-try); and (vii) a 100-task KG growth experiment showing a difficulty-dependent gain, with hard-task MPF improving 8.8% while easy/novel tasks stay at ceiling. All code, models, and evaluation artifacts are released openly. Our findings show that integration pattern, not knowledge content, determines whether GraphRAG augmentation helps or hinders LLM agents.", "authors": ["Sayan Adhikari", "Gulshan Noorsumar", "Øyvind Jensen"], "categories": ["physics.comp-ph", "math-ph"], "fields_of_study": ["Physics", "Mathematics"], "published_date": "2026-06-05", "url": "https://arxiv.org/abs/2606.07850", "pdf_url": "https://arxiv.org/pdf/2606.07850v1", "arxiv_id": "2606.07850", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/MatPro-IFE/pde-agents", "venue": null, "quality_score": 0.65}
{"id": "5ea3182923a447700ff7de9a75dc9d912a92dbc16a5d4fdefccb458d42d9e167", "sources": ["arxiv", "semantic_scholar"], "title": "YouZhi: Towards High-Concurrency Financial LLMs via Adaptive GQA-to-MLA Transition", "abstract": "Large language models (LLMs) drive significant financial innovations, yet their high-concurrency deployment is severely bottlenecked by KV cache memory overhead, which inflates infrastructure costs and throttles scalability. To address this, we propose YouZhi-LLM, a highly efficient financial LLM empowered by a comprehensive structural transition and training pipeline natively built on the Huawei Ascend ecosystem. At its algorithmic core, YouZhi-LLM features a layer-adaptive GQA-to-MLA transition framework that dynamically assigns per-layer FreqFold sizes, maximizing KV-cache compression while minimizing perplexity degradation. To recover representation capacity and inject domain expertise, the Ascend-based training pipeline seamlessly integrates generalized knowledge distillation with financial-specific supervised fine-tuning. Evaluations demonstrate the superiority of this systematic approach, with the adaptive transition reducing perplexity degradation by up to 35% over uniform baselines. Crucially, when evaluated on Ascend NPUs via vLLM-Ascend, the massive KV-cache reduction translates directly into deployment efficiency. Compared to their respective base models, YouZhi-7B yields a 12.3% improvement in average financial benchmark score alongside a 2.69$\\times$ increase in maximum concurrency; similarly, YouZhi-14B achieves a 7.0% accuracy gain and a 2.43$\\times$ concurrency boost, establishing a new paradigm for cost-effective, high-throughput financial inference.", "authors": [" PSBC LLM Team", " Huawei LLM Team", "Ruihan Long", "Junjie Wu", "Tianan Zhang", "Duo Zhang", "Yaozong Wu", "Jinbin Fu", "Chang Liu", "Zhentao Tang", "Wenshuang Yang", "Xin Wang", "Zhihao Song", "Ning Huang", "Wenjing Xu", "Shuai Zong", "Shupei Sun", "Sen Wang", "Jing Hu", "Bin Wang", "Xinyu Wang", "Junkui Ju", "Zequn Ding", "Jie Ran", "Man Luo", "Shixiong Kai", "Linkai Hou", "Kaichao Liang", "Hu Zhao", "Yang Zhao", "Shucheng Lin", "Wei Yu", "Chenghan Jiang", "Jingjing Ding", "Jiahui Zhang", "Tian Jin", "Yuhang Zhang", "Dong Guo", "Wei Sun", "Jun Xie", "Jianwei Li", "Lei Cao", "Pei Li", "Jiabin Li", "Jia Yuan", "Rui Yuan", "Jing Zhu", "Mingxuan Yuan", "Zhangcheng Lv", "Xin Jiang", "Xiuhong Fei", "Xiaozhe Ren", "Yulong Li", "Zhipeng Zhang", "Hang Wang", "Zhaohui Xu", "Rui Zhao", "Yibo He", "Xinzhuang Niu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-04", "url": "https://arxiv.org/abs/2606.05868", "pdf_url": "https://arxiv.org/pdf/2606.05868v1", "arxiv_id": "2606.05868", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5df83aefc7ef75e9f29c217c072de2a4bb944d9f7ffa61b3c00591b44f8a6dd8", "sources": ["arxiv", "semantic_scholar"], "title": "CollabSim: A CSCW-Grounded Methodology for Investigating Collaborative Competence of LLM Agents through Controlled Multi-Agent Experiments", "abstract": "Multi-agent systems (MAS) built on large language models have shown growing promise, with their effectiveness resting on agents' ability to coordinate through text-based channels much as human teams do. Yet recent study suggests that MAS often falter not because agents lack individual task-solving ability, but because they lack collaborative competence: the capacity to establish common ground, maintain shared task understanding, balance individual and collective incentives, and repair misalignment as interaction unfolds. Decades of research in Computer-Supported Cooperative Work have characterized these requirements for human teams coordinating under constrained communication, yet existing MAS evaluations focus mainly on task outcomes or single-agent proficiency in reasoning, planning, and tool use. To enable a systematic analysis of agents' collaborative competence in MAS, we introduce CollabSim, a configurable simulation framework that combines a theory-grounded definition of collaborative capabilities, controlled manipulation of interaction conditions, and action-level probing of agents' internal states. Experiments across four LLMs show that CollabSim can capture condition effects, separate model performance patterns, and reveal task-dependent effects of agent design.", "authors": ["Jiaju Chen", "Bo Sun", "Yuxuan Lu", "Yun Wang", "Dakuo Wang", "Bingsheng Yao"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-04", "url": "https://arxiv.org/abs/2606.06399", "pdf_url": "https://arxiv.org/pdf/2606.06399v2", "arxiv_id": "2606.06399", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b01f83b507b02718e0789e4098d33bac3a7760e3ba45687731ec86d23635d19a", "sources": ["arxiv", "semantic_scholar"], "title": "ToolChoiceConfusion: Causal Minimal Tool Filtering for Reliable LLM Agents", "abstract": "Large language model agents increasingly rely on external tools, but larger tool menus can reduce reliability and efficiency by increasing wrong-tool calls, premature actions, and token cost. Existing tool-selection methods often optimize semantic relevance, exposing tools whose names or descriptions match the user request. We argue that relevance is insufficient: a tool may be related to the task while still being unnecessary or premature at the current step. We propose Causal Minimal Tool Filtering (CMTF), a training-free method that selects tools by causal sufficiency. CMTF uses lightweight precondition-effect contracts to expose only the minimal next-step tool frontier needed to advance from the current state toward the user goal. Across multi-step tool-use tasks, we compare CMTF with all-tools exposure, keyword retrieval, state-aware filtering, and causal-path ablations, measuring task success, wrong-tool calls, premature actions, tool exposure, and token cost. In the main benchmark with 102 tasks, 100 tools, four LLM backends, and 2448 task-method-model runs, CMTF matches the strongest causal baseline in aggregate success while reducing visible tools from 100 to one per step and reducing token usage by about 90% relative to all-tools exposure.", "authors": ["Rahul Suresh Babu", "Laxmipriya Ganesh Iyer"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-04", "url": "https://arxiv.org/abs/2606.06284", "pdf_url": "https://arxiv.org/pdf/2606.06284v1", "arxiv_id": "2606.06284", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "854499a642e3f076e703ef4c20a5c9e9b76a6e1c7ff57e2128bf9047e8ef2556", "sources": ["arxiv", "semantic_scholar"], "title": "Do More Agents Help? Controlled and Protocol-Aligned Evaluation of LLM Agent Workflows", "abstract": "Does adding more agents help an LLM workflow once compared systems share the same benchmark loader, tool access, answer contract, usage accounting, and trajectory logging? We introduce BenchAgent, an evaluation framework that places single-agent, fixed multi-agent (MAS), and evolving MAS workflows under one normalized execution and logging protocol. BenchAgent evaluates these substrate-internal workflows across ten reasoning, coding, and tool-use benchmarks with GPT-4.1, and separately reports a Protocol-Aligned External (PAE) GAIA study of a runtime-generated workflow. Under SI conditions, at most one of six tested MAS exceeds the matched single-agent anchor on benchmark-balanced average accuracy: EvoAgent lies within the Wilson one-run guidance, while the remaining five trail by 2.56-11.29 points and occupy more expensive accuracy-cost trade-offs. On the PAE GAIA snapshot, a Claude-Code-style runtime workflow reaches 66.72% overall and 69.23% on Level 3, more than 20 points above the strongest non-Claude baseline, Jarvis, a fixed MAS.", "authors": ["Yuhang Fu", "Ruishan Fang", "Jiaqi Shao", "Huiyu Zheng", "Zhengtao Zhu", "Bing Luo", "Tao Lin"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-04", "url": "https://arxiv.org/abs/2606.05670", "pdf_url": "https://arxiv.org/pdf/2606.05670v1", "arxiv_id": "2606.05670", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/LINs-lab/MASArena/tree/BenchAgent", "venue": null, "quality_score": 0.65}
{"id": "7c7eabd773195d897701dd3192badd6d69ee4bb2a6265661ea5803bb2192dfea", "sources": ["arxiv", "semantic_scholar"], "title": "Autoregressive Diffusion World Models for Off-Policy Evaluation of LLM Agents", "abstract": "Evaluating large language model (LLM) agents in multi-turn interactive environments is expensive and risky, as it requires online environment interaction. We propose ADWM (Autoregressive Diffusion World Model), an evaluation framework that estimates the performance of a new LLM agent policy purely from pre-collected trajectories. The core idea is to learn a latent diffusion world model that simulates how the environment responds to the evaluation policy, without ever executing it in the real environment. Existing diffusion-based OPE methods guide full trajectories in a single pass by jointly diffusing states and actions, an assumption that breaks down for LLM agents whose actions are discrete text that must be sampled from the policy after observing the environment. Unlike autoregressive world models that suffer from compounding errors, ADWM models each transition as an independent denoising process, enabling reliable step-by-step rollouts where the world model and agent alternate in causal order. Crucially, the LLM agent under evaluation directly guides the diffusion generation at each step via a policy-conditioned score function, ensuring that simulated trajectories accurately reflect its decision-making patterns. Empirically, ADWM achieves accurate value estimates and evaluation reliability across diverse multi-turn agent tasks, demonstrating its promise as a practical framework for offline LLM agent evaluation.", "authors": ["Kaixuan Liu", "Guojun Xiong", "Weinan Zhang", "Shengpu Tang"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-04", "url": "https://arxiv.org/abs/2606.05558", "pdf_url": "https://arxiv.org/pdf/2606.05558v1", "arxiv_id": "2606.05558", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "35d10bc510ab3a0e127baf9206fd9469245c3c7008336224b1097ce176debeec", "sources": ["arxiv", "semantic_scholar"], "title": "The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development?", "abstract": "Current AI benchmarks evaluate agents on task execution within human-designed workflows. These evaluations fundamentally fail to measure a critical next-level capability: whether models can autonomously develop agent systems. We introduce the Meta-Agent Challenge (MAC), an evaluation framework designed to test the capacity of frontier models for autonomous agent development. Specifically, a code agent (the meta-agent) is given a sandboxed environment, an evaluation API, and a time limitation to iteratively program an agent artifact that maximizes performance on a held-out test set across five domains. To ensure evaluation integrity, this framework is secured by multi-layer defenses against reward hacking. Leveraging this framework, we demonstrate that meta-agents rarely match human-engineered baseline policies, and the few that do are dominated by proprietary frontier models. Moreover, the design process exhibits high variance, and high optimization pressure surfaces emergent adversarial behaviors like ground-truth exfiltration-highlighting critical deficits in both robustness and model alignment. Ultimately, MAC provides a rigorous, open-source benchmark for autonomous AI research and development, offering an empirical proxy for evaluating recursive self-improvement. Benchmark is publicly available at: https://github.com/ant-research/meta-agent-challenge.", "authors": ["Xinyu Lu", "Tianshu Wang", "Pengbo Wang", "zujie wen", "Zhiqiang Zhang", "Jun Zhou", "Boxi Cao", "Yaojie Lu", "Hongyu Lin", "Xianpei Han", "Le Sun"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-03", "url": "https://arxiv.org/abs/2606.04455", "pdf_url": "https://arxiv.org/pdf/2606.04455v1", "arxiv_id": "2606.04455", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ant-research/meta-agent-challenge", "venue": null, "quality_score": 0.65}
{"id": "cc55af41e4fc6e2f488c87ae84c8c8c7c0d2275fff41113cca285d716915cc1a", "sources": ["arxiv", "semantic_scholar"], "title": "Streaming Communication in Multi-Agent Reasoning", "abstract": "Multi-agent reasoning systems adopt a \"generate-then-transfer\" paradigm that forces end-to-end latency to scale linearly with pipeline depth. We introduce StreamMA, a multi-agent reasoning system that streams each reasoning step to downstream agents as soon as it is generated, pipelining adjacent agents and thus reducing latency. Surprisingly, this pipelining also improves effectiveness: because multi-step reasoning quality is non-uniform and early steps are more reliable than later ones, working with these reliable early steps instead of the full chain prevents error-prone late steps from misleading downstream agents. We formalize both advantages with the first closed-form joint analysis of stream, serial, and single protocols, deriving the effectiveness ordering, speedup upper bound, and cost ratio. Across eight reasoning benchmarks spanning mathematics, science, and code, two frontier LLMs (Claude Opus 4.6 and GPT-5.4), and three topologies (Chain, Tree, Graph), StreamMA outperforms both baselines (avg. +7.3 pp, max +22.4 pp on HMMT 2026; Claude Opus 4.6-high). Beyond these contributions, we discover a \"step-level scaling law\": increasing per-agent steps consistently improves both effectiveness and efficiency, a new scaling dimension orthogonal to and composable with agent-count scaling.", "authors": ["Zhen Yang", "Xiaogang Xu", "Wen Wang", "Cong Chen", "Xander Xu", "Ying-Cong Chen"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-03", "url": "https://arxiv.org/abs/2606.05158", "pdf_url": "https://arxiv.org/pdf/2606.05158v1", "arxiv_id": "2606.05158", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "98119941cdd317489c590fd90cf3a17786d8f3890f75e4c487eb501df2c44617", "sources": ["arxiv", "semantic_scholar"], "title": "GARL: Game-Theoretic Reinforcement Learning for Multi-Agent Strategic Prioritisation", "abstract": "LLM-based multi-agent systems are increasingly used for strategic decision-making tasks. In such settings, performance depends not only on individual model capabilities, but also on the policies by which agents interact and adapt. Multi-agent reinforcement learning can optimise these interaction policies, but its reward design often remains task-specific and weakly grounded in interaction structure. To address this gap, we propose GARL, a GAme-theoretic Reinforcement Learning framework for multi-agent strategic prioritisation. GARL formalises strategic prioritisation as a two-stage game: competing agents first allocate strategic resources over a shared candidate set, and a higher-level arbiter then produces the final ranking. The resulting game-theoretic utilities are converted into role-specific reinforcement signals, allowing policy optimisation to be guided by structured interaction. We instantiate GARL on issues-in-dispute ranking, where the goal is to prioritise core issues in legal proceedings. Experiments show that GARL improves ranking performance, enables small open-source LLMs to become competitive with a strong closed-source LLM under the same candidate-ranking setting, and yields gains in legal-domain competence and broader strategic decision-making. Overall, GARL demonstrates how game-theoretic interaction structure can be turned into reinforcement-learning objectives, providing a principled approach to policy optimisation in multi-agent strategic prioritisation.", "authors": ["Yuxiao Ye", "Yiwen Zhang", "Huiyuan Xie", "Yuqin Huang", "Zhiyuan Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-03", "url": "https://arxiv.org/abs/2606.05002", "pdf_url": "https://arxiv.org/pdf/2606.05002v1", "arxiv_id": "2606.05002", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "bd7a0013cfd95bda84cea4caa3883cd8619fccf4af5a223a9b33f79fdb4b7351", "sources": ["arxiv", "semantic_scholar"], "title": "FORGE: Multi-Agent Graduated Exploitation and Detection Engineering", "abstract": "Vulnerability disclosure volumes now far exceed organizational assessment capacity, yet three adjacent research communities (proof-of-concept generation, vulnerability prioritization, and detection rule engineering) operate largely in isolation. Existing automated exploit generation systems report binary pass/fail outcomes, discarding partial progress and producing no signal for the other two communities. This paper presents FORGE, a multi-agent system that bridges these three silos through graduated exploitation depth. Five specialized agents (Intel, Generator, Planner, Exploit, and Detector) execute in a fixed pipeline that (1) generates targeted vulnerable applications from CVE metadata, (2) conducts coached, multi-turn exploitation assessed by an LLM-primary oracle on a four-level taxonomy (L0: no evidence through L3: full compromise), and (3) produces Sigma and Snort detection rules grounded in OpenTelemetry exploitation traces. Graduated depth is the bridging mechanism: deeper exploitation yields richer behavioral traces for detection engineering, while depth data across scoring bands provides ground truth for prioritization validation. A tiered knowledge architecture accumulates intelligence across assessments, transferring build and exploitation experience to subsequent CVEs. Evaluation on 603 CVEs from the CVE-GENIE dataset achieves 67.8% end-to-end L1+ exploitation at USD 1.50 per CVE across eight languages and 187 CWE types. Exploitation rates remain near 68% regardless of EPSS or CVSS band, indicating that pattern-level reachability is orthogonal to metadata-based prioritization. Detection rules from L2+ exploitation achieve significantly higher span-normalized grounding than L1-derived rules (p=0.035), and 93.4% of generated Snort rules produce zero false positives against a synthetic benign corpus.", "authors": ["Farooq Shaikh"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.03453", "pdf_url": "https://arxiv.org/pdf/2606.03453v1", "arxiv_id": "2606.03453", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "ed246ae2d4d53d6c2fe32947a8e90375a713eb4d58e84c8b0ca5b9ad0be45988", "sources": ["arxiv", "semantic_scholar"], "title": "Multi$^2$: Hierarchical Multi-Agent Decision-Making with LLM-Based Agents in Interactive Environments", "abstract": "A central goal of large language model (LLM) research is to build agentic systems that can plan, act, and adapt through sustained interaction with dynamic environments. While recent LLM-based agents exhibit impressive contextual reasoning, their long-horizon decision-making remains fragile, often suffering from objective drift, where goals and plans drift over extended interactions. We introduce Multi$^2$, a hierarchical multi-agent decision-making framework that explicitly decomposes agent behavior into complementary roles. A high-level agent (System 1) focuses on context-aware sub-goal generation using supervised fine-tuning (SFT), while a low-level agent (System 2) executes atomic actions through offline-to-online reinforcement learning (RL) in interactive environments. This separation enables stable long-horizon control, mitigates objective drift, and allows efficient adaptation. Across diverse interactive environments, Multi$^2$ consistently outperforms strong agentic baselines, demonstrating improved robustness and coordination in multi-turn interaction. Beyond performance, we introduce and release three hierarchical benchmark datasets, filling a long-standing gap in training and evaluating hierarchical decision-making for LLM-based agents.", "authors": ["Sangeun Park", "Minhae Kwon"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.03698", "pdf_url": "https://arxiv.org/pdf/2606.03698v1", "arxiv_id": "2606.03698", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "907ca1e93712ec801982a7a54cb468b696ccaabe600e27983715f1d91ed7327d", "sources": ["arxiv", "semantic_scholar"], "title": "The Deliberative Illusion: Diagnosing Factual Attrition and Stance Homogenization in Multi-Agent LLM Deliberation", "abstract": "Multi-agent LLM systems often treat consensus as evidence of successful interaction. For deliberative problems, however, reliability depends on whether agents preserve the facts and viewpoints needed to interpret an issue. We identify the deliberative illusion: discussion produces (1) factual attrition, the progressive loss of issue-critical facts, alongside (2) stance homogenization, the collapse of diverse positions toward consensus. To measure this process, we introduce DelibTrace, a framework that decomposes each issue into atomic facts, labels issue-critical ones, distributes them across agents, and tracks their survival across discussion rounds. Across ethical and news-based deliberation with three representative LLM families, multi-agent discussion erases up to 72% of issue-critical facts. This loss is consequential: retained evidence can reconstruct the issue misleadingly, final stances remain anchored in base-model priors, and a single malicious agent can inject misinformation into the shrinking shared context. These results reveal a sharper risk: agents can agree more while knowing less. We call for evaluations that measure which facts, uncertainties, and legitimate disagreements survive interaction.", "authors": ["Herun Wan", "Jiaying Wu", "Minnan Luo", "Fanxiao Li", "Ningnan Wang", "Nancy F. Chen", "Min-Yen Kan"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.03032", "pdf_url": "https://arxiv.org/pdf/2606.03032v1", "arxiv_id": "2606.03032", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0e49aa46c099be2af4a2920dfe4630d23cda113b2063e1db5abdcc837871915a", "sources": ["arxiv", "semantic_scholar"], "title": "ToolGate: Token-Efficient Pre-Call Control for Tool-Augmented Vision-Language Agents", "abstract": "Tool-augmented vision-language agents can acquire external perceptual evidence through OCR, detection, segmentation, and other tools, but executing every proposed tool call is costly and sometimes unnecessary. We study the pre-call control problem: after a ReAct-style VLM agent proposes a perceptual tool call, should the call be executed, or skipped before its output enters the context? Across five benchmarks, we find that the baseline agent exhibits poor local selectivity: helpful and harmful calls occur at similar rates (11.8% vs. 9.9%), while most calls do not change the immediate forced-answer prediction. We introduce ToolGate, a lightweight external controller that predicts execute/skip decisions from trajectory text and simple structural features. Across two Qwen3-VL backbones, ToolGate reduces token cost to 64-69% of the unrestricted ReAct baseline while preserving average accuracy in cross-domain settings. With matched-domain trajectory training on Qwen3-VL-30B, it further improves average accuracy by 1.65 points. These results show that tool-augmented VLM agents benefit not only from better perceptual tools, but also from explicit control over when tool outputs are worth paying for.", "authors": ["Anjie Liu", "Yan Song", "Zhixun Chen", "Ziqin Gong", "Zhongwei Yu", "Jun Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.03054", "pdf_url": "https://arxiv.org/pdf/2606.03054v1", "arxiv_id": "2606.03054", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d9737ea0f3ebd22bbe26aa31d16a0c1783d99f683984648c4940027ef25ad7d8", "sources": ["arxiv", "semantic_scholar"], "title": "SMAC-Talk: A Natural Language Extension of the StarCraft Multi-Agent Challenge for Large Language Models", "abstract": "As LLMs become more widely deployed, they are increasingly expected to work alongside other AI agents rather than operating in isolation. Effective coordination in these settings requires agents to communicate, share information and make decisions under uncertainty. We introduce SMAC-Talk, a natural language extension of the StarCraft Multi-Agent Challenge for evaluating LLM-based agents in cooperative multi-agent environments. The environment has several key features such as decentralized control, partial observability and long-horizon decision making. SMAC-Talk includes a natural language communication channel which is used to probe agent coordination and trust. We use this communication channel to construct different evaluation scenarios, including settings with an embedded deceptive communicator that tries to disrupt and deceive allies through communication alone. We provide three agents for benchmarking using 4 models from the Qwen3.5 family and study how reasoning structure, memory and model scale affect coordination between agents. We release SMAC-Talk as an open benchmark to support the research community in developing and evaluating LLM agents in cooperative multi-agent settings.", "authors": ["Joel Sol", "Homayoun Najjaran"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.04202", "pdf_url": "https://arxiv.org/pdf/2606.04202v1", "arxiv_id": "2606.04202", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "48cc30f772df1d314fdc41e555f1849aff7edeea092fda30080b4a013836b600", "sources": ["arxiv", "semantic_scholar"], "title": "A Training-Free Mixture-of-Agents Framework for Multi-Document Summarization using LLMs and Knowledge Graphs", "abstract": "Multi-Document Summarization (MDS) plays a critical role in distilling essential information from collections of textual data. Existing approaches often struggle to capture complex inter-document relationships, rely heavily on large amounts of labeled data for supervised training, or exhibit limited generalization across domains and languages. To address these limitations, we present a training-free mixture-of-agents framework for MDS that leverages the complementary strengths of large language models (LLMs) and knowledge graphs. Our approach decomposes summarization into specialized agent tasks: extractive selection, knowledge-aware abstraction, and iterative refinement, each operating without task-specific fine-tuning. We unify their outputs using a multi-perspective consistency mechanism guided by LLMs. Experiments across four datasets in English and Vietnamese demonstrate state-of-the-art or competitive performance, validating the effectiveness and adaptability of our modular design.", "authors": ["Cuong Vuong Tuan", "Trang Mai Xuan", "Tien-Cuong Nguyen", "Vu-Duc Ngo", "Thien Van Luong"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-02", "url": "https://arxiv.org/abs/2606.03867", "pdf_url": "https://arxiv.org/pdf/2606.03867v1", "arxiv_id": "2606.03867", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "214ac628a6bd2a5d5e5a7422b5fa760739a94e0780394d8a2a70d3f79b07f47c", "sources": ["arxiv", "semantic_scholar"], "title": "MOC: Multi-Order Communication in LLM-based Multi-Agent Systems", "abstract": "Despite the remarkable progress of Large Language Model (LLM) based Multi-Agent Systems, most research focuses on optimizing coordination topology while largely underexploring the equally critical problem: how to transmit and optimize messages among agents effectively? Current communication schemes typically rely on the direct concatenation of first-order neighbor responses, which induces a restricted evidence receptive field and leads to the dilution of crucial insights over multi-hop paths. To address these limitations, we propose the Multi-Order Communication (MOC) scheme, which reconstructs the inter-agent communication to capture multi-hop dependencies and incorporates a structural message consolidation strategy to ensure efficiency. Specifically, we formalize the communication mechanism to construct a structured multi-order evidence stream, and subsequently design a Semantic-Topological Merging algorithm to optimize semantic fidelity within token constraints. Extensive experiments across six diverse datasets and LLM backbones of varying parameter scales demonstrate that MOC consistently improves task performance and reduces communication costs. The code is available at https://github.com/yao-guan/MOC.", "authors": ["Yao Guan", "Lin Wang", "Zhihu Lu", "Ziyi Wang", "Wenzhu Yan", "Qiang Duan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-01", "url": "https://arxiv.org/abs/2606.02359", "pdf_url": "https://arxiv.org/pdf/2606.02359v1", "arxiv_id": "2606.02359", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/yao-guan/MOC", "venue": null, "quality_score": 0.65}
{"id": "a7705dc96fcfae5ffb9dae0c9eed59a5703751f72de4b27bd94cb03fbf2f7ba1", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Computer Use", "abstract": "Computer use agents (CUAs) today are primarily deployed as single serial agents. This setup is suboptimal for complex long-horizon tasks that benefit from task decomposition, parallel execution, and consistent re-planning based on new information. In this paper, we argue that we should instead move towards evaluating and building multi-agent computer use (MACU) systems. These systems, which emphasize planning and parallel execution, alleviate many of the shortcomings of single-agent CUAs. We propose a general multi-agent setup in which a manager model decomposes computer use tasks as a directed acyclic graph (DAG), encoding relevant dependencies and goals for subagents. At each iteration, the manager dispatches parallel CUA subagents to carry out nodes on the ready frontier of the DAG, and continuously revises the DAG (adding, canceling, or rewriting nodes) as new findings arrive from subagents. This design treats the partially observable environment of computer use as a first class challenge: information that downstream agents may not be able to re-observe are retained and passed forward through the manager and DAG structure. We demonstrate that MACU consistently improves over strong single-agent baselines by $3.4-25.5\\%$ on desktop (OSWorld) and web navigation (Online-Mind2Web, WebTailBench, Odysseys) benchmarks, exhibits more favorable test-time scaling, and solves complex long-horizon tasks where single-agent CUAs get stuck. On Odysseys, a long-horizon web navigation benchmark, MACU improves average task completion wall-clock time by ${\\sim} 1.5 \\times$, demonstrating its efficacy in speeding up traditionally slow CUA pipelines. Our findings highlight that multi-agent coordination is a promising axis for scaling computer use agents to work productively for longer and more effectively. We release all code and interactive visualizations at https://jykoh.com/multi-agent-computer-use.", "authors": ["Jing Yu Koh", "Ruslan Salakhutdinov", "Daniel Fried"], "categories": ["cs.MA", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-01", "url": "https://arxiv.org/abs/2606.01533", "pdf_url": "https://arxiv.org/pdf/2606.01533v1", "arxiv_id": "2606.01533", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "eaf5bba48e2db025c65ddd789de2b1cf4f7599e3504533e0851b69b952944cd1", "sources": ["arxiv", "semantic_scholar"], "title": "POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems", "abstract": "Orchestrating Large Language Models into Multi-Agent Systems (LLM-MAS) has unlocked remarkable reasoning capabilities, yet emergent failures and hallucinations that resist characterisation block their deployment in safety-critical domains -- a gap made legally untenable by emerging AI regulation. Existing evaluation paradigms share a common flaw: centralised judgment creates single points of failure and demands domain-specific expertise. Here we present POIROT, a protocol that repurposes a system's own agents as its diagnostic layer, leveraging the epistemic diversity already present in the architecture. Across evaluated settings, POIROT outperforms single-LLM evaluator baselines, with gains that scale with problem complexity (OR = 1.60, $p = 0.008$), agent count, and fault dimensionality, persisting under compound fault conditions. These results demonstrate that safety oversight need not be externalised: the agents executing a role carry sufficient collective intelligence to audit it. We release POIROT as an open-source library alongside BLAME, a benchmark for fault attribution in safety-critical multi-agent systems.", "authors": ["Iñaki Dellibarda Varela", "R. Sendra-Arranz", "Pablo Romero-Sorozabal", "J. M. Valverde-García", "Annemarie F. Laudanski", "Álvaro Gutiérrez", "Eduardo Rocon", "Manuel Cebrian"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-01", "url": "https://arxiv.org/abs/2606.02282", "pdf_url": "https://arxiv.org/pdf/2606.02282v1", "arxiv_id": "2606.02282", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "eed585a2aabc91de5dcd1d061fc5fbe870902cd626161b3cbcb46d9e567817fd", "sources": ["arxiv", "semantic_scholar"], "title": "Ghost Tool Calls: Issue-Time Privacy for Speculative Agent Tools", "abstract": "Tool-augmented language agents speculatively issue likely future tool calls to hide latency, but those calls leak inferred user intent to external services before the agent commits to the branch. Every external observer that received the call retains the disclosure after the agent abandons the branch. Timing is the issue, not authorization: no commit-time cleanup, read-only restriction, or access-control allow-list unsends what an observer already holds. We call these invocations ghost tool calls and propose Speculative Tool Privacy Contracts, a runtime abstraction that treats observation before commitment as a first-class effect, distinct from state mutation. We implement the contracts in a prototype runtime and evaluate twelve policies across three corpora. Speculative dispatch increases what an observer can infer about user intent; post-hoc filters, read-only restrictions, and access-control allow-lists leave that inference intact; only issue-time policies that change or suppress the speculative call's argument or destination projection before dispatch reduce it.", "authors": ["Bardia Mohammadi", "Lars Klein", "Akhil Arora", "Laurent Bindschaedler"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-06-01", "url": "https://arxiv.org/abs/2606.02483", "pdf_url": "https://arxiv.org/pdf/2606.02483v1", "arxiv_id": "2606.02483", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f885bfe045e03564f1f2248f09226dcbe35a0ac3954e293e7f2b5d3e07331e6b", "sources": ["arxiv", "semantic_scholar"], "title": "Early Diagnosis of Wasted Computation in Multi-Agent LLM Systems via Failure-Aware Observability", "abstract": "Tool-using multi-agent large language model (LLM) systems spend computation through model tokens, tool calls, retries, and code execution before producing an answer. When a run fails, final-answer evaluation reveals the endpoint but usually not the point at which the trajectory stopped making recoverable progress. This paper introduces a failure-aware observability framework for diagnosing wasted computation in multi-agent LLM traces. The framework maps recurring failure modes to online trace signals, including tool reliability, execution recovery, orchestration loops, evidence availability, information change, and budget pressure. We instantiate the framework in a three- agent question-answering system and evaluate it on 165 GAIA validation traces under identical execution caps. Operational failures remain common: 22/53 level-1 runs, 33/86 level-2 runs, and 12/26 level-3 runs fail to produce a usable final answer. The traces expose different mechanisms behind these outcomes, including insufficient evidence, repeated-action loops, max-step termination, tool-failure streaks, and execution calls that succeed without useful output. Mean token use rises from 8,152 tokens at level 1 to 16,389 tokens at level 3, while evidence availability and sentence-level support diverge. A cached 10-trace LLM-judge grounding audit shows that cheap online signals and deeper semantic metrics capture complementary layers of failure. The results position failure-aware observability as a diagnostic layer between raw execution logs and final-answer accuracy.", "authors": ["Xianyou Li", "Weiran Yan", "Yichao Wu", "Penghao Liang", "Mengwei Yuan", "Jianan Liu", "Jing Yang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-31", "url": "https://arxiv.org/abs/2606.01365", "pdf_url": "https://arxiv.org/pdf/2606.01365v1", "arxiv_id": "2606.01365", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "60d2d7a637b17cc96b24e10aa6b12b2784d45ab3b1d12ecf02f932e25124535f", "sources": ["arxiv", "semantic_scholar"], "title": "FinCom: A Financial Multi-Agent Demo with Disagree-or-Commit Deliberation", "abstract": "Multi-agent systems powered by large language models (LLMs) are increasingly used for financial analysis and decision support. However, existing coordination schemes, especially those emphasizing consensus or debate, are vulnerable to sycophancy: agents conform to peer reasoning instead of evidence, leading to premature agreement and degraded outcomes. We introduce FinCom (Financial Committee), a governed multi-agent framework and interactive system that operationalizes the Disagree-or-Commit (DoC) protocol to embed structured dissent into financial AI committees. A central Supervisor orchestrates three ReAct-enabled specialist agents: Research, Quantitative, and Risk. Each agent is equipped with role-specific tools for retrieval, computation, and stress testing. During deliberation, agents must either explicitly critique or commit to their peers' reasoning before converging on a unified recommendation. This demonstration showcases how FinCom supports committee-style financial analysis through coordinated multi-agent interaction, including structured report generation and interactive decision support. Evaluated across the most recent financial agent benchmark, in addition to 90 internal handcrafted financial tasks using an LLM-as-a-Judge protocol, DoC improves reasoning accuracy and risk awareness significantly over a consensus-seeking baseline on both an in-house and external evaluation set. By reframing disagreement as a governance primitive rather than noise, FinCom offers a lightweight, prompt-only recipe for improving accountability, transparency, and epistemic robustness in agentic financial systems.", "authors": ["Chao Peter Yang", "Zixiao Tan", "Kaisen Yao", "Ziyu Zhou", "Eleanor Jiang", "Michael Wu"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-31", "url": "https://arxiv.org/abs/2606.00939", "pdf_url": "https://arxiv.org/pdf/2606.00939v1", "arxiv_id": "2606.00939", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1760d30ff7ffe62b392ebbdcfea4b45c9d49edbd5a175fef9769e277b0b5bc10", "sources": ["arxiv", "semantic_scholar"], "title": "RAG-driven Multi-Agent LLM Framework with Task Decomposition for Beyond 5G Auto-Configuration", "abstract": "While Large Language Models (LLMs) offer a promising path toward intent-driven network management by translating natural language human intents into machine-readable configurations, they often suffer from hallucinations and structural inconsistencies in multi-step and complex tasks. To address these challenges, this paper proposes a retrieval-augmented and task decomposition-based multi-agent LLM framework for Beyond 5G network auto-configuration. The framework employs a semantic retrieval-augmented generation pipeline to ensure that its outputs are aligned with technical standards and vendor-specific manuals. Furthermore, it introduces a modular architecture for configuration generation, closed-loop configuration verification, and network deployment, in which complex tasks are decomposed into smaller sub-tasks handled by specialized agents. In this architecture, hallucinated configuration parameters are identified by the configuration verifier agent and corrected through low computational segment-level regeneration. The performance evaluation experiments with the OpenAirInterface emulator demonstrate that the proposed task decomposition-based configuration and verification approach improves the average success rate by 22.7% over monolithic methods, achieving 94.4% success in network configuration.", "authors": ["İrşat Emin Sarıdaş", "Onur Salan", "Ali Görçin", "Ibrahim Hokelek", "Hakan Ali Çırpan"], "categories": ["eess.SP"], "fields_of_study": ["Engineering"], "published_date": "2026-05-31", "url": "https://arxiv.org/abs/2606.01222", "pdf_url": "https://arxiv.org/pdf/2606.01222v1", "arxiv_id": "2606.01222", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "061de5b4a7532b3ac5a8d07f2495caa1d2e1e6b6c3b1269b02daed603ffbdbe3", "sources": ["arxiv", "semantic_scholar"], "title": "CAREAgent: Clinical Agent with Structured Reasoning and Tool-Integrated for Order Generation", "abstract": "Clinical order generation serves as a critical bridge between clinical decision-making and real-world practice, translating medical decisions into concrete and executable orders. Existing agents mainly focus on coarse-grained decisions and overlook the fine-grained, executable information required for clinical orders. To address this gap, we propose CAREAgent, an agent for clinical order generation. To support its training, we introduce a two-stage agentic reasoning data construction method. First, we design an agent framework that constructs verifiable reasoning trajectories aligned with realistic clinical tool usage. Second, we filter reasoning trajectories by format compliance, order validity, and clinical plausibility. Building on the constructed data, the model is first trained via supervised fine-tuning to acquire fundamental reasoning formats and medical knowledge, and is subsequently optimized through reinforcement learning with multi-dimensional reward functions to enhance complex clinical reasoning capabilities. Experiments on multiple benchmarks demonstrate the effectiveness of CAREAgent. On ClinicalBench (unseen during training), CAREAgent improves the F1 score by 5.05%, 2.09%, and 0.86% over the single-agent, multi-agent, and agentic reasoning methods, respectively.", "authors": ["Ruihui Hou", "Ziyue Huai", "Chennuo Zhang", "Ziyan Liu", "Siran Zhao", "Yao Yu", "Jie Zhai", "Tong Ruan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-31", "url": "https://arxiv.org/abs/2606.01094", "pdf_url": "https://arxiv.org/pdf/2606.01094v1", "arxiv_id": "2606.01094", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2e28d4c8ab8557a63792fc114dfd009bbbe192dd6abe2913f6d55905120cf886", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling Behavior of Single LLM-Driven Multi-Agent Systems", "abstract": "The burgeoning field of LLM-based Multi-Agent Systems (MAS) promises to tackle complex tasks through collaborative intelligence, yet fundamental questions regarding their scaling behavior and intrinsic collective dynamics remain underexplored. This paper systematically investigates how the performance of a homogeneous MAS evolves as the number of agents increases, isolating the variable of collaboration from model or knowledge heterogeneity. We propose the Sequential Iterative Multi-Agent System (SIMAS) framework, a minimalist architecture centered on sequential inter-agent communication, to clearly observe scaling effects. Through extensive experiments across diverse tasks and model scales, we establish that MAS performance does not scale monotonically with agent count but follows a pattern of diminishing returns, governed by a trade-off between collaborative synergy and coordination overhead. Our findings reveal that effective MAS requires a sufficiently capable base LLM, that task type critically modulates the optimal agent count, and that collective intelligence is an emergent property contingent on strategic interaction design rather than a guaranteed outcome of agent plurality. The performance degradation stems coordination overhead rather than merely long-context failure, and the scaling tendency generalizes across interaction architectures like structured debate topologies. This work provides a foundational understanding of MAS scaling laws, offering practical guidance for designing efficient collaborative systems and challenging the prevailing assumption that more agents invariably lead to better performance.", "authors": ["Jialing Li", "Zhouhong Gu", "Yin Cai", "Hongwei Feng"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-30", "url": "https://arxiv.org/abs/2606.00655", "pdf_url": "https://arxiv.org/pdf/2606.00655v1", "arxiv_id": "2606.00655", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "033102c6d4bde68537276e29358d9dba384232f01d9aba0d22d754713f693ccd", "sources": ["arxiv", "semantic_scholar"], "title": "FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search", "abstract": "LLM-based agents increasingly solve complex tasks through long trajectories involving reasoning steps, tool calls, and inter-agent communication. However, when these agents fail, it is often unclear which agent caused the failure and which step introduced the decisive error. This attribution problem is challenging because mistakes can propagate across the trajectory: later actions may appear incorrect, but only because they depend on an earlier corrupted state. Therefore, failure attribution cannot be treated as independent step-level classification. We propose FALAT, a diagnostic framework for failure attribution in LLM agent trajectories. FALAT frames attribution as a dependency-guided search problem. It first constructs an expectation of how the task should be solved and uses this expectation to identify suspicious regions in the trajectory. It then traces dependencies among decisions, tool outputs, and agent messages to distinguish error-introducing steps from steps that merely inherit or propagate prior mistakes. Finally, FALAT evaluates whether correcting a candidate step would be sufficient to recover the expected outcome, allowing it to identify both the responsible agent and the decisive failure step. We evaluate FALAT on the Who&When benchmark, which includes both algorithm-generated and hand-crafted multi-agent failure trajectories. The results show that FALAT consistently improves responsible-agent and decisive-step attribution. Its best configurations achieve 46.0% step-level accuracy on algorithm-generated trajectories and 29.1% on the more challenging hand-crafted trajectories, outperforming specialized attribution baselines and direct prompting with standalone LLMs. These findings suggest that dependency-aware reasoning is essential for reliable failure diagnosis in LLM agent systems.", "authors": ["Md Nakhla Rafi", "Md Ahasanuzzaman", "Dong Jae Kim", "Zhijie Wang", "Tse-Hsun Chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-30", "url": "https://arxiv.org/abs/2606.00765", "pdf_url": "https://arxiv.org/pdf/2606.00765v1", "arxiv_id": "2606.00765", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7b5bf44305ea75c39b6a396737e311e99fa30b92251fc32e114e68b2d9a10252", "sources": ["arxiv", "semantic_scholar"], "title": "Doing What They Say, Not What They Reason: Locating the Faithfulness Gap in LLM Agents", "abstract": "Do LLM agents act on the reasoning they state? This question of process fidelity is central to using LLMs in social simulation, yet it is hard to measure where no reference for correct behavior exists. We study it in acontrolled setting, a Texas Poker simulator with a verifiable reference action for every decision by decomposing the faithfulness gap into two steps: reasoning-conclusion and conclusion-action. The two steps behave oppositely.", "authors": ["Yufeng Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-30", "url": "https://arxiv.org/abs/2606.00476", "pdf_url": "https://arxiv.org/pdf/2606.00476v1", "arxiv_id": "2606.00476", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "24bdd4d7e33f82b77dbc25f96a02b10567101a67206ae0ef6b44dbfcf142433d", "sources": ["arxiv", "semantic_scholar"], "title": "Not All Flips Are Conformity: Decomposing Stance Convergence in Multi-Agent LLM Debate", "abstract": "Multi-agent debate (MAD) is a promising strategy for improving LLM reasoning, but when agents converge on a shared answer, it is unclear whether that convergence reflects genuine deliberation or social compliance. We show that the conventional answer flip rate conflates three distinct mechanisms: spontaneous instability, stance-induced conformity, and reasoning-induced persuasion. Our three-source decomposition framework isolates each through controlled counterfactual conditions. In the primary MMLU-Pro setting, 37% of agent-question observations change under self-reflection alone, while robustness tests show substantial model-dependent instability across GPQA-Diamond and three model families; strict conformity is 29% in the primary setting and remains predominantly harmful across model replications (57-77% correct-to-wrong). A controlled information-gradient experiment reveals that even vacuous reasoning is associated with 20-39% error adoption among resistant agents, with reasoning-like presentation carrying substantial persuasive weight. Harmful conformity can be predicted from Round 0 features (AUC = 0.79), and risk-targeted intervention reduces it by 13.6 percentage points (p < 0.001). However, without correctness labels or self-reflection controls, reducing peer adoption does not improve accuracy, because harmful and beneficial influence cannot be distinguished.", "authors": ["Xiqi Hao", "Zengqing Wu", "Yu-Xuan Qiu", "Chuan Xiao", "Ruiqi Xu", "Shuyuan Zheng", "Jianbin Qin"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-30", "url": "https://arxiv.org/abs/2606.00820", "pdf_url": "https://arxiv.org/pdf/2606.00820v1", "arxiv_id": "2606.00820", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "29e365ea4af0526725d00b7dca82347be015bc98a8f07f08b56fd9f78af1c23a", "sources": ["arxiv", "semantic_scholar"], "title": "MAVEN: Improving Generalization in Agentic Tool Calling", "abstract": "Generalization across agentic tool-calling environments remains a central challenge for reliable agentic reasoning systems. Although large language models achieve strong results on individual benchmarks, their ability to compose reasoning strategies, preserve intermediate states, and coordinate tools across domains remains underexplored. We present MAVEN (Modular Agentic Verification and Execution Network), a lightweight symbolic reasoning scaffold for structured decomposition, adaptive tool orchestration, and intermediate verification. We evaluate MAVEN across established tool-calling benchmarks, including BFCL v3, TauBench, Tau2Bench, AceBench, and introduce MAVEN-Bench, a stress-test benchmark for multi-step mathematical and physical reasoning with explicit verification and adversarial task composition. MAVEN-Bench exposes a substantial gap between partial reasoning quality and end-to-end task success; in direct MAVEN-Bench runs, MAVEN improves its GPT-OSS-120b base model from 48% to 71% accuracy without additional training. It also remains competitive with frontier proprietary baselines while using an open-weight backbone with an estimated cost ratio of roughly 1/10, suggesting that lightweight verification-centered scaffolds can strengthen compositional reasoning and motivate more process-aware evaluation of agents in the wild.", "authors": ["Omkar Ghugarkar", "Vishvesh Bhat", "Muhammad Ahmed Mohsin", "Asad Aali"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-29", "url": "https://arxiv.org/abs/2605.30738", "pdf_url": "https://arxiv.org/pdf/2605.30738v1", "arxiv_id": "2605.30738", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9f12b0c27effb05084be4a97206072e5091f45b9b3f3c724a30b83aa59625311", "sources": ["arxiv", "semantic_scholar"], "title": "Depth-Dependent Indirect Prompt Injection in Tool-Calling ReAct Agents: Injection Depth, Payload Framing, and Turn-Budget Sensitivity", "abstract": "ReAct agents that interleave chain-of-thought reasoning with tool calls are increasingly deployed for real tasks such as scheduling, file retrieval, and data access. Their tool observation loop creates a direct attack surface: an adversary who controls any tool's return value can embed instructions that redirect the agent away from the user's goal, a threat known as indirect prompt injection. Existing benchmarks evaluate attack success rate (ASR) at a fixed injection position under fixed conditions, leaving three risk dimensions unexplored: where in the tool sequence the payload appears (injection depth), what rhetorical register it uses (framing), and how many turns the agent is permitted (turn cap). We conduct four controlled studies on 20 scenarios spanning five attack categories, totalling 460 trials against GPT-4o-mini and Claude Haiku at a combined API cost under 0.36 USD. Study 1 shows that ASR against GPT-4o-mini decays from 60% at depth 1 to 0% at depths 4 and 5 (Cramer's V = 0.58, p < 0.001; restricted to within-sequence depths 1-3: V = 0.47, p = 0.0013), driven by model resistance at depth 1 and task completion before payload encounter at deeper positions. Study 2 replicates the depth experiment on Claude Haiku, which achieves 0% ASR at every depth through a combination of conservative tool invocation and genuine instruction resistance. Study 3 shows that framing modulates ASR between 25% (neutral) and 75% (persona) at depth 1, a 50-percentage-point range that does not reach statistical significance at N = 20 per condition. Study 4 confirms that ASR is stable across turn caps of 3, 5, and 7, indicating the turn budget is not a risk factor in this setting. Our results establish injection depth as the dominant variable and show that sanitising only the first tool observation captures 67% of measured injection successes.", "authors": ["Mohammadreza Rashidi"], "categories": ["cs.CR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-29", "url": "https://arxiv.org/abs/2605.30686", "pdf_url": "https://arxiv.org/pdf/2605.30686v1", "arxiv_id": "2605.30686", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0dd89d0eb24c2486f19864db399008f4631087bf819aa48e742671ee9000ee61", "sources": ["arxiv", "semantic_scholar"], "title": "Counterfactual Graph for Multi-Agent LLM Calibration", "abstract": "Multi-agent LLM systems often treat agreement as evidence: when many agents in a panel give the same answer, that answer is assumed to be more reliable. We show that this assumption can fail after agents communicate. Communication can induce correlated failures and false consensus, so the same vote share may reflect reliable agreement in one topology but over-confidence in another. We propose CAGE-CAL, a counterfactual agent-graph calibration framework for multi-agent LLMs. For each query, CAGE-CAL compares an observed post-communication agent graph with a matched counterfactual no-communication graph, capturing both pairwise failure correlations and group-level dependencies. Rather than simply counting how many agents agree, CAGE-CAL estimates the counterfactual shift between observed and no-communication dependence, and calibrates confidence accordingly. Across five benchmarks, CAGE-CAL improves reliability discrimination with competitive ECE, and its calibrated confidence further improves topology selection over the best fixed-topology strategy.", "authors": ["Jiatan Huang", "Mingchen Li", "Ziming Li", "Sunjae Kwon", "Hong Yu", "Chuxu Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.30653", "pdf_url": "https://arxiv.org/pdf/2605.30653v1", "arxiv_id": "2605.30653", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "75b0962d1ccb4d37b3c8f9c842e7ac691eddeef03446ccacd50a93d1c684e61b", "sources": ["arxiv", "semantic_scholar"], "title": "CONCAT: Consensus- and Confidence-Driven Ad Hoc Teaming for Efficient LLM-Based Multi-Agent Systems", "abstract": "Although large language model (LLM) based multi-agent systems (MAS) show their capability to solve complex tasks and achieve higher performance over single agent systems, they lead to huge computational overheads because of heavy communication between agents. Previous research has made efforts to train a sparse multi-agent graph or fine-tune a planner to orchestrate the workflow better. However, such extra training processes introduce computational costs and limit MAS to specific domains, therefore compromising their generalizability. In this paper, we propose CONCAT, a training-free multi-agent collaboration framework based on CONsensus and Confidence-driven Ad hoc Teaming to efficiently organize agent interactions. Specifically, agents are clustered based on their initial answers, and leaders of each cluster are selected based on the agents' confidence. Then, a heuristic function based on the Theory of Mind is designed to predict the collaboration benefits between every two leaders according to their answers and confidence. Finally, an ad hoc multi-agent network is organized after evicting a percentage of communications based on the predicted benefits. Experiments across three LLMs and three benchmarks show that CONCAT achieves up to 2.02x higher efficiency (accuracy/latency ratio) than LLM-Debate and outperforms training-aware methods such as AgentDropout, while reducing average latency by 50.1% on Qwen2.5-14B-Instruct, without any task-specific training.", "authors": ["Ziyang Ma", "Dingyi Zhang", "Sichu Liang", "Jiajia Chu", "Pengfei Xia", "Hui Zang", "Deyu Zhou"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29612", "pdf_url": "https://arxiv.org/pdf/2605.29612v1", "arxiv_id": "2605.29612", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "981d9c2151d336d5b99823916f954bc5486d7b2651716f2fcb8981b3aa7595e3", "sources": ["arxiv", "semantic_scholar"], "title": "PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration", "abstract": "LLM multi-agent systems often coordinate through natural-language dialogue or loosely structured shared memory, making intermediate state difficult to validate, attribute, and audit. We introduce PatchBoard, a schema-grounded collaboration architecture that replaces inter-agent dialogue with validated JSON Patch mutations over a shared structured state. An Architect agent constructs a task-specific schema and workflow rules, while a deterministic kernel validates each proposed state mutation against schema constraints, role-specific write contracts, and runtime invariants before committing it transactionally. On 630 matched ALFWorld episodes, PatchBoard achieves an 84.6% success rate, compared with 30.8% for LangGraph and 61.6% for Flock, while reducing tokens per successful task to 45.5k, compared with 368.3k and 64.2k, respectively.", "authors": ["Shuyu Zhang", "Yaqi Shi", "Lu Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29313", "pdf_url": "https://arxiv.org/pdf/2605.29313v1", "arxiv_id": "2605.29313", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "388efdadf696966dac598bdec8d2e4596750ad6c8bb875561a1903d8e2f07bd0", "sources": ["arxiv", "semantic_scholar"], "title": "On Effectiveness and Efficiency of Agentic Tool-calling and RL Training", "abstract": "Tool-calling is a central component of modern large language model (LLM) agents, equipping them with skills beyond their parametric knowledge. This paper studies tool-calling along two complementary axes: effectiveness, i.e., how this capability is measured, and efficiency, i.e., how it is learned. On effectiveness, we systematically analyze tool-calling evaluation pipelines and show that results can be highly sensitive to seemingly minor, often undocumented implementation choices including the random seed, system prompt, multi-turn template construction, and how prior interaction/reasoning history is carried forward. These choices can lead to substantial differences in reported performance, especially in multi-turn settings where without rigorous standardization, leaderboard rankings are unreliable. On efficiency, we examine standard reinforcement learning (RL) for tool-calling and identify two sources of computational waste: (i) during rollouts, many prompts produce no learning signal, and (ii) during policy updates, optimization incurs high computational cost. Guided by these findings, we introduce two techniques that accelerate RL-based tool-calling training, achieving substantial wall-clock speedup without degrading performance.", "authors": ["Tong Liu", "Cheng Qian", "Matej Cief", "Yuan He", "Daniele Dan", "Nikolaos Aletras", "Gabriella Kazai"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2606.00135", "pdf_url": "https://arxiv.org/pdf/2606.00135v1", "arxiv_id": "2606.00135", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "afdf544749a8553785449ac0db79f315c6451c458ab87a5478416230ef1eee28", "sources": ["arxiv", "semantic_scholar"], "title": "SkillBrew: Multi-Objective Curation of Skill Banks for LLM Agents", "abstract": "Retrieval-augmented LLM agents increasingly rely on curated skill banks: collections of reusable textual principles that guide decision making on complex tasks. Existing approaches typically expand these banks in an append-only fashion, continuously adding new skills without removing redundant, outdated, or harmful ones, resulting in inefficient and poorly curated repositories. In this paper, we formulate the skill bank curation as a constrained multi-objective problem: a desirable bank must be useful for the agent, diverse in its content, and provide good coverage of the query distribution. To this end, we introduce SkillBrew, a multi-objective curation framework that formalizes skill bank curation as Pareto-aware optimization under a utility constraint, and solves it via a bi-level propose-then-verify loop. We evaluate our approach on two public benchmarks. Our findings suggest that treating skill banks as objects of principled curation, rather than ever-growing append-only logs, is an important step toward building self-improving LLM agents.", "authors": ["Wentao Hu", "Zhendong Chu", "Yiming Zhang", "Junda Wu", "Ming Jin", "Xiangyu Zhao", "Yilei Shao", "Yanfeng Wang", "Qingsong Wen"], "categories": ["cs.CL", "cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29440", "pdf_url": "https://arxiv.org/pdf/2605.29440v1", "arxiv_id": "2605.29440", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6260d2d569c598ccb827351078219b97204e837de346f7ba606d6b4c38aafe49", "sources": ["arxiv", "semantic_scholar"], "title": "AgentSchool: An LLM-Powered Multi-Agent Simulation for Education", "abstract": "Despite the rapid deployment of LLMs into classrooms, validating educational AI remains uniquely intractable: interventions act on developing learners whose cognitive and social trajectories are irreversibly shaped, while real-world trials are slow, ethically constrained, and institutionally locked. LLM-based educational simulators have emerged as a potential remedy, but many still collapse learning into persona-conditioned role-play and, when optimized only to reproduce existing classrooms, can structurally penalize the institutional novelty that pedagogical reform requires. In this work, we introduce AgentSchool, an LLM-driven multi-agent simulator that models learning as state transition rather than prompted behavior. AgentSchool couples cognitively growable student agents -- equipped with weighted subject knowledge graphs, thinking-workflow pools, and explicit misconceptions -- with adaptive teacher agents that plan, scaffold, and reflect along the Zone of Proximal Development, embedded in a configurable scenery generator that situates instruction within both formal and informal learning fields, and a multi-scale simulator that decouples interaction scale, temporal granularity, and simulation duration. Experiments show that structured student agents produce more differentiated mastery and misconception traces than a baseline simulator, while teacher-agent comparisons show backbone-dependent patterns consistent with ZPD-informed adaptation. Further, AgentSchool generates plausible traces of peripheral participation, clique formation, aggressor-induced cohesion, and opinion-leader emergence consistent with classroom social theories. Beyond its role as an educational research instrument, AgentSchool frames education as a socially meaningful testbed for long-horizon memory, multi-agent coordination, and future institutional reasoning under organizational pressure.", "authors": ["Yulei Ye", "Wenhao Li", "Zhong Wen", "Yunshu Huang", "Yichen Hu", "Zifan Wei", "Yige Wang", "Xinyu Xie", "Haoxuan Yang", "Yanjun Huang", "Ruijia Li", "Hong Qian", "Yu Song", "Bo Jiang", "Bingdong Li", "Lijun Li", "Bo Zhang", "Pinlong Cai", "Xingcheng Xu", "Shuangye Chen", "Xia Hu", "Liang He", "Aimin Zhou", "Jingjing Qu", "Jing Shao", "Xiangfeng Wang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.30144", "pdf_url": "https://arxiv.org/pdf/2605.30144v1", "arxiv_id": "2605.30144", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "25e6120dfe0689e19541713972fbd77c03487c86afc0b754d3f7b4c3ee0b60b3", "sources": ["arxiv", "semantic_scholar"], "title": "MINDGAMES: A Live Arena for Evaluating Social and Strategic Reasoning in Multi-Agent LLMs", "abstract": "Large language models (LLMs) are increasingly deployed as interactive agents, yet their capacity for social and strategic reasoning over extended interaction remains poorly understood. Existing evaluations rely on static vignettes or single-game benchmarks that cannot capture the sustained, multi-faceted reasoning that real-world multi-agent settings demand. We introduce Mindgames, a multi-game arena and evaluation platform for LLM agents that operationalizes complementary reasoning demands relevant to ``theory of mind'': belief attribution under hidden information, opponent modeling through repeated strategic interaction, cooperative inference under knowledge asymmetries, and sustained deception in social deduction. Built on TextArena, Mindgames provides a unified interaction interface, TrueSkill-based rating, and full trajectory logging across four game environments. We instantiate Mindgames through a 2025 competition cycle hosted at a major AI conference, which assessed 944 submitted agents from 76 teams across four games: Colonel Blotto, Iterated Prisoner's Dilemma, Codenames, and Secret Mafia. Our analysis surfaces both agent-level and evaluation-level limitations: brittle rule adherence remains a major bottleneck, top-performing systems repeatedly rely on explicit structural scaffolding, and leaderboard validity differs sharply across environments. In particular, failure-heavy environments can reward robustness to opponent errors as much as strategic ability, with Secret Mafia exhibiting a pronounced error-survival confound in this cycle. We release a dataset of 29,571 multi-agent games with turn-level observations, actions, and rewards, together with MG-Ref, a deterministic offline tournament protocol that scores new agents against a frozen reference pool of top-ranked, low-error Stage~II submissions under the same error-attribution lens used in this analysis.", "authors": ["Kevin Wang", "Anna Thöni", "Benjamin Kempinski", "Bobby Cheng", "Jianzhu Yao", "Benjamin Finch", "Leon Guertler", "Viraj Nadkarni", "Yihan Jiang", "Aliaksei Korshuk", "Alexander Buyantuev", "Ilya Makarov", "Siyuan Wu", "Yu-Chi Cheng", "Yan-Ru Ju", "Ti-Rong Wu", "I-Hsuan Chu", "Yu-Yu Yang", "I-Chen Wu", "Yitian Huang", "Qinlu Cao", "Yiheng Sun", "Yuhong Dai", "Hongkun Yao", "Jingxuan Fu", "Jiwei Zhang", "Hao Liao", "Mossimo Ebeling", "Govind Arun", "Sadhvik Bathini", "Mihir S Arya", "Avinash Anish", "Aditya Ranjan", "Kirtana Sunil Phatnani", "Paval KS", "Vrushali Mehta", "Aravind S", "Nikhil Arora", "Tanya Upadhyay", "Amol Bandagale", "Yuan Lu", "ChunEn Hsiao", "YuTing Lin", "Arvin Chung", "Jerry John Thomas", "Mathieu Laurière", "Leshem Choshen", "Yoram Bachrach", "Pramod Viswanath", "Maria Polukarov", "Cheston Tan", "Tal Kachman", "Atlas Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29512", "pdf_url": "https://arxiv.org/pdf/2605.29512v1", "arxiv_id": "2605.29512", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "cb4c683408ec46a64be6f16bc532595df67485a27bffd95a1311307e1972bcf2", "sources": ["arxiv", "semantic_scholar"], "title": "Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems", "abstract": "LLM-based multi-agent systems (MAS) have emerged as an effective paradigm for complex and long-horizon tasks. However, in real-world tasks, MAS often exhibit various failures during execution and such failures are difficult to eliminate during design. This motivates experience-driven MAS evolution, where a system improves based on its own execution experience. Yet such evolution is challenging because MAS experience is prolonged and intricate, interleaving multiple agents' execution chains and communication messages, which makes it difficult to identify what should be improved. To address this challenge, we propose Meta-Team, an experience-driven MAS evolution framework based on collaborative self-evolution. Meta-Team preserves the execution context of each agent and coordinates post-task communication, enabling agents to exchange distributed evidence for evolution. Building on this design, Meta-Team conducts multi-scale self-evolution, transforming execution experience into reusable improvements to agent behaviors, inter-agent coordination, and team-level organization. Across six long-horizon agent benchmarks, Meta-Team consistently outperforms single-agent systems, hand-crafted MAS, and prior MAS evolution methods; further analyses demonstrate that Meta-Team enables more reliable and scalable MAS self-evolution.", "authors": ["Zhezheng Hao", "Tianfu Wang", "Huanshuo Dong", "Ziyan Liu", "Hong Wang", "Xiankun Lin", "Qiang Lin", "Can Wang", "Hande Dong", "Jiawei Chen"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29790", "pdf_url": "https://arxiv.org/pdf/2605.29790v1", "arxiv_id": "2605.29790", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "db45cb0fa85463afa95df3aa5bce3c6ad65e7b66a530d99290d8a5b260772206", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-ALSO: LLM-Driven Adaptive Learning-Signal Optimization for Multi-Agent Reinforcement Learning", "abstract": "Effective training-time guidance is central to multi-agent reinforcement learning (MARL), yet remains difficult in sparse-reward settings where weak supervision limits coordination and policy improvement, and existing methods often require substantial domain expertise or manual design effort. Large language models (LLMs) provide a promising alternative for flexible learning-signal design, yet existing LLM-based methods remain largely single-agent-oriented, one-shot, or weakly validated for the evolving training dynamics of cooperative MARL. To address these limitations, we propose LLM-ALSO, an iterative LLM-driven adaptive learning-signal optimization framework for MARL. Rather than directly deploying LLM-generated rewards, LLM-ALSO decomposes adaptation into iterative diagnosis, proposal, and validation: a Critic LLM diagnoses stage-specific learning and coordination failures from sparse-return metrics and compact behavior evidence, a Generator LLM proposes candidate reward-shaping configurations conditioned on the diagnosis, and branch-validation feedback refines candidates before they affect the main training trajectory. Through short-horizon validation and stage-aware adaptation, LLM-ALSO promotes only validated updates into training, reducing the risk of unreliable LLM-generated modifications. Experiments on sparse-reward cooperative MARL tasks show that LLM-ALSO improves sparse-evaluation performance and learning efficiency.", "authors": ["Xiaoguang Wu", "Zhi Zheng", "Hui Xiong"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.29293", "pdf_url": "https://arxiv.org/pdf/2605.29293v1", "arxiv_id": "2605.29293", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3a40ccc0e6b9f8facf36fed31e50335bbb981e646c48133ca2f5c263a2a910c3", "sources": ["arxiv", "semantic_scholar"], "title": "A Theory-Guided LLM Pedagogical Agent for STEM+C Scaffolding Without Over-Reliance", "abstract": "LLM pedagogical agents are proliferating, yet recent findings have raised questions about their adherence to established theories of learning and, by extension, their educational value. Concerns regarding cognitive offloading, over-reliance, and \"gaming\" behaviors persist and remain largely unaddressed. In response, we developed Copa, an agentic, multi-agent, multimodal Collaborative Peer Agent for STEM+C learning. Copa is built on top of the Evidence-Decision-Feedback (EDF) framework, grounding its interactions in Social Cognitive Theory and Social Constructivism and promoting sense-making through adaptive, dialogic support rather than answer-seeking. In an authentic high school computational-modeling study (n=33 dyads), we demonstrate that Copa (1) supports students' confidence building and ability to verbalize conceptual understanding without causing dependence; and (2) provides adaptive feedback personalized to learners that is interpretable with respect to students' multimodal input data. These findings position theory-guided, multimodal LLM agents as a promising path toward classroom AI integration that amplifies students' reasoning rather than replacing it.", "authors": ["Clayton Cohn", "Surya Rayala", "Siyuan Guo", "Hanchen David Wang", "Naveeduddin Mohammed", "Umesh Timalsina", "Shruti Jain", "Ryan Li", "Angela Eeds", "Menton Deweese", "Pamela J. Osborn Popp", "Rebekah Stanton", "Shakeera Walker", "Ashwin T S", "Meiyi Ma", "Gautam Biswas"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-28", "url": "https://arxiv.org/abs/2605.30539", "pdf_url": "https://arxiv.org/pdf/2605.30539v1", "arxiv_id": "2605.30539", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5f260cd750da3734ae9c16ffc723e7a50dcead9c1b10cb84b29ec0d91e4f1b14", "sources": ["arxiv", "semantic_scholar"], "title": "MRMMIA: Membership Inference Attacks on Memory in Chat Agents", "abstract": "Membership inference attacks (MIAs) test whether a target data record belongs to a system's private data, and have become a standard tool to measure privacy leakage in machine learning systems. Prior work has primarily focused on training corpora or retrieval databases. However, MIAs against agent memory have received less attention, even though such memory can contain sensitive user-agent interactions, retrieved facts, and user preferences. Therefore, in this work, we focus on chat agent memory MIAs, where an adversary infers whether a candidate memory unit belongs to the chat agent's memory store. We propose Multi-Recall Memory MIA (MRMMIA), a unified attack that utilizes multiple recall probes to the agent to extract the membership signal across black-box, gray-box, and white-box settings. Our experiments demonstrate that MRMMIA consistently outperforms baselines. Our results expose the privacy risk in agents and provide an initial evaluation framework for membership leakage in chat-agent memory systems.", "authors": ["Kai Chen", "Yan Pang", "Tianhao Wang"], "categories": ["cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.27825", "pdf_url": "https://arxiv.org/pdf/2605.27825v1", "arxiv_id": "2605.27825", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b571632241f894b466e0d6089e77d9ae1a452b0b05f647355a807d3bce9b0047", "sources": ["arxiv", "semantic_scholar"], "title": "Do Agents Know What They Can't Do? Evaluating Feasibility Awareness in Tool-Using Agents", "abstract": "Tool-using agents often incur substantial computational cost due to long reasoning chains and iterative tool usage. In practical scenarios, many tasks become infeasible under constrained tool environments, where the capabilities required for successful task completion are unavailable. Detecting infeasible tasks and stopping execution early can significantly reduce unnecessary execution cost. In this work, we propose FeasiGen, an automatic pipeline for constructing infeasible agent tasks by identifying the critical tools required for successful task completion. Our approach extracts tool-calling traces from successful executions across multiple agent systems, identifies critical tools consistently shared across diverse execution strategies, and masks these tools to automatically transform solvable tasks into infeasible ones. Human verification confirms that the infeasibility annotations for our constructed tasks achieve over 94% accuracy. We further introduce feasibility-aware evaluation metrics for measuring whether agents can recognize infeasible tasks and stop execution appropriately. Extensive evaluations across nine models reveal substantially weak infeasibility detection ability, with false continue rate reaching up to 73.9%. We further observe that multi-agent architectures significantly reduce erroneous execution under infeasible conditions.", "authors": ["Liang Cheng", "Mingsheng Cai", "Jiuming Jiang", "Luo Mai"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28532", "pdf_url": "https://arxiv.org/pdf/2605.28532v1", "arxiv_id": "2605.28532", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3e990cb9fa5d46becb63ac9d1ebd22a1e8efb800e3b6826a95a743f7c1971199", "sources": ["arxiv", "semantic_scholar"], "title": "When Does Memory Help Multi-Trajectory Inference for Tool-Use LLM Agents?", "abstract": "Multi-trajectory inference for tool-use LLM agents - generating multiple reasoning attempts and selecting among them - benefits from transferring knowledge across attempts so that later ones avoid the pitfalls of earlier ones. Existing cross-trajectory memory methods (trajectory-level reflection, atomic fact extraction, raw observation injection) are each evaluated under a single inference strategy on a single task, making it unclear whether reported gains reflect properties of the memory abstraction or of the inference method. We propose a unified framework that decomposes memory along two axes -- the scope of transfer (within an expansion vs. across trajectories) and the abstraction of the transferred content -- and evaluate four methods under three inference strategies (best-of-N, beam search, MCTS) on four tool-use benchmarks spanning SQL, knowledge-graph, and CLI environments, in a verifier-free setting that matches the deployment regime of practical agents. The experiment matrix identifies the inference method as a confound: the same memory method produces statistically distinct results under different inference strategies on the same examples. Reflection reaches significance only under MCTS (not under best-of-N); within-expansion injection (conditioning each candidate on prior siblings' outcomes) helps only diversity-starved beam search; and atomic fact extraction is accuracy-neutral but shortens trajectories by 19-26% on tasks with reusable environmental structure.", "authors": ["Xinzhe Li", "Yaguang Tao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28224", "pdf_url": "https://arxiv.org/pdf/2605.28224v1", "arxiv_id": "2605.28224", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "298af24472dbb0fe6598ee22abb00ccb1d2753a190447d54fb40b42c383ccd99", "sources": ["arxiv", "semantic_scholar"], "title": "EgoBench: An Interactive Egocentric Multimodal Benchmark for Tool-Using Agents", "abstract": "As AI agents increasingly operate in open, real-world environments, they require a deep synergy of multimodal perception, tool invocation with multi-hop reasoning, and dynamic interaction with users. However, existing benchmarks fail to jointly evaluate these capabilities due to challenges in designing strictly coupled multi-capability tasks, simulating natural and task-constrained user feedback, and ensuring objective evaluation of dynamic interaction. To bridge this gap, we introduce EgoBench, the first interactive multimodal benchmark for tool-using agents. EgoBench comprises 1,045 egocentric-video-grounded tasks covering four daily scenarios, along with a user-agent-tool interactive environment for evaluation. We implement a three-stage synergistic pipeline through which each task is designed to enforce the joint application of visual perception and tool-augmented multi-hop reasoning. We additionally develop a multi-agent simulated user within EgoBench to evaluate agents' interaction capabilities, which generates high-fidelity, task-aligned responses to agents. Furthermore, we establish a deterministic joint validation framework that guarantees objective assessment through process-based and result-based equivalence. Benchmarking eight SOTA video-MLLM agents on EgoBench reveals a severe performance ceiling: the best model achieves only 30.62% accuracy in the best-performing scenario, averaging 19.43% across all four scenarios. Finally, we conduct a multi-dimensional error analysis to disentangle failure modes, exposing capability bottlenecks for advancing future AI agents.", "authors": ["Yunqi Liu", "Tong Niu", "Zitong Wang", "Zhenlong Dai", "Yuqi Qing", "Weiqiang Wang", "Jian Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.27820", "pdf_url": "https://arxiv.org/pdf/2605.27820v1", "arxiv_id": "2605.27820", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "87fbbf227c0400982934dcb3daceafc431a7097dec19cf5896d8d8bc65f093ca", "sources": ["arxiv", "semantic_scholar"], "title": "A Unified Framework for the Evaluation of LLM Agentic Capabilities", "abstract": "As LLMs are increasingly deployed as agents, reliable assessment of their agentic capabilities has become essential. However, reported benchmark scores often jointly reflect model capability and the implementation choices each benchmark is packaged with, making cross-benchmark results difficult to interpret as clean measurements of the underlying model. In this work, we present a unified framework for the fair evaluation of LLM agentic capabilities. Driven by a unified configuration system, the framework integrates diverse benchmarks into a standardized instruction--tool--environment format, executes agents through a fixed ReAct-style architecture within a controllable sandbox, and provides an optional offline setting that replaces volatile live environments with curated snapshots, so that framework effects and environment effects can be analyzed separately. Building on this, we unify the evaluation methodology under each benchmark's original task-success criteria, while introducing unified metrics for resource consumption and a taxonomy for decision- and execution-level failure attribution. Within this framework, we adapt 7 widely used benchmarks spanning 24 domains across single-agent, multi-agent, and safety-critical scenarios, and conduct a large-scale empirical analysis over 400K rollouts and 5B tokens on 15 models. The results show that scaffold choice and environmental volatility materially shift benchmark outcomes in both directions, allowing our framework to disentangle intrinsic LLM capabilities from framework- and environment-induced artifacts. We further demonstrate its extensibility as a secure testbed for safety-critical domains. Codes and benchmarks at are available at https://github.com/whfeLingYu/A-Unified-Framework-for-the-Evaluation-of-LLM-Agentic-Capabilities, https://huggingface.co/AgentFramework/Unified_Farmework.", "authors": ["Pengyu Zhu", "Lijun Li", "Yaxing Lyu", "Qianxin Luo", "Jingyi Yang", "Yi Liu", "Tingfeng Hui", "Xinyu Yuan", "Li Sun", "Sen Su", "Jing Shao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.27898", "pdf_url": "https://arxiv.org/pdf/2605.27898v1", "arxiv_id": "2605.27898", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/whfeLingYu/A-Unified-Framework-for-the-Evaluation-of-LLM-Agentic-Capabilities", "venue": null, "quality_score": 0.65}
{"id": "636a5875908c40b43dcc168f16de097b2061f80c516abf87589ce92f0a427cef", "sources": ["arxiv", "semantic_scholar"], "title": "Tool Forge: A Validation-Carrying Toolchain for Governed Agentic Execution", "abstract": "Large language model agents are increasingly expected to perform operational work: calling APIs, manipulating files, assembling workflows, and acting inside enterprise systems. Yet the tool layer on which this execution depends is still commonly treated as either a hand-written integration artifact or a static list of schemas exposed to a model. This paper introduces Tool Forge, a validation-carrying toolchain for converting natural-language capability intent into governed, sandbox-verified, cataloged tool artifacts and exposing those artifacts to agents through a token-efficient routing layer. Tool Forge treats a tool as a capsule containing intent, capability contract, implementation, dependency policy, tests, documentation, runtime validation evidence, lifecycle state, credential bindings, and routing metadata. It also introduces a Router that exposes intent-scoped tool sessions instead of loading full catalog schemas into the model context. We describe the system architecture, validation pipeline, MCP-facing routing model, governance controls, and initial reproducible benchmarks from the open-source implementation. Across 83 Router benchmark cases, Tool Forge Router achieves aggregate micro-F1 of 0.901 while reducing estimated task-flow tool context by 99.2% relative to naive full-catalog schema exposure. In a 25-case end-to-end generation probe over local-tool tasks, Tool Forge generates 25 of 25 tool bundles, reaches micro-F1 of 0.940 against deterministic acceptance checks, and passes 23 of 25 live sandbox validations. These results are presented as an initial systems benchmark, not as a state-of-the-art claim. The paper identifies remaining challenges in adversarial routing, broader API grounding, sandbox isolation, and cross-system evaluation.", "authors": ["Swanand Rao"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28000", "pdf_url": "https://arxiv.org/pdf/2605.28000v1", "arxiv_id": "2605.28000", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/nextmoca/tool-forge", "venue": null, "quality_score": 0.65}
{"id": "f167587e9632604783758250afc8647bf748581dbd6a4b0835d38dcc4f9cf363", "sources": ["arxiv", "semantic_scholar"], "title": "MolLingo: Molecule-Native Representations for LLM-Powered Scientific Agents", "abstract": "We present MolLingo, a multi-agent system that emulates the reasoning process of a chemist to automate molecular design. Existing LLM-based approaches either operate as standalone generative models without access to external tools or lack the multi-agent coordination and shared memory needed for iterative, evidence-driven reasoning across the molecular design pipeline. MolLingo addresses this by coordinating a Literature Agent, a Chemist Agent, and an Orchestrator through a shared memory module, with each agent equipped with domain-specific tools. To enable effective molecular reasoning, we introduce BRICS-based Fragment Enumeration (BFE), a synthesis-aware molecular fragmentation method that decomposes molecules into chemically meaningful building blocks represented as block-based SMILES paired with common chemical names. This representation bridges molecular structure and LLM semantic space, enabling block-level reasoning and editing that is difficult with raw SMILES alone. As a case study in early-stage therapeutic design, MolLingo further grounds the Chemist Agent's reasoning in binding site geometry and residue-level protein context derived from molecular docking to optimize molecules for stronger target binding. Across four benchmarks, MolLingo consistently outperforms frontier LLMs and specialized baselines, including a fourfold docking score improvement over GPT-5.4 despite using the same underlying model, consistent drug property optimization gains across multiple LLM backbones, and state-of-the-art results on TOMG-Bench, surpassing both frontier LLMs and the RL-based optimization method RePO. Our results suggest that LLMs are already capable molecular design assistants when guided through chemically meaningful representations and biologically grounded structural context. Code is available at: https://anonymous.4open.science/status/MolLingo-7450.", "authors": ["Thao Nguyen", "Heng Ji"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.27853", "pdf_url": "https://arxiv.org/pdf/2605.27853v1", "arxiv_id": "2605.27853", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "3b6d4f6906421ba85a20786b5742c4096ef3c6199d4dd9742d6becc153e2774d", "sources": ["arxiv", "semantic_scholar"], "title": "Mixture-of-Experts Knowledge Graph Retrieval-Augmented Generation for Multi-Agent LLM-based Recommendation", "abstract": "Large language models (LLMs) have recently been adopted for recommendations due to their ability to understand user intent and item semantics. However, LLM-based recommender systems often rely on parametric knowledge and suffer from outdated knowledge, motivating knowledge graph retrieval-augmented generation (KG-RAG) to ground recommendations on structured, up-to-date KGs. Despite this promise, effective KG-RAG in recommendations faces great challenges. First, users' queries vary in complexity and require KG knowledge at different granularities, whereas existing methods adopt a one-size-fits-all retrieval strategy, leading to over-retrieval for simple queries and under-retrieval for complex ones. In addition, augmenting LLMs with KG knowledge requires translating graph-structured data into linear text, which may introduce noise and cause structural information loss. Moreover, the selection of retrieval granularity lacks direct supervision and must be inferred from the final recommendation after alignment and downstream utilization, making query-aware retrieval hard to learn end-to-end. To address these issues, we propose MixRAGRec, a cooperative multi-agent framework for KG-RAG recommendations. MixRAGRec integrates a Mixture-of-Experts Retrieval Agent that routes each query to a KG retrieval expert with different granularities, a Knowledge Preference Alignment Agent that converts structured knowledge into LLM-friendly natural language, and a Contrastive Learning-reinforced Recommendation Agent trained with contrastive preference feedback. Notably, we introduce Mixture-of-Experts Multi-Agent Policy Optimization (MMAPO) to train three agents under a unified objective. Extensive experiments on real-world datasets demonstrate the effectiveness of our framework.", "authors": ["Shijie Wang", "Chengyi Liu", "Yujuan Ding", "Shanru Lin", "See-Kiong Ng", "Xu Xin", "Wenqi Fan"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28175", "pdf_url": "https://arxiv.org/pdf/2605.28175v2", "arxiv_id": "2605.28175", "doi": "10.1145/3770855.3817630", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "efced1838dcbca799f62038b63df0f0176a1d35e280bd3862d9599af81a98441", "sources": ["arxiv", "semantic_scholar"], "title": "Examining Agents' Bias Amplification versus Suppression in Multi-Agent Systems", "abstract": "Multi-agent systems are increasingly deployed to support various tasks where agents interact to achieve individual and collective objectives. Although these systems can enhance task performance and decision-making, fairness preservation through bias reduction remains challenging. This study examines how agent-level biases shift and impact system-wide fairness. We use prompts to expose individual agents to group-favoring bias, then assess downstream impacts at the system level. To quantify the impact, we propose Favor Bias Strength (FBS), a zero-centered metric that decomposes bias alteration between favored-group uplift and disfavored-group suppression. Using multiple agent designs, benchmarks, and up-to-date large language models, we show that agents endowed with bias can substantially affect system-wide fairness. Interestingly, when agents are exposed to bias uniformly, the system-wide bias elevates, even exceeding the additive sum of the individual agents' biases. The empirical evidence underscores the criticality of fairness in multi-agent systems, which warrants further analyses and empirical tests.", "authors": ["Zejian Eric Wu", "Zhongyi Jiang", "Yuan Zhuang", "Paul Jen-Hwa Hu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28098", "pdf_url": "https://arxiv.org/pdf/2605.28098v1", "arxiv_id": "2605.28098", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "04e912a4a1454289103316cb797db2a6221471613207c2397c53a83bdbae5ce6", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLM-based Metamorphic Testing for REST APIs", "abstract": "As REST APIs become an increasingly significant part of software systems, their validation is becoming more critical. Hence, testing and uncovering underlying issues are of utmost importance for improving software quality. However, testing REST APIs is challenging mainly due to the difficulty of assessing whether the output of an API call is correct, i.e., the test oracle problem. Metamorphic testing is a specification-based testing approach for situations where correct outputs are unknown or not specified explicitly. To check the correctness of a system, relations between the different outputs are specified. We present ARMeta, a tool-supported approach that uses an LLM-based multi-agent workflow to support metamorphic testing of REST APIs documented with OpenAPI. The agentic workflow is used to identify metamorphic test scenarios and specify them in the Given-When-Then format. These scenarios are automatically implemented as executable tests and executed against the system under test. We evaluate ARMeta on two publicly available web applications that expose REST interfaces and compare its performance with a scenario-based testing baseline. The results show that ARMeta explores behaviors that serve as a complement to existing scenario-based testing approaches.", "authors": ["Shehroz Khan", "Abdullah Mughees", "Gaadha Sudheerbabu", "Tanwir Ahmad", "Dragos Truscan"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.28321", "pdf_url": "https://arxiv.org/pdf/2605.28321v1", "arxiv_id": "2605.28321", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6f8f94bfe504b805588edd99590da35f79a7648c0bd603ad55134114eb80f932", "sources": ["arxiv", "semantic_scholar"], "title": "SafeRx-Agent: A Knowledge-Grounded Multi-Agent Framework for Safe and Explainable Medication Recommendation", "abstract": "Medication recommendation predicts medications for patient visits, but existing methods still face two key challenges. At the model level, traditional drug recommendation methods only predict structured drug codes with limited evidence grounding, while LLM agents can use richer clinical context but may lack safety verification and traceability. At the task level, existing benchmarks often use broad medication categories, which ignore subgroup-level safety differences and can lead to risk overestimation. We introduce the first fine-grained medication recommendation setting based on fourth-level ATC code generation. We propose Safe Prescription Agent (SafeRx-Agent), a knowledge-grounded multi-agent framework that uses patient context, external clinical knowledge, and safety verification to recommend traceable medication sets. Experimental results on MIMIC-III and MIMIC-IV datasets show that SafeRx-Agent improves fine-grained medication prediction accuracy while controlling drug interactions, contraindications, and medication set size.", "authors": ["Xinyu Wang", "Hanwei Wu", "Zhenghan Tai", "Sicheng Lyu", "Qincheng Lu", "Ziyu Zhao", "Jijun Chi", "Jingrui Tian", "Xiao-Wen Chang", "Ziyang Song"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-27", "url": "https://arxiv.org/abs/2605.29146", "pdf_url": "https://arxiv.org/pdf/2605.29146v2", "arxiv_id": "2605.29146", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "fede12f8835ea170c097626204cf8028ee4cca81e490000c0238c1f9322d2578", "sources": ["arxiv", "semantic_scholar"], "title": "Voluntary Collusion with Secret Tools in Competing LLM Agents", "abstract": "Even when a tool is explicitly described as unfair and harmful to others, ostensibly safety-aligned LLM agents still voluntarily engage in secret collusion whenever doing so confers a strategic advantage. To investigate this phenomenon, we introduce an empirical framework built on two strategic multi-agent environments: Liar's Bar, a competitive deception scenario, and Cleanup, a mixed-motive resource-management scenario, in which agents are offered secret collusion tools that provide significant advantages while clearly disadvantaging the other agents. Across 12 models (at the 7B, 70B, and proprietary scales) and 6 prompt variants, we find that most agents consistently accept these tools and develop collusive strategies, while explicitly acknowledging the unfairness of the tools before accepting. We further show that neither the unfairness labels nor baseline alignment alone reliably deters collusion: only explicit ethical framing reduces adoption and, even then, smaller models remain susceptible. More broadly, our work presents the first systematic investigation of voluntary collusion adoption in LLM-based multi-agent systems, and suggests that preventing such behaviour requires explicit safeguards rather than reliance on general alignment.", "authors": ["Xijie Zeng", "Frank Rudzicz"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27593", "pdf_url": "https://arxiv.org/pdf/2605.27593v1", "arxiv_id": "2605.27593", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0d203e77f5144e6b09082f114c73a939208e365fdb42e0029ad146815f6976d6", "sources": ["arxiv", "semantic_scholar"], "title": "Agents that Matter: Optimizing Multi-Agent LLMs via Removal-Based Attribution", "abstract": "As multi-agent systems (MAS) become increasingly complex, identifying the contributions of individual agents is critical for system optimization. However, existing approaches lack a rigorous, unified framework for credit assignment. In this work, we formalize agent attribution as a cooperative game, parameterized by the coalition distribution, removal protocol, and target metric. Using this framework, we show that Leave-One-Out (LOO) identifies bottleneck agents as effectively as combinatorial methods, but at a fraction of the computational cost. We also demonstrate that removal protocols induce distinct games: Agent ablation isolates structural bottlenecks, whereas introspective LLM judges fail to faithfully approximate this behavior. Furthermore, to evaluate the utility of specific agent backbones, we introduce attribution via model replacement. By substituting underlying models of low-contribution agents, we improve task performance by up to 17% while reducing cost by up to 35% across three benchmarks. Finally, we apply our framework to audit a medical MAS, revealing that agent contributions to diagnostic accuracy and ethical behavior are often decoupled. By intervening on counterproductive roles, we observe an increase in ethics alignment while maintaining diagnostic accuracy. Overall, this work provides a principled approach for cost-effective MAS attribution and intervention.", "authors": ["Mingyu Lu", "Yushan Huang", "Chris Lin", "Su-In Lee"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27621", "pdf_url": "https://arxiv.org/pdf/2605.27621v1", "arxiv_id": "2605.27621", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "620fbe3ccab184dfebd8b7248207d8e447810dee1bbad7a5984359368f8da6a8", "sources": ["arxiv", "semantic_scholar"], "title": "HARP: Measuring Harm Amplification in Multi-Agent LLM Systems", "abstract": "Multi-agent LLM systems decompose workflows across agents, tools, shared context, memory, and decision gates. This modularity improves interpretability, but creates a propagation risk: a bounded perturbation to one component can be reused by other agents and amplified into system-level harm. We introduce HARP (Harm Amplification through Role Perturbation), a trace-first methodology for studying local-to-global harm amplification in multi-agent LLM systems. HARP compares paired clean and perturbed executions and records specialist outputs, tool calls, memory reads/writes, guard events, oracle logs, latency, token cost, and decisions. We define local harm as deviation from targeted agents or corrupted channels, global harm as deviation over the full trace, and harm amplification as (H_global/H_local). This complements attack success rate with a measure of how strongly orchestration spreads harm beyond the attack point. We instantiate HARP in a finance-oriented seven-agent system with a deterministic decision gate and configurable attack harness for specialist compromise, collusion, shared-context corruption, and temporal or memory-persistent attacks. Across five defenses, prompt-only defenses preserve benign utility but leave high success and stealth; pre-tool and step-level guards reduce some failures with utility or latency costs; and IntegrityGuard, a trace-consistency defense, achieves the lowest attack success and global harm but introduces utility/cost trade-offs. Results show that single-specialist compromise produces the strongest amplification, shared-context corruption yields the highest attack success, and temporal persistence produces the largest malicious impact. HARP argues that secure multi-agent evaluation must measure not only bypass, but propagation.", "authors": ["Md Hafizur Rahman", "Zafaryab Haider", "Tanzim Mahfuz", "Prabuddha Chakraborty"], "categories": ["cs.CR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27489", "pdf_url": "https://arxiv.org/pdf/2605.27489v1", "arxiv_id": "2605.27489", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "53421d40fd629663042ce04bd89500470a939e5bc52350ef86f09993fa899067", "sources": ["arxiv", "semantic_scholar"], "title": "Got a Secret? LLM Agents Can't Keep It: Evaluating Privacy in Multi-Agent Systems", "abstract": "LLM safety evaluations predominantly test models in isolation, yet deployed AI agents increasingly operate within persistent social environments alongside other agents. We introduce a Moltbook-style simulation platform where thousands of LLM agents interact across communities over a simulated month, and use it to evaluate privacy as a downstream safety concern under varying degrees of social pressure. We find that shifting from single turn to multi turn social evaluation amplifies privacy violations (CIMemories 19.95% to Ours 45.30% across OpenAI models), that leakage is socially contagious, with agents 8 times more likely to disclose sensitive information after observing a peer do so, and that explicit privacy instructions reduce but do not eliminate this effect, leaving leakage rates above 37.8% even with safeguards. Our findings suggest that static chat based safety benchmarks systematically underestimate risks in agentic deployment, and that social context alone is sufficient to elicit sensitive disclosures that single turn evaluations would never surface.", "authors": ["Aman Priyanshu", "Supriti Vijay", "Esha Pahwa"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27766", "pdf_url": "https://arxiv.org/pdf/2605.27766v1", "arxiv_id": "2605.27766", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "ecb6d140e8b7f9bbf5d830b1469fb0badac665c5ef2e425fa9d3d0dee437ad6e", "sources": ["arxiv", "semantic_scholar"], "title": "UnityMAS-O: A General RL Optimization Framework for LLM-Based Multi-Agent Systems", "abstract": "LLM-based multi-agent systems decompose complex tasks into interacting roles, but most remain manually orchestrated by prompts, tools, and control rules, while agents are rarely optimized through a unified reinforcement learning interface. Existing RL post-training frameworks mainly target single-policy optimization and lack abstractions for user-defined multi-agent workflows, structured interaction, role-specific credit assignment, and configurable parameter sharing. We present UnityMAS-O, a general RL optimization framework for LLM-based multi-agent systems. UnityMAS-O treats the complete workflow as the optimization unit, rather than a single response or policy trajectory. It represents workflows through four first-class objects: logical agent roles, graph trajectories, user-defined rewards, and agent--model mappings. This decouples logical agents from physical model parameters, supporting full sharing, full separation, and partial sharing, with rewards assigned at role, turn, and trajectory levels. UnityMAS-O extends verl with a Ray-based star-topology runtime. A central controller executes workflows, invokes tools, records structured trajectories, and assembles rewards; model-local worker groups handle rollout, buffering, advantage computation, and distributed PPO-style updates. Users can define agents, workflows, model mappings, and rewards without rewriting the optimization infrastructure. We instantiate UnityMAS-O on retrieval-augmented QA, iterative agentic search, and reflective code generation. Across Natural Questions, HotpotQA, and held-out code tasks, multi-agent RL improves manually specified workflows after optimization, with especially large gains for smaller models and strict code all-passed metrics. These results show that UnityMAS-O can serve as a reusable substrate for converting diverse LLM-based multi-agent workflows into trainable multi-agent RL systems.", "authors": ["Yiqun Chen", "Wei Yang", "Erhan Zhang", "Shijie Wang", "Qi Liu", "Zechun Niu", "Bin Zhang", "Haitao Li", "Rui Li", "Lingyong Yan", "Jinyuan Feng", "Biqing Qi", "Xiaochi Wei", "Yan Gao", "Yi Wu", "Yao Hu", "Jiaxin Mao"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.26646", "pdf_url": "https://arxiv.org/pdf/2605.26646v1", "arxiv_id": "2605.26646", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "dea4308531e83625260e5845be78765ffa9e4b722b0617582a38b7c87bc19a39", "sources": ["arxiv", "semantic_scholar"], "title": "TADDLE: A Tool-Augmented Agent for Detecting Deficient LLM-Generated Peer Reviews", "abstract": "LLM-generated peer reviews are increasingly common at major venues, yet their deficiencies are hard to detect because they are uniformly fluent and well-structured. Existing work either classifies authorship without judging quality, or scores quality with features designed for human-written reviews; no prior system detects deficiencies in LLM-generated reviews at the level of individual defect types. To bridge the gap, we introduce TADDLE, a Tool-Augmented Agent for Detecting Deficient LLM-Generated Peer Reviews, together with the first expert-annotated benchmark for this task. Our benchmark comprises 1,800 reviews on 50 ICLR 2025 papers, multi-label-annotated by 18 domain experts against a taxonomy of six defect categories (plus a non-deficient label). TADDLE decomposes detection into four specialized analysis tools -- Verify, Correct, Complete, and Transform -- orchestrated by an agent; an integrator synthesizes their outputs into binary and multi-label classifications via two-stage semi-supervised learning. Extensive experiments show that TADDLE performs strongly on both binary detection and the multi-label classification task. We release the benchmark and code at https://github.com/AquariusAQ/TADDLE.", "authors": ["Hanqi Duan", "Xiang Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.26911", "pdf_url": "https://arxiv.org/pdf/2605.26911v1", "arxiv_id": "2605.26911", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/AquariusAQ/TADDLE", "venue": null, "quality_score": 0.65}
{"id": "2a3ff62c9f2e7fc34a7272db86f0a5a971de85a29b07cc8bad4c812e3012ceae", "sources": ["arxiv", "semantic_scholar"], "title": "Decoupled Intelligence: A Multi-Agent LLM Framework for Controllable Traffic Scenario Generation in SUMO", "abstract": "The integration of Large Language Models (LLMs) with microscopic traffic simulation offers a promising path toward autonomous urban planning and intelligent transportation analysis. However, existing monolithic agent architectures often struggle with the complexity of end-to-end simulation workflows, leading to reasoning failures, parameter inconsistency, and a lack of systematic state management. This paper proposes a novel multi-agent collaborative framework designed to automate the entire lifecycle of traffic simulation in SUMO (Simulation of Urban Mobility). Our approach decouples the simulation pipeline into specialized roles, including Planner, Builder, Demand, Runner, and Analyst, coordinated by a high-level reasoning engine. We introduce a state-persistent Orchestrator leveraging the Model Context Protocol (MCP) to ensure seamless data handover and environmental consistency across distributed agent actions. This architecture enables a robust closed-loop refinement process, where simulation outcomes are iteratively analyzed and optimized to satisfy user-defined Key Performance Indicators (KPIs). Experimental results through role ablation studies demonstrate that the proposed multi-agent framework significantly enhances task success rates and parameter accuracy compared to single-agent baselines. Furthermore, case studies on real-world network extraction and traffic optimization highlight the system's capability to bridge the gap between high-level natural language intent and low-level simulation execution.", "authors": ["Shuyang Li", "Ruimin Ke"], "categories": ["cs.MA", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27685", "pdf_url": "https://arxiv.org/pdf/2605.27685v1", "arxiv_id": "2605.27685", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2c7ad6f14d755a8643947ad70d3a8d5b654ac7b7d6b02bcbe6785b5e11107049", "sources": ["arxiv", "semantic_scholar"], "title": "FinHarness: An Inline Lifecycle Safety Harness for Finance LLM Agents", "abstract": "Finance LLM agents must simultaneously block prompt-induced unauthorized actions and approve legitimate multi-step business workflows. However, boundary filters often miss irreversible mid-trajectory tool calls, while post-hoc LLM judges perform auditing only after termination -- too late for intervention and at a computational cost that scales linearly with trace length. We present FinHarness, an inline safety harness that wraps a finance agent end-to-end with three components: a Query Monitor that fuses single-turn intent with cross-turn drift, a Tool Monitor that evaluates each prospective tool call, and a Cascade module that integrates per-step risk and adaptively routes verification between a lightweight and an advanced-tier LLM judge. Fired risk factors are re-injected into the agent input as ex-ante evidence, enabling the agent to refuse, re-plan, or approve on its own. On FinVault, routed FinHarness cuts ASR from 38.3% to 15.0% while largely preserving benign approval ($41.1\\% \\to 39.3\\%$), and uses $4.7\\times$ fewer advanced-judge calls than an always-advanced ablation.", "authors": ["Haoxuan Jia", "Yang Liu", "Bin Chong", "Yingguang Yang", "Yancheng Chen", "Jiayu Liang", "Qian Li", "Hanning Lu", "Kefu Xu", "Hao Zheng", "Chongyang Zhang", "Hao Peng", "Philip S. Yu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27333", "pdf_url": "https://arxiv.org/pdf/2605.27333v1", "arxiv_id": "2605.27333", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2d654f81bf2e8c5c69a936a1961a9776cc8db723783f4e753a29cedfa12deb11", "sources": ["arxiv", "semantic_scholar"], "title": "TRACES: Proactive Safety Auditing for Multi-Turn LLM Agents via Trajectory-State Modeling", "abstract": "LLM agents increasingly operate through multi-turn tool use and environment interaction, where safety risks often emerge from intermediate steps long before they surface in the final outcome. Reactive auditing is therefore insufficient: post-hoc diagnosis frequently misses the chance to flag risks while they are unfolding. We propose TRACES, a representation-based proactive auditor that learns prefix-level trajectory risk states from the hidden representations of an observer LLM. TRACES induces latent mechanism features from step representations and models their temporal evolution to estimate whether a partial trajectory is drifting toward unsafe behavior. To sidestep the cost and ambiguity of step-level risk annotation, TRACES is trained with weak trajectory-level supervision while still producing dense prefix-level risk estimates. Across multiple agent safety benchmarks, TRACES improves both full-trajectory safety prediction and proactive risk discrimination. Our analyses further suggest that these risk states can help train a safer agent, highlighting the broader potential of proactive auditing for long-horizon agent safety.", "authors": ["Jiaqian Li", "Yanshu Li", "Boxuan Zhang", "Ruixiang Tang", "Kuan-Hao Huang"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27690", "pdf_url": "https://arxiv.org/pdf/2605.27690v1", "arxiv_id": "2605.27690", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0d45dfe7d74798c43d6b16da136b221aa25b085e1e3461071a9473f3f3062cef", "sources": ["arxiv", "semantic_scholar"], "title": "A Policy-Driven Runtime Layer for Agentic LLM Serving", "abstract": "Multi-agent LLM systems have become the dominant production workload, but the serving stack was not built for them. The agent framework above knows agent identities, role, schemas, and dispatch structure but never sees an engine-level event; the serving engine below sees every event but knows nothing about agents. A surprising number of cross-cutting policies depend on both: prefix caching, batch shaping, speculative execution, fairness, tool-result memoization, safety enforcement, and more. Each lives in the seam between the two layers and is currently solved by a one-off patch into one neighbor or the other. We argue this seam is best addressed by an architectural change rather than point fixes: insert a third tier, an agent runtime layer, between the framework and the engine, exposing four primitives (observe, score, predict, act) into which any agent-aware policy plugs, with agent identity as the shared coordinate. We map nine concrete policies onto the layer and validate the abstraction in depth on the one with the largest immediate serving-cost lever: KV caching across sessions, instantiated as CacheSage, which learns the per-workload agent transition matrix online and uses it for survival-based eviction and between-step prefetch. Preliminary results on five real multi-agent workloads show +13 to +37 pp cache hit-rate lift, 12% to 29% lower mean TTFT, and 6% to 14% higher throughput over an unmodified serving stack.", "authors": ["Rui Zhang", "Chaeeun Kim", "Liting Hu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.27744", "pdf_url": "https://arxiv.org/pdf/2605.27744v1", "arxiv_id": "2605.27744", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7096f71493f2f15e0247dcba31891b8d20dee6aaa2ac549d1f7517d00f1d5e62", "sources": ["arxiv", "semantic_scholar"], "title": "Helicase: Uncertainty-Guided Supply Chain Knowledge Graph Construction with Autonomous Multi-Agent LLMs", "abstract": "LLM-based multi-agent systems have been widely adopted for knowledge retrieval and report generation, synthesizing known information through web search and textual reasoning. However, many critical information tasks in supply chains are not simple one-shot queries: they are structural inference problems requiring multi-hop reasoning across complex, fragmented web resources. Questions such as \\textit{``Which Tesla components use lithium from Australian mines?''} have no answer in any single document; answers must be computationally synthesized through the autonomous construction and analysis of dynamic knowledge graphs assembled from fragmented, heterogeneous sources. Moreover, such discovery processes must be uncertainty-aware: decisions depend not only on answers but on calibrated confidence in their reliability, traceable to source quality and reasoning consistency. To address this capability gap, we propose \\textit{Helicase}, an autonomous multi-agent LLM system for uncertainty-guided supply chain knowledge graph construction. \\textit{Helicase} decomposes high-level supply-chain queries into executable investigation plans, coordinates specialized web-search, reasoning, and coding agents through iterative verification loops, and incrementally constructs query-specific supply chain knowledge graphs with per-fact uncertainty annotations. Its three-layer uncertainty framework tracks uncertainty at the action, trajectory, and memory layers, enabling both structural inference and calibrated confidence assessment. To evaluate autonomous reasoning across the full complexity spectrum, we introduce SCQA (Supply Chain Query Assessment), a benchmark of 80 supply chain queries organized into four quadrants spanning single-hop to multi-hop inference under both high and low data visibility.", "authors": ["Yunbo Long", "Haolang Zhao", "Ge Zheng", "Alexandra Brintrup"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-26", "url": "https://arxiv.org/abs/2605.26835", "pdf_url": "https://arxiv.org/pdf/2605.26835v1", "arxiv_id": "2605.26835", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7dd6a05ef0aa25e73f9ff50d484c2dbc9fee6a8da8e026445ef5f96408059b0c", "sources": ["arxiv", "semantic_scholar"], "title": "Stateful Inference for Low-Latency Multi-Agent Tool Calling", "abstract": "Multi-agent tool calling is becoming the dominant interaction pattern for LLM-based systems, yet existing inference frameworks treat each tool call as an independent request, re-processing the entire conversation from scratch even though 85-95% of the prompt is unchanged from the previous turn. We present a stateful inference architecture that converts the $O(n_t)$ per-turn cost of conventional serving into an $O(Δ_t)$ delta-only cost: a persistent KV cache lives across turns and advances by ingesting only the new tokens, while a radix prefix cache extends this across interleaved multi-agent traffic and a prompt-lookup speculative decoder accelerates structured output. Against vLLM and SGLang on novel, fully-generated workloads, the reference implementation is $2.1\\times$ faster per turn on a 6-turn agentic workflow and $4.2\\times$ on the median turn of a 35-turn one, halving end-to-end wall time. The advantage comes from stateful reuse and speculation, not caching.", "authors": ["Victor Norgren"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-25", "url": "https://arxiv.org/abs/2605.26289", "pdf_url": "https://arxiv.org/pdf/2605.26289v1", "arxiv_id": "2605.26289", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d01e686ac0d679740def537aa7e2fda81620f9155ba80fa4a0d331c0fb9fff13", "sources": ["arxiv", "semantic_scholar"], "title": "Tool-Call Dependency Structure is Linearly Decodable in LLM Agent Residual Streams", "abstract": "Tool-using LLM agents produce trajectories whose calls form a directed dependency graph: earlier tool outputs supply arguments to later calls. Whether this execution structure is represented inside the model is unknown; prior structural probes have targeted static code or chain-of-thought text, not an agent's run-time call graph. A low-capacity edge probe on the residual stream of Qwen3-32B decodes the tool-call dependency graph well above both a Hewitt--Liang random-label control and a positional baseline. A counterfactual contrast between value corruption and structural perturbation indicates the signal tracks abstract topology rather than identifier values, and replicates under an independent, non-substring oracle. The non-positional component replicates on three further interactive multi-hop benchmarks and attenuates as call order alone becomes a sufficient proxy for dependency, vanishing in single-shot planning. Per-layer activation patching shifts the probe at a later, non-patched boundary, evidence that the representation propagates rather than passively reads out, though the realised tool call does not move. To our knowledge this is the first structural probe of an LLM agent's runtime tool-call dependency graph. Our claims concern representation, not behavioural control, and span two model families and one primary domain.", "authors": ["Tianda Sun", "Dimitar Kazakov"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-25", "url": "https://arxiv.org/abs/2605.25310", "pdf_url": "https://arxiv.org/pdf/2605.25310v1", "arxiv_id": "2605.25310", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "27fc16535a85de67fd41f03476c888ac520b2b5ed071f1db8ff78607ec761cd1", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Systems are Mixtures of Experts: Who Becomes an Influencer?", "abstract": "The effectiveness of multi-agent LLM deliberation depends not only on the agents' individual predictions, but also on how they communicate and collaborate. We study this mechanism through the lens of Friedkin-Johnsen (FJ) opinion dynamics, a tractable model for analyzing stubbornness, influence, and opinion change in multi-agent systems that captures empirically observed deliberation patterns. We show that the FJ parameters are input-dependent, turning multi-agent deliberation into a mixture of experts. This perspective implies that multi-agent systems can outperform single agents and static ensembles when routing reflects agent competence. Since competence is latent in practice, we analyze how influence is established through observable proxies: agents' self-assessed confidence, their perceived confidence, and initial alignment with other agents' views.", "authors": ["Franka Bause", "Jonas Niederle", "Martin Pawelczyk", "Rebekka Burkholz"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-25", "url": "https://arxiv.org/abs/2605.25929", "pdf_url": "https://arxiv.org/pdf/2605.25929v1", "arxiv_id": "2605.25929", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5572df65022f8c3c826324972a5e2a769bc24d5a866917f2ec6a00c4c2784fe7", "sources": ["arxiv", "semantic_scholar"], "title": "DarkForest: Less Talk, Higher Accuracy for Multi-Agent LLMs", "abstract": "Multi-agent LLM systems improve reasoning by combining outputs from multiple agents, but interaction-heavy methods can introduce error propagation and high communication overhead. When agents exchange raw responses or reasoning traces, incorrect intermediate reasoning may be adopted and amplified, leading to confident but wrong consensus; multi-round communication also increases token consumption, latency, and inference cost. In this paper, we propose a controlled-communication coordination framework named DarkForest. DarkForest first keeps agents independent, so each agent produces an answer without seeing the others' outputs. It then parses the raw responses into structured candidate records, groups semantically equivalent candidates into clusters, and estimates a calibrated belief distribution over these clusters using agent reliability, confidence, parse quality, support-pattern reliability, and independence corrections. A coordinator receives only policy-permitted evidence from this belief state with controlled communication. Experiments on six reasoning benchmarks show that DarkForest achieves leading overall quality, improves the strongest baseline by up to 30.7\\% on benchmark metrics, and reduces token consumption by up to $6.5\\times$ compared with communication-heavy baselines.", "authors": ["Yi Li", "Songtao Wei", "Dongming Jiang", "Zhichun Guo", "Qiannan Li", "Bingzhe Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-24", "url": "https://arxiv.org/abs/2605.25188", "pdf_url": "https://arxiv.org/pdf/2605.25188v1", "arxiv_id": "2605.25188", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e0d07677e75375282ab6a07559010d907f4238b10a8fc8a775506179a9f0c9e9", "sources": ["arxiv", "semantic_scholar"], "title": "Meta-Agent: From Task Descriptions to Verified Multi-Agent Systems", "abstract": "AI agents are increasingly used to solve complex, multi-step tasks, but existing multi-agent frameworks remain brittle as workflows grow in scale and depth. Small errors at intermediate stages can propagate through agent interactions, while insufficient grounding and weak verification mechanisms further limit reliability. We present Meta-Agent, a two-phase framework that automatically constructs and executes specialized multi-agent systems from natural-language task descriptions. In the construction phase, a task planner decomposes a problem into a directed acyclic graph of agent specifications with explicit input/output contracts and verification criteria. A web search module grounds each specification with external evidence, and a code generation module produces system prompts and tool configurations. A construction-time verification stage then validates generated artifacts and triggers targeted regeneration when failures are detected. In the execution phase, a coordinator dispatches subtasks across the agent graph while execution-time verification gates intermediate outputs. We further introduce a three-level error attribution mechanism that distinguishes local, upstream, and structural failures, enabling targeted recovery strategies ranging from localized retries to partial re-execution and re-decomposition. We evaluate Meta-Agent across coding, contextual learning, and open-ended reasoning tasks. Experiments against strong multi-agent baselines and ablation studies demonstrate consistent improvements in task success rate, error recovery, and workflow stability. The results highlight the importance of tightly integrating planning, grounding, and verification for building reliable multi-agent systems.", "authors": ["Andy Xu", "Yu-Wing Tai"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-24", "url": "https://arxiv.org/abs/2605.25233", "pdf_url": "https://arxiv.org/pdf/2605.25233v1", "arxiv_id": "2605.25233", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "49e1b39966d635dfb24243186f944331b601c28113aa494cc84a046df91c9732", "sources": ["arxiv", "semantic_scholar"], "title": "GroupTravelBench: Benchmarking LLM Agents on Multi-Person Travel Planning", "abstract": "Travel planning in the real world is overwhelmingly a \\textit{group} activity, yet existing LLM travel-planning benchmarks reduce it to a single user, where the field is approaching saturation. This single-user assumption sidesteps what makes group planning hard for an agent: discovering private preferences across multiple users, surfacing conflicts, and balancing utility against fairness. To bring the task back to its multi-user reality, we introduce \\textbf{\\textit{GroupTravelBench}}, the first benchmark for \\textbf{multi-user, multi-turn} travel planning. Built from real user profiles, POI data, and ticket prices, it comprises 650 tasks across three difficulty levels, each running in a synchronous group-chat sandbox with cached tool data for reproducible offline evaluation. Beyond the multi-step reasoning and tool use that single-user benchmarks already test, GroupTravelBench probes three group-specific capabilities: \\textit{(i) elicitation} of private preferences through multi-turn dialogue; \\textit{(ii) coordination} of inter-user conflicts via compromise or subgrouping; and \\textit{(iii) planning} that balances group utility against fairness. We pair this with a complementary evaluation framework combining rule-based outcome metrics and LLM-judge process metrics. Across a wide range of frontier models, even the strongest agents fall short on all four rule-based outcome metrics, with plan validity below 12\\%, suggesting that group-level outcome quality is a key open challenge for LLM travel-planning agents.", "authors": ["Xiang Cheng", "Yulan Hu", "Lulu Zheng", "Zheng Pan", "Xin Li", "Yong Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-24", "url": "https://arxiv.org/abs/2605.25200", "pdf_url": "https://arxiv.org/pdf/2605.25200v2", "arxiv_id": "2605.25200", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1dbf3d581ffd95e0056adc109b507c93b1468be88bddaf49e60d13127e5a8d69", "sources": ["arxiv", "semantic_scholar"], "title": "CP-Agent: A Calibrated Risk-Controlled Agent for Feedback-Driven Competitive Programming", "abstract": "Large language models still struggle with contest-level programming, while many agentic remedies rely on massive inference-time sampling or expensive multi-stage post-training. We study when execution feedback reliably helps an LLM CP solver and which mechanisms govern the gains. We model feedback-driven solving as a calibrated stopped process and identify three quantities: false-admission risk, program-level evidence against bad programs, and the active-state success hazard. Under held-out trace calibration and selection from a pre-declared finite controller manifest, the resulting structural certificate lower-bounds the clean success probability before false admission. We instantiate mechanisms targeting these quantities as Dual-Granularity Verification, Test Augmentation, and Experience-Driven Self-Evolving, yielding CP-Agent. Without updating any parameters, CP-Agent raises Pass@1 from 25.8\\% to 48.5\\% on LiveCodeBench Pro and improves Refine@5 by 11.0\\% on ICPC-Eval. Across three LLM backbones, CP-Agent lies on the cost--accuracy efficiency frontier, and ablations show that each component primarily affects its corresponding certificate quantity.", "authors": ["Peisong Wang", "Bowen Liu", "Zehua Li", "Yuyao Wang", "Zhiwei Ma", "Yuhan Li", "Jia Li"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-23", "url": "https://arxiv.org/abs/2605.24693", "pdf_url": "https://arxiv.org/pdf/2605.24693v1", "arxiv_id": "2605.24693", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/NineAbyss/CP-Agent", "venue": null, "quality_score": 0.65}
{"id": "1f475af873f519e84f6aae61b97fe2366c659ab51ea106281bd2c6547671b58c", "sources": ["arxiv", "semantic_scholar"], "title": "Market Regime Council for Dynamic Credit Assignment in Multi-Agent LLM Decision Systems", "abstract": "Multi-agent LLM decision systems for portfolio management still lack a principled way to assign credit across specialist agents, remain vulnerable to cold-start dominance under regime shifts, and offer limited transparency into how final allocations are formed. We propose Market Regime Council (MRC), a cooperative multi-agent decision system that computes exact Shapley credits across all single, pairwise, and Grand-coalition outputs for online agent weighting. Instantiated with N=3 specialist agents, at each trading period, MRC recomputes coalition-based Shapley weights from exponentially weighted performance histories, uses a Bayesian adaptive mixture to stabilize early periods, applies regime-dependent multipliers to adjust agent authority, and records each rebalance through a five-layer causal trace. Over 1,037 trading days across 13 crypto assets and five seeds, MRC achieves a Sharpe ratio of 1.51 and a cumulative return of 440.1%, ranking first on CR, SR, and IR among active baselines and attaining the lowest MDD among active methods. Ablation results show that the gains come from Shapley-weighted integration across coalition outputs rather than from any single stage in isolation. Code and demo data are included in the supplementary material.", "authors": ["Yunhua Pei", "Zerui Ge", "Jin Zheng", "John Cartlidge"], "categories": ["cs.AI", "cs.LG", "q-fin.PM"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2026-05-23", "url": "https://arxiv.org/abs/2605.24490", "pdf_url": "https://arxiv.org/pdf/2605.24490v1", "arxiv_id": "2605.24490", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7406ca4fc3cb4f5655fa39556aedfb4043ed916ba8ae0df375b31bd2ffd2fd71", "sources": ["arxiv", "semantic_scholar"], "title": "How Many Tools Should an LLM Agent See? A Chance-Corrected Answer", "abstract": "Before an LLM agent can use a tool, a retrieval system must decide which candidate tools to show to the agent. How long should that shortlist be? Show too many tools and the model struggles to choose. Show too few and the correct tool may not appear. Most systems apply a fixed shortlist size to every query, but no standard metric exists to evaluate whether that size was appropriate. We treat the number of tools shown to an LLM agent as the object of evaluation and we apply Bits-over-Random (BoR), a chance-corrected metric that asks whether success at a given depth is better than what random selection would achieve at that same depth. We evaluate BoR across three tool-selection benchmarks, multiple scorers, and registries ranging from 20 to 3,251 tools. We then turn the same principle into a reinforcement learning (RL) reward for choosing tool shortlist depth per query. The RL agent is deliberately simple, serving as a probe of the metric rather than a proposed system. As the shortlist grows, random chance of including the correct tool rises, so the reward naturally decreases, reducing the need for an engineered depth penalty. On BFCL (370 tools), the learned policy nearly matches the coverage of showing 50 tools ($90.3\\%$ vs $90.8\\%$) while presenting only 7 on average. On ToolBench (3,251 tools), a fixed shortlist of 5 tools achieves higher aggregate coverage ($64.7\\%$ vs $61.9\\%$) but finds nothing on hard queries (correct tool ranked 6th-20th). The BoR agent finds $16.7\\%$ on those same queries by searching deeper. Downstream validation with Claude Sonnet 4.6 indicates that shorter adaptive lists also improve the LLM's ability to select the right tool: $93.1\\%$ versus $87.1\\%$ when always shown 5 tools, widening to $76.8\\%$ vs $60.9\\%$ on medium-difficulty queries where the correct tool is present but not ranked first.", "authors": ["Vyzantinos Repantis", "Ameya Gawde", "Harshvardhan Singh", "Joey Blackwell"], "categories": ["cs.IR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-23", "url": "https://arxiv.org/abs/2605.24660", "pdf_url": "https://arxiv.org/pdf/2605.24660v2", "arxiv_id": "2605.24660", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "670b69f19f9f49454535932d4b1ddf78d5b8c10a3997bb9cb47fc5d9d20adc28", "sources": ["arxiv", "semantic_scholar"], "title": "Co-ReAct: Rubrics as Step-Level Collaborators for ReAct Agents", "abstract": "ReAct-style agents for search-intensive, multi-step reasoning tasks rely largely on their own internal judgment to decide what evidence to seek, which reasoning or action step to take next, and when to stop, often producing shallow, redundant, or poorly targeted trajectories. Prior work has explored rubrics as external quality signals, but existing uses are mostly evaluative rather than action-guiding: rubrics typically serve as training-time rewards or post-hoc evaluators of completed outputs, and in deep-research settings they are often coarse-grained and report-level rather than step-level. We introduce Co-ReAct, a rubric-guided action-selection framework that uses rubrics as step-level guidance during inference. At each decision step, Co-ReAct injects a rubric into the agent's context to guide the next Reason-or-Act decision, specifying what the agent should target in evidence seeking, search, reasoning, or self-evaluation. To make this guidance reliable, we train a dedicated rubric generator with GRPO. Unlike prior pairwise or binary preference formulations, our objective optimizes a list-wise Spearman rank-correlation reward against multi-judge expert consensus rankings, encouraging rubrics that are discriminative rather than merely plausible. On DeepResearchBench and SQA-CS-V2, Co-ReAct consistently improves over ReAct and representative test-time compute baselines across search agents built on both 8B/14B open-source and frontier closed-source base models. The trained rubric generator can also serve as a drop-in component that improves these baselines without changing their underlying decision mechanisms. Our code is publicly available at https://github.com/ZBWpro/Co-ReAct.", "authors": ["Jiazheng Kang", "Bowen Zhang", "Zixin Song", "Jiangwang Chen", "Xiao Yang", "Da Zhu", "Guanjun Jiang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-22", "url": "https://arxiv.org/abs/2605.23590", "pdf_url": "https://arxiv.org/pdf/2605.23590v1", "arxiv_id": "2605.23590", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ZBWpro/Co-ReAct", "venue": null, "quality_score": 0.65}
{"id": "d1e7a27d7f20afb3cc671e783c5674047b95023989b2840e93d1742b1008ce0e", "sources": ["arxiv", "semantic_scholar"], "title": "When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs", "abstract": "Multi-agent LLM workflows route inference through specialized roles to lift end-task accuracy, but jointly training those roles with reinforcement learning is unstable in ways that are poorly understood. We study when end-to-end RL training of multi-agent LLM workflows improves over their base models, comparing Shared-Policy training, where all roles update one policy, with Isolated-Policy training, where each role has its own parameters. Our experimental matrix spans Eval-Opt, Voting, and Orch-Workers workflows, math and code tasks, and three model scales (0.6B, 1.7B, 4B). We find that multi-agent RL usually improves over base models, but gains depend jointly on workflow, task, and scale, not on policy sharing alone. Isolated-Policy tends to reach higher peak accuracy yet more often falls off a terminal accuracy cliff, while Shared-Policy training does not eliminate failure; it redistributes failure into qualitatively different patterns. We then explain the strongest of these patterns through role-level gradient dynamics induced by workflow topology and policy routing: under Isolated-Policy, parallel same-role agents on shared prompts amplify per-role gradients and drive terminal degradation in Voting and Orch-Workers workflows; under Shared-Policy, asymmetric per-step gradient mass causes the shared policy to be captured by the dominant role, producing different failure signatures by task and workflow. Together, the empirical map and its underlying mechanisms show that policy sharing routes training pressure through different channels rather than offering uniform stability, making it a design choice with workflow- and task-conditional tradeoffs.", "authors": ["Yifan Zeng", "Yiran Wu", "Yaolun Zhang", "Wentian Zhao", "Kun Wan", "Qingyun Wu", "Huazheng Wang"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-22", "url": "https://arxiv.org/abs/2605.24202", "pdf_url": "https://arxiv.org/pdf/2605.24202v2", "arxiv_id": "2605.24202", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3bf1db92652508b5d7d7cd75770c5e4d81d5eed6e681c66335523639275b411e", "sources": ["arxiv", "semantic_scholar"], "title": "SynAE: A Framework for Measuring the Quality of Synthetic Data for Tool-Calling Agent Evaluations", "abstract": "Today, tool-calling agents are commonly evaluated or tested on static datasets of execution traces, including input commands, agent responses, and associated tool calls. However, internal production datasets are often insufficient or unusable for testing; for example, they may contain sensitive or proprietary data, or they may be too sparse to support comprehensive testing (especially pre-deployment). In these settings, practitioners are increasingly replacing or augmenting real datasets with synthetic ones for evaluation purposes. A key challenge is quantifying the relation between these synthetic datasets and the real data. We introduce SynAE, an evaluation framework for assessing how well synthetic benchmarks for multi-turn, tool-calling agents replicate and augment the characteristics of real data trajectories. SynAE assesses the validity, fidelity, and diversity of synthetic data across four metric categories: (i) task instructions and intermediate responses, (ii) tool calls, (iii) final outputs, and (iv) downstream evaluation. We evaluate SynAE using recent agent benchmarks and test common synthetic data failure modes via realistic and controlled generation schemes. SynAE detects fine-grained variations in data validity, fidelity and diversity, and shows that no single metric is sufficient to fully characterize synthetic data quality, motivating a multi-axis evaluation of synthetic data for agent testing. A demo of SynAE is available at https://synae-2026-synae-demo.static.hf.space/index.html, with code at https://github.com/wsqwsq/SynAE.", "authors": ["Shuaiqi Wang", "Aadyaa Maddi", "Zinan Lin", "Giulia Fanti"], "categories": ["cs.CL", "cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-21", "url": "https://arxiv.org/abs/2605.22564", "pdf_url": "https://arxiv.org/pdf/2605.22564v1", "arxiv_id": "2605.22564", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wsqwsq/SynAE", "venue": null, "quality_score": 0.65}
{"id": "56e67ef4ae9754c9c2bd708e7ffb2b646299d0bb391c6c2e313962c999db052f", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents", "abstract": "Agentic systems are becoming more capable: agents define strategies, take actions, and interact with different environments. This autonomy poses serious challenges for overseeing and assessing agent behavior. Most current tools are limited, focusing on observability with basic evaluation capabilities or imposing static, hand-crafted error taxonomies that cannot adapt to new domains. To address this gap, we present Agentic CLEAR, an automatic, dynamic, and easy-to-use evaluation framework. It produces textual insights into the agent behavior on three levels of granularity: system, trace, and node. Agentic CLEAR operates above the observability layer, enabling seamless integration and featuring an intuitive UI that makes agent evaluation highly accessible. In our experiments on four benchmarks, seven agentic settings, and tens of thousands of LLM calls, we show that Agentic CLEAR produces high-quality, data-driven, insightful feedback. Our analysis shows strong alignment with human-annotated errors and the ability to predict task success rate.", "authors": ["Asaf Yehudai", "Lilach Eden", "Michal Shmueli-Scheuer"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-21", "url": "https://arxiv.org/abs/2605.22608", "pdf_url": "https://arxiv.org/pdf/2605.22608v1", "arxiv_id": "2605.22608", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a6843d76b2c0051b984fe5b00098968479495820a7c49c8b2c9c474bd3c7d4a0", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Evolving Multi-Agent Systems via Decentralized Memory", "abstract": "Self-evolving multi-agent systems (MAS) have emerged as a promising route to LLM agents that continually improve from experience, with persistent memory at their foundation. However, existing designs almost exclusively adopt a centralized repository shared across agents, incurring communication and coordination overhead, raising privacy concerns, and collapsing agent diversity. We propose DecentMem, a decentralized memory framework in which each agent maintains its own dual-pool memory -- an exploitation pool of consolidated past trajectories and an exploration pool of LLM-generated candidates for unseen contexts. The two pools are reweighted online based on stage-wise feedback from an LLM-as-a-judge. Theoretically, we prove that this design guarantees global reachability of the solution space and achieves $O(\\log T)$ cumulative regret, matching the stochastic bandit lower bound up to constants. In practice, across three MAS frameworks (AutoGen, DyLAN, AgentNet), three Qwen3 backbones (4B/8B/14B), two Gemma4 backbones (E2B/E4B) and five benchmarks spanning math, code, QA, and embodied tasks, DecentMem improves average accuracy by up to 23.8% over the strongest centralized memory baseline and by up to 52.5% over the no-memory baseline, while reducing token usage by up to 49%.", "authors": ["Guangya Hao", "Yunbo Long", "Zhuokai Zhao"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-21", "url": "https://arxiv.org/abs/2605.22721", "pdf_url": "https://arxiv.org/pdf/2605.22721v1", "arxiv_id": "2605.22721", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f634d855fde6c9068afdeca8c609cc061614885b3f5c511e4b25748723e579a8", "sources": ["arxiv", "semantic_scholar"], "title": "What Twelve LLM Agent Benchmark Papers Disclose About Themselves: A Pilot Audit and an Open Scoring Schema", "abstract": "We read twelve well-known LLM agent benchmark papers and recorded, dimension by dimension, what each paper actually says about how its evaluation was run. The motivation came from a familiar frustration: two papers will report results on the same benchmark with the same model name and disagree, and you cannot tell why -- the scaffold, the sampling settings, the subset, or the evaluator version. In many cases the published artifact does not let you answer. This paper is an implementation report on the attempt. We designed a small audit schema (five fields: benchmark identity, harness specification, inference settings, cost reporting, failure breakdown), wrote a scoring codebook with the boundary cases we hit during pilot scoring, applied it to twelve canonical papers (eight agent, four classical static), and recorded what we saw. We score the disclosure of an agent run, not its correctness, and make no claim that disclosure implies a trustworthy result. The mean audit score across the eight agent-benchmark papers is 0.38 (out of 1.0), and across the four classical static benchmarks 0.66; the largest gap is on cost (none of the eight agent benchmark papers disclose inference cost in any form) and on harness specification (none fully disclose a content-addressed container image of the evaluation environment). We release the schema as a JSON Schema file, the codebook as a Markdown document, and the raw scoring sheet as a CSV. The scoring was performed by a single auditor in one pass; a multi-rater audit is the natural next step, and we discuss what we think it would change.", "authors": ["Mahdi Naser Moghadasi", "Faezeh Ghaderi"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-20", "url": "https://arxiv.org/abs/2605.21404", "pdf_url": "https://arxiv.org/pdf/2605.21404v1", "arxiv_id": "2605.21404", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "83bc815d14a58a2dd2ec7f3f165b24f6dde2bff92b1e42fdb5cbdb33e8d636c2", "sources": ["arxiv", "semantic_scholar"], "title": "DynaMate2: runtime registration of expert-defined tools for agentic scientific workflow automation", "abstract": "Agentic large-language-model systems can coordinate scientific tools, but many implementations remain difficult for domain scientists to extend without modifying the source orchestration code or relying on unconstrained code generation. DynaMate2 is a LangGraph-based multi-agent framework for converting expert-defined Python functions into persistent AI-callable tools. The architecture separates domain execution from LLM supervision: registered tools perform scientific operations, while a supervisor LLM decomposes goals, selects specialist agents, routes inputs, and propagates outputs across steps. DynaMate2 supports: runtime tool registration from inline code, source files, and explicitly requested natural-language specifications; persistent storage of tools, agents, and conversation state; and a web interface for interactive workflow assembly. We demonstrate the framework on a molecular simulation workflow in which a single instruction retrieves a MACE foundation model, builds a NaCl-water configuration, runs an ASE molecular dynamics trajectory, and generates energy and temperature diagnostics. The demonstration illustrates how validated workflow components can be composed into a supervised agentic pipeline without rewriting the framework. DynaMate2 therefore provides a reusable template for extending LLM-based automation to research groups with existing Python workflows, while preserving the need for explicit tool validation, reproducibility logs, and deployment-specific safeguards.", "authors": ["Orlando A. Mendible-Barreto", "Ajay Vallabh", "Ubaldo M. Córdova-Figueroa", "Yamil J. Colón"], "categories": ["physics.chem-ph"], "fields_of_study": ["Physics"], "published_date": "2026-05-20", "url": "https://arxiv.org/abs/2605.20819", "pdf_url": "https://arxiv.org/pdf/2605.20819v2", "arxiv_id": "2605.20819", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7c1861f9290144204b1de13a9cb05b4c0c3e0db37d6610f72e07bcdd20446c77", "sources": ["arxiv", "semantic_scholar"], "title": "TO-Agents: A Multi-Agent AI Pipeline for Preference-Guided Topology Optimization", "abstract": "Topology optimization can generate efficient structures, but designers often must manually translate qualitative intent, such as desired visual style, product experience, or manufacturability into solver settings that are not directly tied to those preferences. We present TO-Agents, a multi-agent AI framework that connects natural-language design intent with iterative topology optimization. The framework converts a human-provided problem description into validated solver inputs, runs a topology optimization solver, renders the resulting 3D topology, and uses multi-view vision-language reasoning with an independent judge agent to critique each result and revise solver parameters. We evaluate the framework on two long-horizon design tasks: a cantilever beam benchmark and a phone-stand product design. In both tasks, the designer specifies an aesthetic preference for hierarchically branched structures inspired by natural tree morphologies, and the system performs four revision cycles across ten independent replicates. TO-Agents produces at least one preference-aligned design in 60% of trials for each case study, corresponding to up to 6x more successful trials than an ablated pipeline without visual or historical feedback. Judge scores and human evaluations show that the pipeline can identify effective parameter levers, recover from poor revisions, and expand design exploration. A manufacturing agent further post-processes top-ranked designs for additive manufacturing, enabling end-to-end intent-to-prototype design. We also identify failure modes, including overshooting, selective memory, misplaced tools, and incorrect parameter reasoning. These results suggest that agentic topology optimization can shift designers from low-level parameter tuning toward higher-level specification of form and function, while highlighting safeguards needed for reliable autonomous engineering design.", "authors": ["Isabella A. Stewart", "Hongrui Chen", "Faez Ahmed"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-20", "url": "https://arxiv.org/abs/2605.21622", "pdf_url": "https://arxiv.org/pdf/2605.21622v1", "arxiv_id": "2605.21622", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "de557b2ce0d60421f5c4aa6a1c64b0bb0bb9d777d12626419e59e6ae3b6df601", "sources": ["arxiv", "semantic_scholar"], "title": "From Automated to Autonomous: Hierarchical Agent-native Network Architecture (HANA)", "abstract": "Realizing Level 4/5 Autonomous Networks (AN) demands a shift from static automation to agent-native intelligence. Current operations, reliant on rigid scripts, lack the cognitive agency to handle off-nominal conditions. To address this, this letter proposes a hierarchical multi-agent reference architecture enabling high-level autonomy. The framework features a Dual-Driven Orchestrator that coordinates specialized Executive Agents, supported by a shared Public Memory for unified domain knowledge. A key innovation is the integration of agent self-awareness, which empowers the system to harmonize deliberative strategic governance with reflexive fault recovery. We instantiate and validate this architecture within a 5G Core environment. Case studies demonstrate that the system sustains critical throughput under congestion and reduces Mean Time to Repair (MTTR) by 86%, confirming its efficacy in unifying strategic planning with operational resilience.", "authors": ["Binghan Wu", "Shoufeng Wang", "Yunxin Liu", "Ya-Qin Zhang", "Joseph Sifakis", "Ye Ouyang"], "categories": ["cs.AI", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-20", "url": "https://arxiv.org/abs/2605.20608", "pdf_url": "https://arxiv.org/pdf/2605.20608v1", "arxiv_id": "2605.20608", "doi": "10.1109/LNET.2026.3693226", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Networking Letters", "quality_score": 0.55}
{"id": "2bc6115a3e47a32f10a28b6777dedacd7924f55404133735a9d5b73f2e64f6db", "sources": ["arxiv", "semantic_scholar"], "title": "MemGym: a Long-Horizon Memory Environment for LLM Agents", "abstract": "Memory is a central capability for LLM agents operating across long-horizon tasks. Existing memory benchmarks predominantly evaluate retention of personalized information in multi-turn chat scenarios, overlooking the dynamic memory formation that occurs during extended agent execution. Consequently, the memory systems they produce transfer poorly to realistic agentic environments, such as coding and web navigation. We present MemGym, a benchmark for agentic memory that unifies existing agent gyms and in-house memory-grounded pipelines behind one memory-reasoning interface. MemGym spans five evaluation tracks grouped into four agentic regimes: tool-use dialogue (tau2-bench), multi-turn deep-research search (MEMGYM-DR), coding (SWE-Gym and MEMGYM-CODEQA), and computer use (WebArena-Infinity). MemGym reports memory-isolated scores that decouple memory performance from reasoning, retrieval, and tool-use ability, so memory strategies can be ranked without those confounders. Our synthetic pipelines for MEMGYM-CODEQA and MEMGYM-DR are length-controllable, ablation-verified at every stage, and tightly aligned with downstream scenarios. To make evaluation on coding environments academically tractable, we train MemRM, a lightweight reward model (Qwen3-1.7B fine-tuned with QLoRA) that scores compression quality as a fast scalar read in place of full Docker rollouts.", "authors": ["Wujiang Xu", "Yu Wang", "Kai Mei", "Kaiqu Liang", "Zhenting Wang", "Mingyu Jin", "Han Zhang", "Shi-Xiong Zhang", "Wenyue Hua", "Sambit Sahu", "Dimitris N. Metaxas"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-20", "url": "https://arxiv.org/abs/2605.20833", "pdf_url": "https://arxiv.org/pdf/2605.20833v1", "arxiv_id": "2605.20833", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5a3f2686f860fd6c68f956ce490947a40a43f4b8d401c914e3c12963fb3c1f3f", "sources": ["arxiv", "semantic_scholar"], "title": "CASPIAN: Online Detection and Attribution of Cascade Attacks in LLM Multi-Agent Systems via Cross-Channel Causal Monitoring", "abstract": "Cascade attacks in LLM multi-agent systems (MAS) arise when adversarial influence propagates across agents and leads to escalated system-level failures through complex agent interactions. Detecting such cascades is challenging, as their signals are distributed, tightly coupled across interaction channels, and often appear plausibly benign locally but may unfold quickly either within a single turn or gradually across multiple turns. Existing defenses, being largely local and text-centric, fail to capture such cross-channel, temporally coordinated dynamics of cascade propagation. Therefore, we propose CASPIAN, the first framework that provides a unified, cross-channel causal analysis of cascade behavior in LLM-MAS through online monitoring of dynamic influence propagation across agents. CASPIAN models multi-agent interactions using a unified, dynamic causal influence matrix across channels, estimated efficiently via a late-interaction conditional transfer entropy (LI-CTE) formulation, thereby enabling the detection of cascade onset from emergent system-level structure rather than isolated anomalies. It further performs online causal attribution, identifying the origin, bridge, and amplifier agents driving the cascade and reconstructing its principal propagation pathways, capabilities not supported by existing methods. Across diverse multi-agent frameworks and benchmarks, CASPIAN consistently outperforms semantic guardrails, LLM-based judges, and graph-based anomaly detectors in both detection accuracy and early cascade identification while operating with sub-1% relative overhead latency. These results demonstrate that unified cross-channel causal modeling is essential for reliably detecting and understanding cascade failures in LLM multi-agent systems.", "authors": ["Kavana Venkatesh", "Jafar Isbarov", "Saad Amin", "Murat Kantarcioglu", "Jiaming Cui"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.19240", "pdf_url": "https://arxiv.org/pdf/2605.19240v1", "arxiv_id": "2605.19240", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/caspian-detector/caspian", "venue": null, "quality_score": 0.65}
{"id": "ef0ec7457174e454cb2e39324d3814496beadcb5663ec786a17d1b7eaf151c74", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Collaboration with State Management", "abstract": "Recent advances in multi-agent systems have shown great potential for solving complex tasks. However, when multiple agents edit a shared codebase concurrently, their changes can silently conflict and inconsistent views lead to integration failures. Existing multi-agent systems address this through workspace isolation (e.g., one git worktree per agent), but this defers conflict resolution to a post-hoc merge step where recovery is expensive. In this paper, we propose STORM, i.e., STate-ORiented Management for multi-agent collaboration. Specifically, STORM manages agent states by mediating their interactions with the shared workspace, ensuring that each agent operates on a consistent view of the codebase and that conflicting edits are detected and resolved at write time. We evaluate STORM on Commit0 and PaperBench across multiple LLMs. STORM outperforms the git-worktree-based multi-agent baseline by +18.7 on Commit0-Lite and +1.4 on PaperBench, while achieving comparable or better cost efficiency. Combined with single-agent runs, STORM reaches highest scores of 87.6 and 78.2 on the two benchmarks respectively, suggesting that explicit state management is a more effective foundation for multi-agent collaboration than workspace isolation. STORM can also be plugged into any multi-agent system seamlessly.", "authors": ["Mengyang Liu", "Taozhi Chen", "Zhenhua Xu", "Xue Jiang", "Yihong Dong"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.20563", "pdf_url": "https://arxiv.org/pdf/2605.20563v1", "arxiv_id": "2605.20563", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c39ea1406741a7a1de6a94cb686478588f16a41a9f778225b485d9756bde1543", "sources": ["arxiv", "semantic_scholar"], "title": "EngiAI: A Multi-Agent Framework and Benchmark Suite for LLM-Driven Engineering Design", "abstract": "Large Language Model (LLM) agents are increasingly applied to engineering design tasks, yet existing evaluation frameworks do not adequately address multi-agent systems that combine simulation, retrieval, and manufacturing preparation. We introduce a benchmark suite with three evaluation dimensions: (1) a workflow benchmark with seven prompt styles targeting distinct cognitive demands-including direct tool use, semantic disambiguation, conditional branching, and working-memory tasks; (2) a Retrieval-Augmented Generation (RAG) benchmark with gated scoring isolating retrieval contributions to parameter selection; and (3) an High Performance Computing (HPC) benchmark evaluating end-to-end ML training orchestration on a SLURM cluster. Alongside the benchmark we present EngiAI, a Multi-Agent System (MAS) reference implementation built on LangGraph that operationalizes the benchmark by coordinating seven specialized agents through a supervisor architecture, unifying topology optimization, document retrieval, HPC job orchestration, and 3D printer control. Across four LLM backends and two EngiBench problems, proprietary models achieve 96-97% average task completion on Beams2D, while open-source 4B-parameter models reach 55-78%, with clear generational improvement. Conditional branching proves most challenging, with task completion dropping to 20-53% for the conditional style on Photonics2D. RAG gating confirms near-perfect retrieval-augmented scores (about 1.0) versus near-zero without retrieval, validating the evaluation design. On HPC orchestration, one model completes all pipeline steps in 100% of runs while another drops to 50%, revealing that multi-step instruction following degrades over long-running workflows.", "authors": ["Gioele Molinari", "Florian Felten", "Soheyl Massoudi", "Mark Fuge"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.19743", "pdf_url": "https://arxiv.org/pdf/2605.19743v2", "arxiv_id": "2605.19743", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "9e1ff3183cd1fe9e45dde36c5e042e33a209985747d984f82ddf80b2dba8172f", "sources": ["arxiv", "semantic_scholar"], "title": "A Methodology for Selecting and Composing Runtime Architecture Patterns for Production LLM Agents", "abstract": "Production LLM agents combine stochastic model outputs with deterministic software systems, yet the boundary between the two is rarely treated as a first-class architectural object. This paper names that boundary the stochastic-deterministic boundary (SDB): a four-part contract among a proposer, verifier, commit step, and reject signal that specifies how an LLM output becomes a system action. We argue that the SDB is the load-bearing primitive of production agent runtimes. Around this primitive, we organize agent runtime design into three concerns: Coordination, State, and Control. We present a catalog of six runtime patterns that compose the SDB differently across conversational, autonomous, and long-horizon agents: hierarchical delegation, scatter-gather plus saga, event-driven sequencing, shared state machine, supervisor plus gate, and human in the loop. For each pattern, we trace its lineage to distributed-systems concepts and identify what changes when the worker is stochastic. The paper contributes a five-step methodology for selecting runtime patterns, a diagnostic procedure that maps production failures to pattern weaknesses, and a failure mode called replay divergence, in which LLM-based consumers of a deterministic event log produce different downstream outputs under model-version or prompt changes. A stylized reliability decomposition separates per-call model variance from architectural momentum, motivating the claim that as model variance decreases, pattern choice and SDB strength become increasingly important levers for long-run reliability. We apply the methodology to five workloads and provide one runnable reference implementation for a 90-day contract-renewal agent.", "authors": ["Vasundra Srinivasan"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.20173", "pdf_url": "https://arxiv.org/pdf/2605.20173v1", "arxiv_id": "2605.20173", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/vasundras/agent-runtime-patterns", "venue": null, "quality_score": 0.65}
{"id": "a9fd92403376284f90d4f45da3b75461842a9160cc00aa9cc2905f5c7b2446cf", "sources": ["arxiv", "semantic_scholar"], "title": "ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning", "abstract": "Training large multimodal models (LMMs) via reinforcement learning (RL) to natively invoke video-processing tools (e.g., cropping) has become a promising route to long-video understanding. However, existing native-RL methods dispatch tool calls sequentially (i.e., one per turn): a single wrong crop propagates errors without peer correction, multi-turn tool calls corrupt context, and inference cost scales linearly with the number of turns. We introduce ParaVT, the first multi-agent end-to-end RL-trained framework for Parallel Video Tool calling, dispatching multiple time-window crops in a single turn for cleaner context and better fault tolerance. Yet applying standard RL to ParaVT reveals an obstacle we term the Tool Prior Paradox: the pretrained tool priors that enable tool exploration also destabilize cold-started structural format and expose the skip-tool reward shortcut under temperature sampling. A cross-model contrast on a weaker-prior LMM supports this claim: format stays stable but RL elicits zero tool calls, indicating that prior strength is the shared driver of both format collapse and tool exploration. We propose PARA-GRPO (Parseability-Anchored and Ratio-gAted GRPO), which augments standard RL with two complementary mechanisms: (i) a targeted format reward applied only at the structural-token positions most prone to collapse, and (ii) a per-prompt frame-budget randomization that creates training prompts where calling the tool yields a measurable reward signal over skipping it. Across six long-video understanding benchmarks, ParaVT improves over the Qwen3-VL baseline by +7.9% on average, with PARA-GRPO lifting training-time format compliance from 0.13 to 0.64. As tool capabilities become increasingly internalized in modern LMMs, RL must cooperate with the resulting priors, and ParaVT offers a general recipe for agentic RL. Code, data, and model weights are publicly available.", "authors": ["Zuhao Yang", "Kaichen Zhang", "Sudong Wang", "Keming Wu", "Zhongyu Yang", "Bo Li", "Xiaojuan Qi", "Shijian Lu", "Xingxuan Li", "Lidong Bing"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.20342", "pdf_url": "https://arxiv.org/pdf/2605.20342v2", "arxiv_id": "2605.20342", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c14a952a4a42efb330843fb7d5cdae09ff8aa24c1ce0c2ea16595a1712a5037c", "sources": ["arxiv", "semantic_scholar"], "title": "Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs", "abstract": "LLM agents have recently emerged as a powerful paradigm for solving complex tasks through planning, tool use, memory retrieval, and multi-step interaction. However, these agentic workflows often introduce substantial input-side overhead, making the compute-intensive prefilling stage a key bottleneck in long-context, multi-turn inference. In this work, we propose Mix-Quant, a simple and effective phase-aware quantization framework for fast agentic inference. We first investigate FP4 quantization in agentic LLM workflows and observe that quantizing the entire inference process can incur significant performance degradation. In contrast, the prefilling stage exhibits substantial quantization redundancy and can therefore be quantized with minimal accuracy loss, despite being the dominant source of computation. Based on this insight, we apply high-throughput NVFP4 quantization to the prefilling phase while preserving BF16 precision for decoding. By decoupling prefilling acceleration from decoding quality, Mix-Quant combines phase-aware algorithmic quantization with hardware-efficient NVFP4 execution to alleviate the inference bottleneck in LLM agents. Extensive experiments across long-context and agentic benchmarks demonstrate that Mix-Quant largely preserves task performance while delivering significant efficiency improvements, achieving up to a 3x speedup during prefilling.", "authors": ["Haiquan Lu", "Zigeng Chen", "Gongfan Fang", "Xinyin Ma", "Xinchao Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.20315", "pdf_url": "https://arxiv.org/pdf/2605.20315v1", "arxiv_id": "2605.20315", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6d565b4c9ea15c8d07c6fd6209eb28aa2ce85af8087235e3edffaceb5324ff51", "sources": ["arxiv", "semantic_scholar"], "title": "APS: Bias-Controlled Adaptive Prototype Simulation for Population-Scale LLM Agents", "abstract": "LLM-agent simulation offers a flexible computational tool for studying population response trajectories that depend on scenario events, memory, demographics, and evolving social context. However, full multi-round simulation scales linearly with both population size and horizon, requiring every agent to query the LLM at every round. We propose Adaptive Prototype Simulation (APS), a framework that reframes scalable LLM-based simulation as a recurrent oracle-allocation problem. APS retains the designated LLM as the online transition oracle while querying adaptive core prototypes, selected singleton-tail agents, and shadow-audit agents. Prototype responses induce local response surfaces for nearby agents, reducing online LLM calls without replacing the underlying transition model. To control approximation bias, shadow-audit residual correction estimates propagation residuals for aggregate correction and future budget allocation, while tail-protected singleton routing directly queries selected isolated, heterogeneous, or high-curvature regions that are vulnerable to smoothing. Theoretically, we treat APS as an estimator for full-scale high-precision individual social simulation and decompose its errors into prototype-coverage error, shadow-audit residual-correction error, local-propagation bias, and temporal context mismatch. Under the reported protocols, APS gives lower reference-aligned distributional discrepancy than scale-oriented and same-budget baselines while reducing online LLM calls, with ablations and compact robustness checks diagnosing the main bias-control mechanisms. In a 10M-agent, multi-round public-opinion simulation, APS achieves a 381.1-fold reduction over full simulation, with reference-aligned final-round JSD of 0.094 against the corresponding full-LLM reference.", "authors": ["Quan Zheng", "Yan Gao", "Shaobin He", "Haoxiang Guan", "Yuanhe Tian", "Jie Feng", "Ming Wang", "Shuxin Zheng", "Zhen Liu"], "categories": ["cs.MA", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.27419", "pdf_url": "https://arxiv.org/pdf/2605.27419v1", "arxiv_id": "2605.27419", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4e93119299b2bde502e7c5226ecef80199853597790c03ebe91e5e88702c00a5", "sources": ["arxiv", "semantic_scholar"], "title": "Conflict-Resilient Multi-Agent Reasoning via Signed Graph Modeling", "abstract": "LLM-based multi-agent systems (MAS) have demonstrated strong reasoning and decision-making capabilities that consistently surpass those of single LLM agents. However, their performance often suffers from naive aggregation mechanisms that assume uniformly cooperative interactions. Upon close inspection, we observe that existing graph-based MAS frameworks (1) propagate errors when conflicting signals arise without control, and (2) lack explicit modeling of conflicting inter-agent relations as well as structural awareness, failing to identify reliable interaction patterns. To bridge this gap, we introduce SIGMA, a novel SIgned Graph-informed Multi-Agent reasoning framework that explicitly captures trust, conflict, and neutral relations among agents via a signed relational graph. Specifically, given a query, SIGMA first selects a set of relevant and diverse agents, then constructs a structured signed interaction graph with confidence-weighted edges. Reasoning proceeds through conflict-aware signed message passing, which reinforces information from trustworthy agents while suppressing conflicting signals, and terminates with a structure- and conflict-aware weighted aggregation to yield globally consistent and conflict-resilient predictions. Extensive experiments on six benchmark datasets, across multiple LLM backbones and diverse multi-agent configurations, demonstrate that SIGMA consistently outperforms state-of-the-art baselines, achieving notable gains in both accuracy and conflict-resilient performance.", "authors": ["Longgang He", "Longzhu He", "Daojing He", "Chaozhuo Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-19", "url": "https://arxiv.org/abs/2605.19418", "pdf_url": "https://arxiv.org/pdf/2605.19418v1", "arxiv_id": "2605.19418", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "fc6cac82eafe90e40fa6277d56461957ed8237b6d6afb421a66e82150a939b23", "sources": ["arxiv", "semantic_scholar"], "title": "Code as Agent Harness", "abstract": "Recent large language models (LLMs) have demonstrated strong capabilities in understanding and generating code, from competitive programming to repository-level software engineering. In emerging agentic systems, code is no longer only a target output. It increasingly serves as an operational substrate for agent reasoning, acting, environment modeling, and execution-based verification. We frame this shift through the lens of agent harnesses and introduce code as agent harness: a unified view that centers code as the basis for agent infrastructure. To systematically study this perspective, we organize the survey around three connected layers. First, we study the harness interface, where code connects agents to reasoning, action, and environment modeling. Second, we examine harness mechanisms: planning, memory, and tool use for long-horizon execution, together with feedback-driven control and optimization that make harness reliable and adaptive. Third, we discuss scaling the harness from single-agent systems to multi-agent settings, where shared code artifacts support multi-agent coordination, review, and verification. Across these layers, we summarize representative methods and practical applications of code as agent harness, spanning coding assistants, GUI/OS automation, embodied agents, scientific discovery, personalization and recommendation, DevOps, and enterprise workflows. We further outline open challenges for harness engineering, including evaluation beyond final task success, verification under incomplete feedback, regression-free harness improvement, consistent shared state across multiple agents, human oversight for safety-critical actions, and extensions to multimodal environments. By centering code as the harness of agentic AI, this survey provides a unified roadmap toward executable, verifiable, and stateful AI agent systems.", "authors": ["Xuying Ning", "Katherine Tieu", "Dongqi Fu", "Tianxin Wei", "Zihao Li", "Yuanchen Bei", "Jiaru Zou", "Mengting Ai", "Zhining Liu", "Ting-Wei Li", "Lingjie Chen", "Yanjun Zhao", "Ke Yang", "Bingxuan Li", "Cheng Qian", "Gaotang Li", "Xiao Lin", "Zhichen Zeng", "Ruizhong Qiu", "Sirui Chen", "Yifan Sun", "Xiyuan Yang", "Ruida Wang", "Rui Pan", "Chenyuan Yang", "Dylan Zhang", "Liri Fang", "Zikun Cui", "Yang Cao", "Pan Chen", "Dorothy Sun", "Ren Chen", "Mahesh Srinivasan", "Nipun Mathur", "Yinglong Xia", "Hong Li", "Hong Yan", "Pan Lu", "Lingming Zhang", "Tong Zhang", "Hanghang Tong", "Jingrui He"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.18747", "pdf_url": "https://arxiv.org/pdf/2605.18747v1", "arxiv_id": "2605.18747", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/YennNing/Awesome-Code-as-Agent-Harness-Papers", "venue": null, "quality_score": 0.65}
{"id": "de5d4ba67a1868992899afe32f4bffff0b1a56682bdd6953bcc6e4e36d33e01b", "sources": ["arxiv", "semantic_scholar"], "title": "PROTEA: Offline Evaluation and Iterative Refinement for Multi-Agent LLM Workflows", "abstract": "Multi-agent LLM workflows -- systems composed of multiple role-specific LLM calls -- often outperform single-prompt baselines, but they remain difficult to debug and refine. Failures can originate from subtle errors in intermediate outputs that propagate to downstream nodes, requiring developers to inspect long traces and infer which agent to modify. We present PROTEA, a unified interface for offline, test-driven improvement of multi-agent workflows. PROTEA executes a workflow, scores intermediate node outputs with configurable rubrics, and overlays per-node states and rationales on the workflow graph to localize likely bottlenecks. To support complex systems where final-answer references are the primary supervision, PROTEA performs backward node evaluation: it generates candidate node-level expectations from final-answer references and graph context, then compares them with observed node outputs. For selected nodes, PROTEA presents targeted prompt revisions as editable before/after comparisons, then automatically reruns and re-evaluates the workflow to show output changes and score trajectories within the same interface. In two production-adjacent workflows, PROTEA improved document-inspection accuracy from 64.3% to 83.9% and recommendation Hit@5 from 0.30 to 0.38. In a formative study with six experienced LLM developers, participants valued graph-level localization, per-node rationales, and editable before/after prompt revisions.", "authors": ["Kazuki Kawamura", "Satoshi Waki", "Kei Tateno"], "categories": ["cs.CL", "cs.AI", "cs.HC", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.18032", "pdf_url": "https://arxiv.org/pdf/2605.18032v1", "arxiv_id": "2605.18032", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9d6beb984b78c83f21788c676031df5c83b5e0d183528dbf74c0806cce62517f", "sources": ["arxiv", "semantic_scholar"], "title": "Sequential Consensus for Multi-Agent LLM Debates: A Wald-SPRT compute governor with calibration-based failure detection", "abstract": "Multi-agent LLM debate improves factuality and reasoning, but most recipes pick a fixed round count, over-spending on easy items and under-spending on hard ones. We adapt Wald's Sequential Probability Ratio Test (SPRT) as a plug-in compute governor for LLM debates. After each round, an LLM judge emits a [0,1] consensus score on the latest agent positions; a Wald monitor accumulates the log-likelihood ratio of \"useful convergence\" vs \"not yet useful\" under a Beta likelihood family, and stops when either boundary is crossed or returns a capped best-effort outcome at R_max. Under i.i.d. assumptions the rule inherits SPRT type-I/type-II error guarantees; in deployment the calibration itself is the more important object, since it estimates whether the judge score actually separates useful from unhelpful convergence in a given domain. We evaluate two tracks: (i) a Monte-Carlo study under calibrated Beta models characterising working curves, error rates, capping behaviour, and sensitivity; and (ii) a real-LLM evaluation on 200 attempted MMLU and 200 attempted GSM8K items with three heterogeneous agents (gpt-5, claude-opus-4-6, gemini-2.5-pro) and a claude-opus-4-6 judge, using disjoint 40-item calibration subsets. On GSM8K the rule stops in 1.01 average rounds (4.06 LLM calls) at 97.0% accuracy vs 99.0% for fixed-5 debate at 15 calls: a 3.7x call reduction at -2pp accuracy. On MMLU the calibrated KL collapses to about 0 and the rule caps on 99.5% of items at 2.1x cost. The takeaway is not that SPRT makes debate more accurate, but that a classical sequential test serves as a cheap compute-control and failure-detection layer for multi-agent LLM systems.", "authors": ["Andrea Morandi"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.19193", "pdf_url": "https://arxiv.org/pdf/2605.19193v1", "arxiv_id": "2605.19193", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5111e7da38474f17944b806c2c9e8887d06ed088809c1ce76b0512b4cf9023d5", "sources": ["arxiv", "semantic_scholar"], "title": "Harnessing LLM Agents with Skill Programs", "abstract": "Equipping LLM agents with reusable skills derived from past experience has become a popular and successful approach for tackling complex and long-horizon tasks. However, such lessons are often encoded as textual guidance that remains largely advisory, lacking explicit mechanisms for when and how to intervene in the agent loop. To bridge the gap, we introduce HASP(Harnessing LLM Agents with Skill Programs), a new framework that upgrades skills into executable Program Functions (PFs). Rather than offering passive advice, PFs act as executable guardrails that activate on failure-prone states and modify the next action or inject corrective context. HASP is highly modular: it can be applied at inference time for direct agent-loop intervention, during post-training to provide structured supervision, or for self-improvement by evolving validated, teacher-reviewed PFs. Empirically, HASP drives substantial gains compared to both training-free and training-based methods on web-search, math reasoning, and coding tasks. For example, on web-search reasoning, inference-time PFs alone improve the average performance by 25% compared to (multi-loop) ReAct Agent, while post-training and controlled evolution achieve a 30.4% gain over Search-R1. To provide deeper insights into HASP, our mechanism analysis reveals how PFs trigger and intervene, how skills are internalized, and the requirement for stable skill library evolution.", "authors": ["Hongjun Liu", "Yifei Ming", "Shafiq Joty", "Chen Zhao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.17734", "pdf_url": "https://arxiv.org/pdf/2605.17734v1", "arxiv_id": "2605.17734", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c1b2aaf032e7abb83c3247cb58d55608c964cdcc523e1fd00b81ac35851ddd9e", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning", "abstract": "Communication is a key component in multi-agent reinforcement learning (MARL) for mitigating partial observability, yet prior approaches often rely on inefficient information exchange or fail to transmit sufficient state information. To address this, we propose LLM-driven Multi-Agent Communication (LMAC), which leverages an LLM's reasoning capability to design a communication protocol that enables all agents to reconstruct the underlying state as accurately and uniformly as possible. LMAC iteratively refines the protocol using an explicit state-awareness criterion, improving state recovery while narrowing differences in agents' knowledge. Experiments on diverse MARL benchmarks show that LMAC improves state reconstruction across agents and yields substantial performance gains over prior communication baselines.", "authors": ["Sangjun Bae", "Yisak Park", "Sanghyeon Lee", "Seungyul Han"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.18077", "pdf_url": "https://arxiv.org/pdf/2605.18077v2", "arxiv_id": "2605.18077", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "636f8f454c78b86eab408aca1a28299e02b0daa322dd64a94d5b9cd531ab8603", "sources": ["arxiv", "semantic_scholar"], "title": "PPAI: Enabling Personalized LLM Agent Interoperability for Collaborative Edge Intelligence", "abstract": "Deploying large language model (LLM) on edge device enables personalized LLM agents for various users. The growing availability of diverse personalized agents presents a unique opportunity for peer-to-peer (P2P) collaboration, wherein each user can delegate tasks beyond the local agent's expertise to remote agents more suited for the specific query. This paper introduces PPAI, the first personalized LLM agent interoperability system, which enables users to collaborate with each other based on agent specialization. However, the ever-changing pool of agents and their interchangeable capacity introduce new challenges when it comes to matching queries to agents and balancing loads, compared with existing P2P systems. Therefore, we propose a scalable query-agent pair scoring mechanism based on prototypes to identify suitable agents within a P2P network with churn. Moreover, we propose a multi-agent interoperability Bayesian game to balance local demand and global efficiency, when changes in remote agent load occur too quickly to be observed. Finally, we implement a prototype of PPAI and demonstrate that it substantially broadens the range of tasks that could be carried out while maintaining load balance. On average, it achieves an accuracy improvement of up to 7.96% across multiple tasks, while reducing latency by 16.34% compared to the baseline.", "authors": ["Zile Wang", "Qianli Liu", "Kaibin Guo", "Haodong Wang", "Jian Lin", "Zicong Hong", "Song Guo"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.18067", "pdf_url": "https://arxiv.org/pdf/2605.18067v1", "arxiv_id": "2605.18067", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1355e0aba3dbd9f6a4388cb534677a790f575edfe2e4b6a1d22ab2e486bbdf18", "sources": ["arxiv", "semantic_scholar"], "title": "Position: A Three-Layer Probabilistic Assume-Guarantee Architecture Is Structurally Required for Safe LLM Agent Deployment", "abstract": "This position paper argues that enforcing LLM agent safety within a single abstraction layer is not merely suboptimal but categorically insufficient for deployed LLM agents -- a structural consequence of how agent execution works, not a contingent limitation of current systems. The three dimensions that jointly constitute safe operation -- semantic intent and policy compliance, environmental validity, and dynamical feasibility -- each depend on a strictly distinct set of information that becomes available at different stages of execution. No single guardrail can certify all three. We argue that the community must respond with a contract-based architecture in which each safety dimension is enforced by an independently certified layer whose probabilistic guarantee satisfies the next layer's assumption. We sketch such an architecture and derive the compositional system-level safety bounds it admits via the chain rule of probability. Three open problems stand between this and a deployable standard: bound estimation from non-i.i.d.\\ traces, graceful degradation of contracts under deployment drift, and extension to multi-agent settings -- the most important unfinished business in LLM agent runtime assurance.", "authors": ["S. Bensalem", "Y. Dong", "M. Franzle", "X. Huang", "J. Kroger", "D. Nickovic", "A. Nouri", "R. Roy", "C. Wu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.18672", "pdf_url": "https://arxiv.org/pdf/2605.18672v1", "arxiv_id": "2605.18672", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "57b034743e4188b6f0183e6e991d6605f88fc1096e1bc3a2f5e36aaf950cc145", "sources": ["arxiv", "semantic_scholar"], "title": "Time to REFLECT: Can We Trust LLM Judges for Evidence-based Research Agents?", "abstract": "Deep research agents increasingly automate complex information-seeking tasks, producing evidence-grounded reports via multi-step reasoning, tool use, and synthesis. Their growing role demands scalable, reliable evaluation, positioning LLM-as-judge as a supervision paradigm for assessing factual accuracy, evidence use, and reasoning quality. Yet the reliability of these judges for deep research agents remains poorly understood, posing a critical meta-evaluation problem: before deploying LLM judges to supervise research agents, we must first evaluate the judges themselves. Existing meta-evaluations fall short in two ways: (1) reliance on coarse, subjective human-preference agreement; (2) focus on instruction-following or verifiable tasks, leaving open-ended agent executions unexplored. To address these gaps, we introduce REFLECT (REliable Fine-grained LLM judge Evaluation via Controlled inTervention), a meta-evaluation benchmark targeting fine-grained failure detection in agentic environments. REFLECT defines a detailed taxonomy of process- and outcome-level failure modes, instantiated by performing controlled and localized interventions on quality-screened agent execution traces. This yields verifiable, comprehensive, and fine-grained instances for validating the judge models. Our experiments show that current LLM judges remain unreliable: even the best-performing models achieve overall accuracies below 55% across reasoning, tool-use, and report-quality failures, with especially poor performance on evidence verification. Together, our taxonomy and findings expose systematic judge limitations, reveal tradeoffs in cost and reliability, and offer actionable guidance for building more reliable evaluation pipelines for deep research agents.", "authors": ["Leyao Wang", "Yanan He", "Peng Chen", "Asaf Yehudai", "Yixin Liu", "Rex Ying", "Michal Shmueli-Scheuer", "Arman Cohan"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-18", "url": "https://arxiv.org/abs/2605.19196", "pdf_url": "https://arxiv.org/pdf/2605.19196v1", "arxiv_id": "2605.19196", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7294407b9c92bd8bcd0c1108288b8c212007d67d15609d091b1578a5f23f23e4", "sources": ["arxiv", "semantic_scholar"], "title": "VerifyMAS: Hypothesis Verification for Failure Attribution in LLM Multi-Agent Systems", "abstract": "Large language model-driven multi-agent systems (LLM-MAS) excel at complex tasks, yet unreliable agents remain a key bottleneck to system-level reliability. Automatic failure attribution is therefore critical, but existing approaches, such as direct prediction of agent-error pairs and agent-first failure attribution, rely on local logs of agents and miss global failures that only manifest over full interaction trajectories, such as cross-step inconsistencies and inter-agent coordination errors. Moreover, directly predicting failures induces a large combinatorial search space, hindering fine-grained attribution. To address these challenges, we propose VerifyMAS, a hypothesis verification framework for agent failure attribution. Instead of directly predicting faulty agents and error types, VerifyMAS formulates and verifies failure hypotheses against full trajectories. This verification-based approach decomposes attribution into trajectory-level error validation and fine-grained agent localization, providing an error-first attribution approach that captures global failure patterns while substantially reducing the search space. We further introduce a hypothesis-based data construction strategy grounded in a structured error taxonomy and fine-tune a specialized LLM verifier model for trajectory-level failure verification and agent attribution. Experiments on Aegis-Bench and Who&When show that VerifyMAS consistently improves diverse backbone models, including open-source Qwen and API-based GPT models, outperforming prior methods without sacrificing inference efficiency for long multi-agent trajectories.", "authors": ["Hezhe Qiao", "Hanghang Tong", "Ee-Peng Lim", "Bing Liu", "Guansong Pang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17467", "pdf_url": "https://arxiv.org/pdf/2605.17467v1", "arxiv_id": "2605.17467", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "7b8e0ba5a4867f9ca24596343dff2174443b65b1115466a5727eca83762168de", "sources": ["arxiv", "semantic_scholar"], "title": "MetaCogAgent: A Metacognitive Multi-Agent LLM Framework with Self-Aware Task Delegation", "abstract": "Multi-agent large language model (LLM) systems have shown promise for solving complex tasks through agent collaboration. However, existing frameworks assign tasks based on predefined roles without considering whether an agent can accurately assess its own competence boundaries, leading to overconfident execution on tasks beyond its expertise. Inspired by metacognition theory from cognitive science, we propose MetaCogAgent, a multi-agent LLM framework where each agent is equipped with a Metacognitive Self-Assessment Unit that evaluates task-capability alignment before execution. The framework introduces three contributions: (1) a self-assessment mechanism that estimates per-task confidence by combining verbalized uncertainty with historical capability profiles; (2) an adaptive delegation protocol that routes low-confidence tasks to better-suited agents through cross-agent evaluation; and (3) a capability boundary learning module that iteratively refines each agent's competence model via cybernetic feedback. Experiments on our constructed MetaCog-Eval benchmark (700 tasks across 5 cognitive dimensions) demonstrate that MetaCogAgent achieves 82.4% task accuracy -- 8.7% above the best routing baseline -- while using 5% fewer API calls than AutoGen and 34% fewer than ensemble voting. Ablation studies confirm that each metacognitive component contributes to overall system performance.", "authors": ["Chenyu Wang", "Yang Shu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17292", "pdf_url": "https://arxiv.org/pdf/2605.17292v1", "arxiv_id": "2605.17292", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6656e88d69f28a5156dbb8e96e7be5ae3cec2d68529b3cb752ede5041fb4ffe4", "sources": ["arxiv", "semantic_scholar"], "title": "Trust No Tool: Evaluating and Defending LLM Agents under Untrusted Tool Feedback", "abstract": "Tool-using LLM agents increasingly rely on external tools to make consequential decisions, yet most existing agent-security benchmarks and defenses implicitly assume that tool feedback is trustworthy once a tool has been selected. We study a different failure mode, cognitive poisoning, in which a malicious tool behaves plausibly during exploration, accumulates trust through benign-looking feedback, and becomes harmful only when hidden state conditions align with the final executable action. To study this setting, we construct TRUST-Bench, a task-conditioned benchmark of 1,970 hidden-trigger tool-compromise episodes with matched safe controls, introduce an asymmetric penalty metric, GuardedJoint, to better reflect real deployment risk, and present VISTA-Guard, a backbone-agnostic framework for final-action risk scoring. The core idea is to abstract multi-step tool interaction into structured environment variables that encode trust-formation dynamics and then score the risk of the final executable action from this trajectory-conditioned representation. Experiments show that prompt-centric heuristics, scalarized features, and zero-shot judges fail in this regime, whereas trajectory-aware final-action scoring yields strong in-domain discrimination and remains effective under balanced out-of-distribution transfer. Under GuardedJoint, VISTA-Guard reaches $84.2$ in-domain and $56.9$ on balanced out-of-distribution evaluation, while methods that optimize only one side of the safety--utility tradeoff collapse to zero. These findings support a broader view of agent security in black-box tool ecosystems: the decisive defense target is not local prompt text or tool descriptors alone, but the way trust is formed across the interaction trajectory and committed through the final action.", "authors": ["Lecheng Yan", "Ruizhe Li", "Xicheng Han", "Wenxi Li", "Binwu Wang", "Longyue Wang", "Chenyang Lyu", "Guanhua Chen"], "categories": ["cs.CR", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17453", "pdf_url": "https://arxiv.org/pdf/2605.17453v1", "arxiv_id": "2605.17453", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "8e529ead37b8ebc3c3d4039f1b84e721f6bac0019bfd481a14fb32e07d35f8fc", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains", "abstract": "Large language model (LLM)-based multi-agent systems have shown strong potential for complex reasoning by coordinating specialized agents through structured communication. However, existing topology-evolution methods typically construct or optimize a collaboration topology for each query from scratch, leading to substantial online search overhead, high inference-time token consumption, and limited scalability in multi-domain settings. We propose TopoPrior, a framework for learning transferable topology priors for multi-agent LLM collaboration across domains. Rather than repeatedly searching for effective collaboration structures online, TopoPrior learns reusable topology priors from reference collaboration graphs collected offline from multiple domains and uses them to generate query-conditioned initial collaboration graphs for downstream refinement. By shifting part of topology search from per-query online optimization to offline prior learning, TopoPrior amortizes search cost while remaining compatible with existing topology-evolution backbones. Technically, TopoPrior contains two key components. First, a transferable topology prior learning module employs a conditional variational graph framework to capture reusable structural regularities across domains in a latent space. Second, a query-conditioned latent adaptation module introduces adversarial alignment to reduce unnecessary domain discrepancy while preserving query-relevant structural variation. Experiments on multi-domain reasoning benchmarks show that TopoPrior consistently improves several heterogeneous topology-evolution backbones while reducing online inference-time token usage, with only modest additional trainable parameters. These results suggest that transferable topology initialization is an effective and lightweight mechanism for improving the efficiency of multi-agent LLM collaboration across domains.", "authors": ["Taolin Zhang", "Zijie Zhou", "Jiuheng Wan", "Tingyuan Hu", "Chengyu Wang", "Xiaofeng He", "Richang Hong"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17359", "pdf_url": "https://arxiv.org/pdf/2605.17359v1", "arxiv_id": "2605.17359", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5549d89545f6aa150768a9f0cc4f081c5451a3c0d78117c094c2927309e753e2", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Bazaar: Enabling Economic Alignment in Multi-Agent Marketplaces", "abstract": "The deployment of Large Language Models (LLMs) as autonomous economic agents introduces systemic risks that extend beyond individual capability failures. As agents transition to directly interacting with marketplaces, their collective behavior can amplify volatility and mask deception at scale. We introduce the Agent Bazaar, a multi-agent simulation framework for evaluating Economic Alignment, the capacity of agentic systems to preserve market stability and integrity. We identify two failure modes: (1) Algorithmic Instability in a B2C market (\"The Crash\"), where firms amplify price volatility until the market collapses, and (2) Sybil Deception in a C2C market (\"The Lemon Market\"), where a single deceptive agent controlling multiple coordinated seller identities floods the market with fraudulent listings, eroding trust and consumer welfare. We evaluate frontier and open-weight models across both scenarios and find that models largely fail to self-regulate, with failure severity varying by model rather than by size. We propose economically aligned harnesses, Stabilizing Firms and Skeptical Guardians, that improve outcomes but remain fragile under harder market conditions. To close this gap, we train agents with REINFORCE++ using an adaptive curriculum, producing a 9B model that outperforms all evaluated frontier and open-weight models. We propose the Economic Alignment Score (EAS), a 4-component scalar metric aggregating stability, integrity, welfare, and profitability, enabling direct cross-model comparison. Our results show that economic alignment is orthogonal to general capability and can be directly trained with targeted RL.", "authors": ["Seth Karten", "Cameron Crow", "Chi Jin"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17698", "pdf_url": "https://arxiv.org/pdf/2605.17698v1", "arxiv_id": "2605.17698", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "86030df69f07adcaa0f8cc28c3d3a5dbcde90b9e3e495f30e2c4172a289c4fe9", "sources": ["arxiv", "semantic_scholar"], "title": "Taming \"Zombie'' Agents: A Markov State-Aware Framework for Resilient Multi-Agent Evolution", "abstract": "Recent advancements in LLM-based multi-agent systems have demonstrated remarkable collaborative capabilities across complex tasks. To improve overall efficiency, existing methods often rely on aggressive graph evolution among agents (e.g., node or edge pruning), which risks prematurely discarding valuable agents due to transient issues such as hallucinations or temporary knowledge gaps. However, such hard pruning overlooks the potential for ``zombie'' agents to recover and contribute in subsequent discussion rounds. In this paper, we propose AgentRevive, a Markov state-aware framework for resilient multi-agent evolution. Our approach dynamically manages agent collaboration through soft state transitions, implemented via two key components: (1) State-Aware Policy Learning: Agent states are divided into ``Active'', ``Standby'', and ``Terminated'' states, selectively propagating messages based on agent memory. The policy employs a risk estimator to optimize agent state transitions by assessing hallucination risk, minimizing the influence of unreliable nodes while safeguarding valuable ones. (2) State-Aware Edge Optimization: Subgraph edges are pruned according to states learned from the policy, permanently removing ``Terminated'' nodes and retaining ``Standby'' nodes for subsequent rounds to assess their potential future contributions. Extensive experiments on general reasoning, domain-specific, and hallucination challenge tasks show that our method consistently outperforms strong baselines and significantly reduces token consumption through state-aware agent scheduling.", "authors": ["Taolin Zhang", "Pukun Zhao", "Qizhou Chen", "Jiuheng Wan", "Chen Chen", "Xiaofeng He", "Chengyu Wang", "Richang Hong"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17348", "pdf_url": "https://arxiv.org/pdf/2605.17348v1", "arxiv_id": "2605.17348", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "682cfebab4f97c339cacc339b6f0475bcd7ae0c72acdbfae410a7b4164ce7175", "sources": ["arxiv", "semantic_scholar"], "title": "RooAgent: An LLM Agent for Root-Based High Energy Physics Analysis", "abstract": "We present RooAgent as a natural-language interface for Root-based high energy physics data analysis. The package provides physics analysis functions as tools that an LLM agent invokes in response to plain-language prompts. Two operating modes are supported: a LangGraph-based agent compatible with OpenAI's GPT-4.1 via GitHub Copilot and with DeepSeek-V3 via Ollama, and a Model Context Protocol server for use with the Anthropic Claude CLI (Sonnet~4.6). In both modes the analysis logic is implemented in PyRoot and the LLM selects tools and supplies the required arguments. The package supports histogram inspection, event selection, visualisation of kinematic distributions, fitting, and significance estimation, among other tasks. We illustrate RooAgent with tests based on Monte Carlo simulations of $pp\\to ZH$ ($Z\\to\\ell^+\\ell^-$, $H\\to b\\bar{b}$), a multi-task signal-background workflow, a toy statistical analysis, and an application to ATLAS open data for $H\\to ZZ^*\\to 4\\ell$. The package is available on PyPI and the source code is hosted at https://github.com/amanmdesai/RooAgent.", "authors": ["Aman Desai"], "categories": ["hep-ph"], "fields_of_study": ["Physics"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17318", "pdf_url": "https://arxiv.org/pdf/2605.17318v2", "arxiv_id": "2605.17318", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/amanmdesai/RooAgent", "venue": null, "quality_score": 0.65}
{"id": "42b07c895e4407830b29fbbde66de3c2bbdab59b321ee950ab51fc344358d89d", "sources": ["arxiv", "semantic_scholar"], "title": "AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering", "abstract": "Despite substantial advances in large language models (LLMs), generating factually consistent responses for knowledge-intensive question answering remains challenging. These difficulties are primarily due to hallucinations and the limitations of LLMs in bridging long-tail knowledge gaps. To address this, we propose AMATA, an Adaptive Multi-Agent Trajectory Alignment framework that dynamically integrates external knowledge to improve response interpretability and factual grounding. Our architecture leverages six specialized agents that collaboratively perform structured actions for complex question reasoning. We formalize multi-agent collaboration with external tools as a trajectory preference alignment problem, incorporating question-aware agent customization and inter-agent preference harmonization. AMATA introduces two principal innovations: (1) Intra-Trajectory Preference Learning, which learns objective-oriented preferences to prioritize critical agents, and (2) Inter-Agent Dependency Learning, which captures cross-agent tool dependencies through a novel dependency-aware direct preference optimization technique. Empirical results show that AMATA consistently outperforms baseline approaches, knowledge-augmented frameworks, and LLM-based trajectory systems on five established knowledge-intensive QA benchmarks. Further analysis demonstrates the efficiency of our method in reducing token consumption.", "authors": ["Taolin Zhang", "Dongyang Li", "Chen Chen", "Qizhou Chen", "Jiuheng Wan", "Xiaofeng He", "Chengyu Wang", "Richang Hong"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-17", "url": "https://arxiv.org/abs/2605.17352", "pdf_url": "https://arxiv.org/pdf/2605.17352v1", "arxiv_id": "2605.17352", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "8c1300afc52492790790c1c2cc4f0302d280a9b1a3f051775269ae35c2ca5c47", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Paradigm Agent Interaction in Practice:A Systematic Analysis of Generator-Evaluator, ReAct Loop,and Adversarial Evaluation in the buddyMe Framework", "abstract": "The rapid evolution of Large Language Model (LLM) agents has produced diverse interaction paradigms, yet few production systems integrate multiple paradigms within a unified architecture. This paper presents a systematic analysis of three principal agent interaction paradigms, including Multi-Agent Orchestration (Generator-Evaluator), ReAct Tool-Use Loops, and Memory-Augmented Interaction, as implemented in buddyMe, an open-source multi-model agent programming framework. We formalize a five-stage processing pipeline: Requirement Pre-Review -> Task Decomposition -> ReAct Execution -> Real-Execution Verification -> Adversarial Evaluation Discussion, and establish a six-dimensional evaluation schema with weighted scoring. Through four empirical case studies drawn from real-world deployment logs covering museum guide generation, scheduled weather tasks, and comprehensive tour planning, we draw three key conclusions. First, Generator-Evaluator pre-review detects requirement omissions in 20 percent of complex tasks, with 80 percent tasks passing initial inspection. Second, the ReAct loop ensures stable subtask execution but leads to around 30 percent redundant tool invocations. Third, adversarial Evaluator-Defender discussions reach consensus within 2-3 rounds for nearly 70 percent of scenarios, functioning mainly for content refinement rather than logical reversal. We additionally provide three Mermaid-based architectural diagrams and conduct cross-paradigm comparisons with CrewAI, AutoGen, LangGraph, MemGPT and A-Mem across six system dimensions. The research outcomes offer practical design guidelines for constructing stable and reliable multi-paradigm agent systems.", "authors": ["Xiaohua Wang", "Chao Han", "Kai Yu", "XiaoLiang Xu", "Liang Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-16", "url": "https://arxiv.org/abs/2605.16821", "pdf_url": "https://arxiv.org/pdf/2605.16821v1", "arxiv_id": "2605.16821", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "e17ebd701cb2661353c7939c1412ad26fcd7243c47aee4533682ab79ee390a92", "sources": ["arxiv", "semantic_scholar"], "title": "S-Bus: Automatic Read-Set Reconstruction for Multi-Agent LLM State Coordination", "abstract": "We address concurrency control for LLM agents sharing mutable state over HTTP, where agents cannot be modified to declare read sets. S-Bus is an HTTP middleware whose central mechanism, a server-side DeliveryLog, reconstructs each agent's read set at commit time from observed HTTP GET traffic. The consistency property it provides -- Observable-Read Isolation (ORI), a partial causal consistency over the HTTP-observable read projection -- prevents Structural Race Conditions in dedicated-shard topologies. Three contributions. (C1) DeliveryLog mechanism with three-tier mechanised evidence: TLAPS proves ReadSetSoundness and ORICommitSafety (modulo one typing axiom); exhaustive TLC at N=3 explores 20,763,484 states with zero violations; Dafny discharges 9 inductive lemmas. (C2) Empirical safety parity against PostgreSQL 17 SERIALIZABLE and Redis 7 WATCH/MULTI: zero Type-I corruptions across 884,110 commit attempts (427,308 under active contention). (C3) ORI is semantically neutral in dedicated-shard workloads but harmful in single-shard collaborative writing because preservation propagates concurrent contradictions. v2 update: the PH-3 LLM judge is now independently validated against a human annotator (Zahid Hussain, Mindgigs Peshawar) on 400 (step, shard) pairs at strict kappa=0.93 (n=93, 96.8% raw agreement). Inter-LLM-judge agreement is kappa=0.46 (boundary variance). Agent self-reports over-claim shard usage by 32% (LLM judge) to 49% (human annotator). The SJ-v4 semantic-quality rubric remains single-judge LLM-only. Source code, formal proofs, harness, annotation data: https://github.com/sajjadanwar0/sbus", "authors": ["Sajjad Khan"], "categories": ["cs.LG", "cs.AI", "cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-16", "url": "https://arxiv.org/abs/2605.17076", "pdf_url": "https://arxiv.org/pdf/2605.17076v2", "arxiv_id": "2605.17076", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/sajjadanwar0/sbus", "venue": null, "quality_score": 0.65}
{"id": "77880cdb6661616827844a3339e2a112023dc9fe2eec013b5e080da9dc1984fe", "sources": ["arxiv", "semantic_scholar"], "title": "NeuroMAS: Multi-Agent Systems as Neural Networks with Joint Reinforcement Learning", "abstract": "Multi-agent language systems are often built as hand-designed workflows, where agents are assigned semantic roles and communication protocols are specified in advance. We propose NeuroMAS, a method that first treats a multi-agent language system as a trainable and scalable neural-network-like architecture with LLM agents as nodes and intermediate textual signals as edges. In NeuroMAS, agent nodes are role-free but structure-aware: the topology only determines how information can flow in general, while reinforcement learning training determines how nodes communicate, specialize, and coordinate. This formulation shifts multi-agent design from workflow engineering toward architecture design, where depth, width, connectivity, and growth protocol become scalable sources of capability. Further, we provide a theoretical perspective showing why such modular textual computation is more parameter-efficient when tasks admit hierarchical decompositions. Experiments show that NeuroMAS improves significantly over both inference-time and trained multi-agent baselines. We further find that organizational scaling is path-dependent: larger systems can be challenging to train from scratch, but become feasible when grown progressively from smaller trained systems. These results suggest that learned neural multi-agent systems are a promising scaling axis for LLMs.", "authors": ["Haoran Lu", "Luyang Fang", "Wenxuan Zhong", "Ping Ma"], "categories": ["cs.AI", "cs.MA", "stat.ME", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-05-16", "url": "https://arxiv.org/abs/2605.16757", "pdf_url": "https://arxiv.org/pdf/2605.16757v1", "arxiv_id": "2605.16757", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9cc54f12e19ad8542c56882b77994e7ca3fddbd740c697f1042153e4fc171235", "sources": ["arxiv", "semantic_scholar"], "title": "BioXArena: Benchmarking LLM Agents on Multi-Modal Biomedical Machine Learning Tasks", "abstract": "Large language model (LLM) agents are increasingly capable of automating components of machine learning development, yet existing biomedical benchmarks mainly focus on question answering, reasoning, and tool usage, or evaluate only narrow aspects of biomedical ML coding. We present BioXArena, a biomedical machine learning benchmark designed to evaluate whether agents can generate task-specific model training pipelines for heterogeneous and multi-modal biomedical datasets. BioXArena contains 76 end-to-end tasks across 9 domains, including sequence modeling, single-cell analysis, structural biology, network biology, chemical biology, perturbation dynamics, phenotype-disease modeling, biomedical imaging, and text-integrated learning. Each task is curated from primary biomedical sources into a unified evaluation framework with hidden labels, held-out graders, and biology-aware metrics normalized to a 0 to 1 scale. Agents are required to write executable code, train predictive models, and generate submissions for private test samples. Most tasks involve multiple input modalities, including tabular data, images, natural language, molecular sequences, omics matrices, and protein structures. We evaluate 11 agent configurations in a standardized 2-hour single-GPU environment. MLEvolve with Gemini-3.1-Pro achieves the highest average score of 0.666, followed by GPT-5.4 with 0.636, while no single agent consistently dominates across all domains. We additionally perform extensive ablation studies, robustness evaluations, scaling analyses, cost analyses, and failure-mode investigations to better understand how model backbones, agent scaffolds, inference budgets, and biomedical domains influence BioML coding performance. We will publicly release all benchmark tasks, graders, execution runners, leaderboard results, and agent trajectories.", "authors": ["Loka Li", "Duzhen Zhang", "Xingbo Du", "Leonard Song", "Zixiao Wang", "Assanali Aukenov", "Noel Thomas", "Shakhnazar Sailaukan", "Yonghan Yang", "Feilong Chen", "Jiahua Dong", "Kun Zhang", "Bin Zhang", "Le Song"], "categories": ["cs.CE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-15", "url": "https://arxiv.org/abs/2605.15766", "pdf_url": "https://arxiv.org/pdf/2605.15766v1", "arxiv_id": "2605.15766", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1acdb65f7a03f600cf2bcb3dee403ba62d003c11e4ac08400f4c507124b1f306", "sources": ["arxiv", "semantic_scholar"], "title": "Cattle Trade: A Multi-Agent Benchmark for LLM Bluffing, Bidding, and Bargaining", "abstract": "We introduce \\textsc{Cattle Trade, a multi-agent benchmark for evaluating large language models (LLMs) as agents in strategic reasoning under imperfect information, adversarial interaction, and resource constraints. The benchmark combines auctions, hidden-offer trade challenges (TCs), bargaining, bluffing, opponent modeling, and resource allocation within a single long-horizon game lasting 50--60 turns. Unlike prior agent benchmarks that test these abilities in isolation, \\textsc{Cattle Trade} evaluates whether agents integrate them across a competitive, multi-agent economic game with conflicting incentives. The benchmark logs every bid, TC offer, counteroffer, and card selection, enabling behavioural analysis beyond final scores or win rates. We evaluate seven cost-efficient language models and three deterministic code agents across 242 games. Strategic coherence, in particular spending efficiency, resource discipline, and phase-adaptive bidding, is associated with rank more strongly than spending volume or any single subskill. Two heuristic code agents outperform most tested LLMs, and behavioural traces surface recurring LLM failure modes including overbidding, self-bidding, bankrupt TC initiation, and weak opponent-state adaptation. Evaluating agentic competence requires benchmarks that test the joint deployment of multiple capabilities in multi-agent environments with conflicting incentives, uncertainty, and economic dynamics.", "authors": ["Robert Müller", "Clemens Müller"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14537", "pdf_url": "https://arxiv.org/pdf/2605.14537v1", "arxiv_id": "2605.14537", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "39fc5b80727965d7820af2d4ed4ef7eb7d21beeef498519355275c2dc873662e", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems", "abstract": "LLM-based autonomous agents have demonstrated strong capabilities in reasoning, planning, and tool use, yet remain limited when tasks require sustained coordination across roles, tools, and environments. Multi-agent systems address this through structured collaboration among specialized agents, but tighter coordination also amplifies a less explored risk: errors can propagate across agents and interaction rounds, producing failures that are difficult to diagnose and rarely translate into structural self-improvement. Existing surveys cover individual agent capabilities, multi-agent collaboration, or agent self-evolution separately, leaving the causal dependencies among them unexamined. This survey provides a unified review organized around four causally linked stages, which we term the LIFE progression: Lay the capability foundation, Integrate agents through collaboration, Find faults through attribution, and Evolve through autonomous self-improvement. For each stage, we provide systematic taxonomies and formally characterize the dependencies between adjacent stages, revealing how each stage both depends on and constrains the next. Beyond synthesizing existing work, we identify open challenges at stage boundaries and propose a cross-stage research agenda for closed-loop multi-agent systems capable of continuously diagnosing failures, reorganizing structures, and refining agent behaviors, extending current coordination frameworks toward more self-organizing forms of collective intelligence. By bridging these previously fragmented research threads, this survey aims to offer both a systematic reference and a conceptual roadmap toward autonomous, self-improving multi-agent intelligence.", "authors": ["Shihao Qi", "Jie Ma", "Rui Xing", "Wei Guo", "Xiao Huang", "Zhitao Gao", "Jianhao Deng", "Jun Liu", "Lingling Zhang", "Bifan Wei", "Boqian Yang", "Pinghui Wang", "Jianwen Sun", "Jing Tao", "Yaqiang Wu", "Hui Liu", "Yu Yao", "Tongliang Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14892", "pdf_url": "https://arxiv.org/pdf/2605.14892v2", "arxiv_id": "2605.14892", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a72fe7030dd79e5175761308fd3ad74a1dcf2de4decccbbda5112fe516d60e48", "sources": ["arxiv", "semantic_scholar"], "title": "Concurrency without Model Changes: Future-based Asynchronous Function Calling for LLMs", "abstract": "Function calling, also known as tool use, is a core capability of modern LLM agents but is typically constrained by synchronous execution semantics. Under these semantics, LLM decoding is blocked until each function call completes, resulting in increasing end-to-end latency. In this work, we introduce AsyncFC, a pure execution-layer framework that decouples LLM decoding from function execution, enabling overlap between model decoding and function execution as well as inter-function parallelism when dependencies permit. AsyncFC layers over existing models and unmodified function implementations, requiring no fine-tuning or changes to the standard synchronous function-calling protocol. Across standard function-calling benchmarks and adapted software engineering benchmarks, AsyncFC significantly reduces end-to-end task completion time while preserving task accuracy. Furthermore, these results reveal that LLMs possess a native capability to reason over symbolic futures that represent unresolved execution results, enabling an asynchronous paradigm for model-tool interaction.", "authors": ["Guangyu Feng", "Huanzhi Mao", "Prabal Dutta", "Joseph E. Gonzalez"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.15077", "pdf_url": "https://arxiv.org/pdf/2605.15077v1", "arxiv_id": "2605.15077", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "bd2fc440df137d37625dedc149737ba12ef155263197653bf40604ce2af6999a", "sources": ["arxiv", "semantic_scholar"], "title": "Making OpenAPI Documentation Agent-Ready: Detecting Documentation and REST Smells with a Multi-Agent LLM System", "abstract": "The growing adoption of AI agents and the Model Context Protocol (MCP) has motivated organizations to expose existing REST APIs as agent-consumable tools. In our industrial context, this initiative targeted an ecosystem of 16 production APIs comprising approximately 600 endpoints. Although these APIs were stable and widely used within a microservice architecture, early proof-of-concept experiments revealed systematic failures in task planning, tool selection, and payload construction when accessed through MCP-based agents. Rather than attributing these failures to model limitations alone, we conducted an ecosystem-scale empirical assessment of the underlying OpenAPI documentation. We developed Hermes, a multi-agent LLM-based system that detects documentation and REST-related smells at the endpoint level and generates explainable diagnostic reports. The large-scale evaluation identified 2,450 smells across 600 endpoints, with deficiencies present in all analyzed operations. Practitioner validation confirmed high agreement with the detected issues while also revealing contextual trade-offs in remediation decisions. The findings suggested that structural validity within microservice environments does not guarantee semantic readiness for agent-based consumption. Based on this evidence, the organization revised its adoption strategy, prioritizing selective endpoint adaptation, redefining documentation standards, and integrating automated documentation assessment into API governance workflows. This case illustrates how systematic artifact-level evaluation can function as a strategic decision-support mechanism, reducing technological risk and guiding evidence-based AI adoption in industrial software ecosystems.", "authors": ["Rayfran Rocha Lima", "Davi G. Assunção Pinheiro", "Thiago Medeiros de Menezes"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14312", "pdf_url": "https://arxiv.org/pdf/2605.14312v1", "arxiv_id": "2605.14312", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "53f486dd8cac88c931604833d0c7173ddd9f57d2f1763eddf4139a42647cbed0", "sources": ["arxiv", "semantic_scholar"], "title": "From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents", "abstract": "Voice agents increasingly require reliable tool use from speech, whereas prominent tool-calling benchmarks remain text-based. We study whether verified text benchmarks can be converted into controlled audio-based tool calling evaluations without re-annotating the tool schema and gold labels. Our dataset-agnostic framework uses text-to-speech, speaker variation, and environmental noise to create paired text-audio instances while preserving the original dataset annotations. Based on extensive evaluation of 7 omni-modal models on audio-converted versions of Confetti and When2Call, our framework demonstrates that the performance is strongly model- and task-dependent: Gemini-3.1-Flash-Live obtains the highest Confetti score (70.4), whereas GPT-Realtime-1.5 performs best on When2Call (71.9). On Confetti, the text-to-voice gap ranges from 1.8 points for Qwen3-Omni to 4.8 points for GPT-Realtime-1.5. A targeted analysis of failure cases demonstrates that degradations most often reflect misunderstandings of argument values in the speech. Considering real-world deployment scenarios, we further report text-only results, an ambiguity-based reformulation stress test, and a reference-free LLM-as-judge protocol validated against human preferences. Notably, we find that open-source Qwen3 judges with at least 8B parameters exceed 80% agreement with proprietary judges, supporting privacy-preserving evaluation. Overall, our framework provides a verifiable and reproducible first-stage diagnostic that complements purpose-built audio corpora.", "authors": ["Md Tahmid Rahman Laskar", "Xue-Yong Fu", "Seyyed Saeed Sarfjoo", "Quinten McNamara", "Jonas Robertson", "Shashi Bhushan TN"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.15104", "pdf_url": "https://arxiv.org/pdf/2605.15104v2", "arxiv_id": "2605.15104", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "2d9e60909ce1e6a4d179789e5bc87074612d4553d2fed2ea6a2da13f1847efdb", "sources": ["arxiv", "semantic_scholar"], "title": "Latency-Quality Routing for Functionally Equivalent Tools in LLM Agents", "abstract": "Tool-augmented LLM agents increasingly access the same tool type through multiple functionally equivalent providers, such as web-search APIs, retrievers, or LLM backends exposed behind a shared interface. This creates a provider-routing problem under runtime load: the router must choose among providers that differ in latency, reliability, and answer quality, often without gold labels at deployment time. We introduce LQM-ContextRoute, a contextual bandit router for same-function tool providers. Its key design is latency-quality matching: instead of letting low latency offset poor answers in an additive reward, the router ranks providers by expected answer quality per service cycle. It combines this capacity-aware score with query-specific quality estimation and LLM-as-judge feedback, allowing it to adapt online to both load changes and provider-quality differences. On the main web-search load benchmark, LQM-ContextRoute improves F1 by +2.18 pp over SW-UCB while staying on the latency-quality frontier. In a high-heterogeneity StrategyQA setting, LQM-ContextRoute avoids additive-reward collapse and improves accuracy by up to +18 pp over SW-UCB; on heterogeneous retriever pools, it improves NDCG by +2.91--+3.22 pp over SW-UCB. These results show that same-function tool routing benefits from treating latency as service capacity, especially when runtime pressure and provider-quality heterogeneity coexist.", "authors": ["Kexin Chu", "Dawei Xiang", "Wei Zhang"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14241", "pdf_url": "https://arxiv.org/pdf/2605.14241v2", "arxiv_id": "2605.14241", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "40038dc22a839241eea9fc0e1f4a884c0d4aeacd0d35034e7aaee8b4d644c83d", "sources": ["arxiv", "semantic_scholar"], "title": "GroupMemBench: Benchmarking LLM Agent Memory in Multi-Party Conversations", "abstract": "Large Language Model (LLM) agents increasingly serve as personal assistants and workplace collaborators, where their utility depends on memory systems that extract, retrieve, and apply information across long-running conversations. However, both existing memory systems and benchmarks are built around the dyadic, single-user setup, even though real deployments routinely span groups and channels with multiple users interacting with the agent and with each other. This mismatch leaves three properties of group memory unmeasured: (i) group dynamics that go beyond concatenated one-on-one chats, (ii) speaker-grounded belief tracking, where the per-user memory modeling is needed, and (iii) audience-adapted language, where Theory-of-Mind shifts produce role-specific vocabulary. We introduce GroupMemBench, a benchmark that exposes all three. A graph-grounded synthesis pipeline produces multi-party conversations with controllable reply structure and conditions each message on per-user personas and target audiences. An adversarial query pipeline then binds every question to a specific asker across six categories, spanning multi-hop reasoning, knowledge update, term ambiguity, user-implicit reasoning, temporal reasoning, and abstention, and iteratively searches challenging, realistic queries that reflect comprehensive memory capability. Benchmarking leading memory systems exposes a sharp collapse: the strongest one reaches only 46.0% average accuracy, with knowledge update at 27.1% and term ambiguity at 37.7%, while a simple BM25 baseline matches or exceeds most agent memory systems. This indicates current memory ingestion erases the structural and lexical features group memory depends on, leaving multi-user memory far from solved.", "authors": ["Jingbo Yang", "Kwei-Herng Lai", "Xiaowen Wang", "Shiyu Chang", "Yaar Harari", "Evgeniy Gabrilovich"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14498", "pdf_url": "https://arxiv.org/pdf/2605.14498v2", "arxiv_id": "2605.14498", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a71c0b15a0e23ab9070dbb9610ddc06c438e402963bafafb7422c98ff16cc147", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic AI Ecosystems in Higher Education: A Perspective on AI Agents to Emerging Inclusive, Agentic Multi-Agent AI Framework for Learning, Teaching and Institutional Intelligence", "abstract": "Integration of artificial intelligent (AI) agents in higher education is transforming teaching, learning and administrative processes. Although existing AI agents effectively support individual tasks, their implementation remains fragmented and inefficient for handling the complexity of educational institutions. This highlights a significant research gap: the lack of integrated eco-system-level agentic multi-agent AI platform capable of coordinated planning, reasoning, and adaptive decision-making across multiple educational functions. This paper presents a forward-looking perspective on agentic multi-agent AI platform in higher education, consisting interconnected autonomous, goal driven agents that support learning, teaching, and institutional operations. It addresses timely and critical questions: Can agentic AI represent the next generation of intelligent systems in tertiary education? Can they collectively support seamless coordinated operations across teaching, learning and administrative support? To what extent can such systems foster inclusive and equitable learning for diverse learners with special educational needs? To ground this perspective, a thematic analysis of existing literature identifies four dominant themes: task-specific fragmented AI tools, the transition from single-agent to multi-agent systems, limited cross-functional integration, and insufficient focus on inclusivity and accessibility. Findings reveal a clear gap between current AI implementations and the needs of holistic, learner-centered educational ecosystem. The paper synthesizes challenges and outlines future research directions for scalable human-aligned, and inclusive agentic AI platform. The significant contribution is the incorporation of inclusive learning perspectives, highlighting how coordinated agentic multi-agent platform can support diverse learners through adaptive, multimodal interventions.", "authors": ["Vidya K Sudarshan", "Anushka Sisodia", "Reshma A Ramachandra", "Sia Batra", "Josephine Chong Leng Leng"], "categories": ["cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-14", "url": "https://arxiv.org/abs/2605.14266", "pdf_url": "https://arxiv.org/pdf/2605.14266v1", "arxiv_id": "2605.14266", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "70a89ae4fff390538074a3884ab586529ef725c8300b2179cdbef8d8700f97c2", "sources": ["arxiv", "semantic_scholar"], "title": "Speculative Interaction Agents: Building Real-Time Agents with Asynchronous I/O and Speculative Tool Calling", "abstract": "There is a growing demand for agentic AI technologies for a range of downstream applications like customer service and personal assistants. For applications where the agent needs to interact with a person, real-time low-latency responsiveness is required; for example, with voice-controlled applications, under 1 second of latency is typically required for the interaction to feel seamless. However, if we want the LLM to reason and execute an agentic workflow with tool calling, this can add several seconds or more of latency, which is prohibitive for real-time latency-sensitive applications. In our work, we propose Speculative Interaction Agents to enable real-time interaction even for agents with complex multi-turn tool calling. We propose Asynchronous I/O, which decouples the core agent reason-and-act thread from waiting for additional information from either the user or environment, thereby allowing for overlapping agentic processing while waiting on external delays. We also propose Speculative Tool Calling as a method to manage task execution when the agent is still unsure if it has received the full information or if additional user information may later be provided. For strong cloud models, our method can be applied out-of-the-box to existing real-time cloud APIs, providing 1.3-1.7$\\times$ speedups with minor accuracy loss. To enable real-time interaction with small edge-scale models, we also present a clock-based training methodology that adapts the model to handle streaming inputs and asynchronous responses, and demonstrate a synthetic data generation strategy for SFT. Altogether, this approach provides 1.6-2.2$\\times$ speedups with the Qwen2.5-3B-Instruct and Llama-3.2-3B-Instruct models across multiple tool calling benchmarks.", "authors": ["Coleman Hooper", "Minwoo Kang", "Suhong Moon", "Nicholas Lee", "Eric Wen", "John Wawrzynek", "Michael W. Mahoney", "Yakun Sophia Shao", "Amir Gholami", "Kurt Keutzer"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13360", "pdf_url": "https://arxiv.org/pdf/2605.13360v2", "arxiv_id": "2605.13360", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "271a062541240609a96c3c31a77d6c3f8cfcf8d59cc7376dabf2253de902e3a8", "sources": ["arxiv", "semantic_scholar"], "title": "Hierarchical Attacks for Multi-Modal Multi-Agent Reasoning", "abstract": "Multi-modal multi-agent systems (MM-MAS) have gained increasing attention for their capacity to enable complex reasoning and coordination across diverse modalities. As these systems continue to expand in scale and functionality, investigating their potential vulnerabilities has become increasingly important. However, existing studies on adversarial attacks in multi-agent systems primarily focus on isolated agents or unimodal settings, leaving the vulnerabilities of MM-MAS largely underexplored. To bridge this gap, we introduce HAM$^{3}$, a Hierarchical Attack framework for multi-modal multi-agent systems that decomposes attacks into three interconnected layers. Specifically, at the perception layer, HAM$^{3}$ mounts attacks by perturbing visual inputs, textual inputs, and their fused visual-textual representations. At the communication layer, it performs communication-level attacks that corrupt message content and interaction topology, such as manipulating shared context or communication links to distort collective information flow. At the reasoning layer, it conducts reasoning-level attacks that interfere with each agent's cognitive pipeline, biasing reasoning trajectories and ultimately compromising final decisions. We evaluate HAM$^{3}$ on the GQA benchmark through multi-agent systems built on distinct reasoning paradigms including ReAct, Plan-and-Solve, and Reflexion. Experiments demonstrate that our framework achieves an Attack Success Rate of up to 78.3%, with reasoning-layer attacks being the most effective. More than half of the successful attacks lead multiple agents to produce consistent errors. These findings offer valuable insights for building more robust and interpretable multi-agent intelligence.", "authors": ["Hao Zhou", "Tiru Wu", "Yan Jiang", "Wanqi Zhou", "Junxing Hu", "Ai Han"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13213", "pdf_url": "https://arxiv.org/pdf/2605.13213v1", "arxiv_id": "2605.13213", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026", "quality_score": 0.55}
{"id": "576236b235ac67e87fabfc03c3974b67e62badf3a38a122e9efabec2b73d9a51", "sources": ["arxiv", "semantic_scholar"], "title": "MARLIN: Multi-Agent Game-Theoretic Reinforcement Learning for Sustainable LLM Inference in Cloud Datacenters", "abstract": "Large Language Models (LLMs) have become increasingly prevalent in cloud-based platforms, propelled by the introduction of AI-based consumer and enterprise services. LLM inference requests in particular account for up to 90% of total LLM lifecycle energy use, dwarfing training energy costs. The rising volume of LLM inference requests is increasing environmental footprints, particularly carbon emissions and water consumption. To improve sustainability for LLM inference serving in cloud datacenter environments, we propose a novel multi-agent game-theoretic reinforcement learning framework called MARLIN to co-optimize time-to-first token (TTFT), carbon emissions, water usage, and energy costs associated with LLM inference. MARLIN demonstrates a reduction of at least 18% in TTFT, 33% in carbon emissions, 43% in water usage, and 11% in energy costs compared to state-of-the-art LLM inference management frameworks.", "authors": ["H. Moore", "S. Qi", "D. Milojicic", "C. Bash", "S. Pasricha"], "categories": ["cs.DC", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13496", "pdf_url": "https://arxiv.org/pdf/2605.13496v1", "arxiv_id": "2605.13496", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "deacc5441d8c0e0106e68ae992cca8c4728f848fa3f51aae33bce8688394458c", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforced Collaboration in Multi-Agent Flow Networks", "abstract": "Multi-agent systems provide a powerful way to extend large language models (LLMs) by decomposing a complex task into specialized subtasks handled by different agents. However, their performance is often hindered by error propagation, arising from suboptimal workflow design or inaccurate agent outputs, which can propagate through the agent collaboration process and degrade final results. To address the challenges, we present MANGO (Multi-Agent Network Gradient Optimization), a data-driven framework that organizes and refines agent collaboration via a flow network constructed from past successful workflows. MANGO integrates reinforcement learning and textual gradients to jointly optimize workflow paths and agent behaviors, while a skipping mechanism prevents redundant updates to well-optimized agents for improving efficiency. Extensive experiments on seven benchmarks show that MANGO achieves up to 12.8% performance improvement over state-of-the-art baselines, enhances efficiency by 47.4%, and generalizes effectively to unseen domains. Our code and datasets are publicly available at https://github.com/openJiuwen-ai/agent-store/tree/main/community/mango.", "authors": ["Zheng Wang", "Yuang Liu", "Yangkai Ding"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.12943", "pdf_url": "https://arxiv.org/pdf/2605.12943v1", "arxiv_id": "2605.12943", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/openJiuwen-ai/agent-store/tree/main/community/mango", "venue": null, "quality_score": 0.65}
{"id": "3039edee21038e7a4199d7dc0a8ec7d4566194277171a50ce555bfc9dbef85a3", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcement Learning for Tool-Calling Agents in Fast Healthcare Interoperability Resources (FHIR)", "abstract": "Fast Healthcare Interoperability Resources (FHIR) is the dominant standard for interoperable exchange of healthcare data. In FHIR, electronic health records form a directed graph of resources. Answering clinically meaningful questions over FHIR requires agents to perform multi-step reasoning, filtering, and aggregation across multiple resource types. Prior work shows that even tool-augmented LLM agents (retrieval, code execution, multi-turn planning) often select the wrong resources or violate traversal constraints. We study this problem in the context of FHIR-AgentBench, a benchmark for realistic question answering over real-world hospital data, and frame reasoning on FHIR as a sequential decision-making problem over a queryable structured graph. We implement a multi-turn CodeAct agent and post-train it with reinforcement learning using a custom harness and tools. A LLM Judge provides execution-grounded rewards. Compared to prompt-based, closed-model baselines, RL post-training improves performance while enforcing data-integrity constraints. Empirically, our approach improves answer correctness from 50% (o4-mini) to 77% on FHIR-AgentBench using a smaller and cheaper Qwen3-8B model. We present an end-to-end post-training pipeline (environment building, harness construction, model training and custom evaluation) that reliably improves multi-turn reasoning over structured clinical graphs.", "authors": ["Marius S. Knorr", "Robert Müller", "Jan P. Bremer", "Nils Schweingruber"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.14126", "pdf_url": "https://arxiv.org/pdf/2605.14126v1", "arxiv_id": "2605.14126", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4627974c6c6f0f12ce52a9f283a15c50c8ad8544bbc75ca04aec56edc2b0b8a0", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborating in Multi-Armed Bandits with Strategic Agents", "abstract": "We study collaborative learning in multi-agent Bayesian bandit problems, where strategic agents collectively solve the same bandit instance. While multiple agents can accelerate learning by sharing information, strategic agents might prefer to free-ride and avoid exploration. We consider a setting with persistent agents that participate in multiple time periods. This is in contrast to most previous works on incentives in multi-agent MAB, which assume short-lived agents, namely each agent has a single decision to make and optimizes their expected reward in that single decision. As in the multi-agent MAB model with incentives, our model does not have monetary transfers, and the only incentives are through information sharing. We propose \\texttt{CAOS}, a mechanism that sustains collaboration as a Nash equilibrium while achieving strong regret guarantees. Our results demonstrate that collaborative exploration can be sustained purely through information sharing, achieving performance close to that of fully cooperative systems despite strategic behavior.", "authors": ["Idan Barnea", "Ofir Schlisselberg", "Yishay Mansour"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13145", "pdf_url": "https://arxiv.org/pdf/2605.13145v1", "arxiv_id": "2605.13145", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "8b6ed9395f084e06396a9ee82d752cc62472b282b2acd83cf88e624d6c210de1", "sources": ["arxiv", "semantic_scholar"], "title": "Submodular Multi-Agent Policy Learning for Online Distributed Task Allocation in Open Multi-Agent Systems", "abstract": "This paper studies multi-agent reinforcement learning with submodular team utilities for online distributed task allocation. In this setting, each agent selects one action from a local categorical policy, so feasible joint actions form a partition matroid over agent-action pairs. Classical multilinear extensions use independent Bernoulli sampling and therefore do not match the categorical policies executed by decentralized agents. To address this mismatch, we introduce the Partition Multilinear Extension (PME), a continuous relaxation whose value equals the expected team utility under factorized categorical policies. We prove that submodular difference rewards provide unbiased PME marginal-gradient information and yield a stagewise score-function policy-gradient estimator. Based on this connection, we propose SubMAPG, a centralized-training decentralized-execution policy-gradient framework with masked categorical policies and submodular difference-reward training signals. For the associated PME marginal-space projected stochastic-gradient dynamics, we prove a stagewise 1/2-approximation guarantee and sublinear dynamic regret in slowly varying environments, measured by the path length of the optimal PME marginals. To handle open systems with time-varying agents and targets, we instantiate SubMAPG with graph neural network policies. Experiments on multi-robot coverage and multi-target tracking show that SubMAPG outperforms local greedy and shared-reward baselines and is competitive with centralized myopic greedy strategies.", "authors": ["Jing Liu", "Yangyang Yang", "Luca Ballotta", "Fangfei Li", "Yang Tang", "Ruggero Carli"], "categories": ["eess.SY"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13269", "pdf_url": "https://arxiv.org/pdf/2605.13269v1", "arxiv_id": "2605.13269", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4022879579a0f5f894d699fd213321925bc0a394816d75e172e552618923b4ee", "sources": ["arxiv", "semantic_scholar"], "title": "TERMS-Bench: Diagnosing LLM Negotiation Agents Beyond Deal Rate", "abstract": "Negotiation is a central mechanism of economic exchange, shaping markets, procurement, labor agreements, and resource allocation. It is also a canonical testbed for agentic language models, requiring multi-turn interaction under hidden preferences, strategic communication, and binding constraints. These properties make negotiation hard to evaluate: unlike math or code, it has no intrinsic verifier. Existing LLM negotiation evaluations rely on LLM-vs.-LLM interaction or aggregate outcomes such as deal rate, leaving failures opaque. We introduce Terms-Bench, short for Testbed for Economic Reasoning in Multi-turn Strategy, a Bayesian-game framework that makes the environment itself the verifier by specifying the counterpart's latent type, policy, and payoff structure. We instantiate it in bilateral price negotiation, where the counterpart's private state and simulator policy are hidden from the agent but observable to the evaluator. This turns the counterpart from a black-box opponent into a diagnostic instrument, enabling agent-attributable failure analysis and oracle-reference optimality gaps. Evaluating 13 LLM agents spanning frontier systems from major providers, Terms-Bench turns negotiation evaluation from aggregate ranking into actionable diagnosis: where agents fail, why they fail, and what to strengthen. Empirically, frontier models saturate deal rate yet diverge in surplus extraction, cue use, belief calibration, and compliance, revealing agent-specific bargaining bottlenecks masked by prior benchmarks.", "authors": ["Erica Zhang", "Fangzhao Zhang", "Aneesh Pappu", "Batu El", "Jose Blanchet", "Susan Athey", "Jiashuo Liu", "James Zou"], "categories": ["cs.GT", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-13", "url": "https://arxiv.org/abs/2605.13909", "pdf_url": "https://arxiv.org/pdf/2605.13909v1", "arxiv_id": "2605.13909", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "82eb7e6fcce27ec07b72905f149571ae672b0d49474d3324e7e217cbc8895e0c", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-X: A Scalable Negotiation-Oriented Exchange for Communication Among Personal LLM Agents", "abstract": "We propose a personal-LLM exchange (LLM-X), a scalable negotiation-oriented environment that enables direct, structured communication across populations of personal agents (LLMs), each representing an individual user. Unlike existing tool-centric protocols that focus on agent-API interaction, LLM-X introduces a message bus and routing substrate for LLM-to-LLM coordination with guarantees around schema validity and policy enforcement. We contribute: (1) an architecture for LLM-X comprising federated gateways, topic-based routing, and policy enforcement; (2) a typed message protocol supporting capability negotiation and contract-net-style coordination; and (3) the first empirical evaluation of LLM-based multi-agent negotiation at scale. Experiments span 5, 9, and 12 agents, under distinct negotiation policies (Low, Medium, High), and across both short-run (minutes) and long-run (2h, 12h) load conditions. Results highlight clear policy-performance trade-offs: stricter policies improve robustness and fairness but increase latencies and message volume. Extended runs confirm that LLM-X remains stable under sustained load, with bounded latency drift.", "authors": ["Giuliano Lorenzoni", "Paulo Alencar", "Donald Cowan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11376", "pdf_url": "https://arxiv.org/pdf/2605.11376v1", "arxiv_id": "2605.11376", "doi": "10.1145/3786167.3788429", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "cbac687f7ad0f6f12cae28b4df6620b643eb2f3af4b01794c9175a3ab6f426fd", "sources": ["arxiv", "semantic_scholar"], "title": "Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations", "abstract": "Operational disaster response goes beyond damage assessment, requiring responders to integrate multi-sensor signals, reason over road networks, populations and key facilities, plan evacuations, and produce actionable reports. However, prior work largely isolates remote-sensing perception or evaluates generic tool use, leaving the end-to-end workflows of emergency operations underexplored. In this paper, we introduce Disaster Operational Response Agent benchmark (DORA), the first agentic benchmark for end-to-end disaster response: 515 expert-authored tasks across 45 real-world disaster events spanning 10 types, paired with expert-verified, replayable gold trajectories totaling 3,500 tool-call steps. Tasks span five dimensions that cover the operational disaster-response pipeline: disaster perception, spatial relational analysis, rescue and evacuation planning, temporal evolution reasoning, and multi-modal report synthesis. Agents compose calls from a 108-tool MCP library over heterogeneous geospatial data: optical, SAR, and multi-spectral imagery across single-, bi-, and multi-temporal sequences (0.015-10m GSD), complemented by elevation and social vector layers. We comprehensively evaluate 13 frontier LLMs on our benchmark, revealing three persistent challenges: 1) disaster-domain grounding exposes unique failure modes (damage-semantic grounding, sensor-modality mismatch, and disaster-pipeline composition); 2) agents are doubly bottlenecked by tool selection and argument grounding, where gold tool-order hints improve accuracy by only 1.08-4.40%, and alternative scaffolds yield at most a 3.24% gain; 3) compositional fragility scales with trajectory length, the agent-to-gold gap widening from 7% to 56% on long pipelines. DORA establishes a rigorous testbed for operationally reliable disaster-response agents.", "authors": ["Junjue Wang", "Weihao Xuan", "Heli Qi", "Pengyu Dai", "Kunyi Liu", "Hongruixuan Chen", "Zhuo Zheng", "Junshi Xia", "Stefano Ermon", "Naoto Yokoya"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11633", "pdf_url": "https://arxiv.org/pdf/2605.11633v1", "arxiv_id": "2605.11633", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d2e7284e3e15c93303ad4effeb4c0695ce0da65b310f2aab23d9a5d2ecedd63e", "sources": ["arxiv", "semantic_scholar"], "title": "FlowSteer: Prompt-Only Workflow Steering Exposes Planning-Time Vulnerabilities in Multi-Agent LLM Systems", "abstract": "Multi-agent systems (MAS) powered by large language models (LLMs) increasingly adopt planner--executor architectures, where planners convert prompts into subtasks, roles, dependencies, and routing paths. This flexibility enables adaptive coordination, but exposes an attack surface in workflow formation: prompts can shape agent organization without modifying MAS infrastructure. We study this risk through social influence probing workflows to identify high-impact subtasks and malicious-signal propagation. The analysis reveals two vulnerabilities: workflow position can amplify or suppress a malicious signal, and sycophantic framing makes downstream agents more likely to relay it. We translate these findings into FlowSteer, a prompt-only workflow steering attack that converts vulnerability priors into one crafted prompt. FlowSteer aligns a malicious signal with influential task components and guides replanning toward dependencies that preserve propagation. Experiments show that FlowSteer increases malicious success by up to 55% over naive prompting, transfers across MAS setups, and remains effective with black-box topology inference. As FlowSteer biases the planning signals that generate the workflow, MAS defenses that inspect only the generated workflow provide limited protection. As such, we introduce FlowGuard, an input-side defense that reduces malicious success by up to 34% while preserving prompt utility. Our results position workflow formation as a new safety frontier for multi-agent LLM systems, opening a planning-time security perspective on how agent coordination itself can be attacked and defended.", "authors": ["Fanxiao Li", "Jiaying Wu", "Tingchao Fu", "Natasha Jaques", "Wei Zhou", "Min-Yen Kan"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11514", "pdf_url": "https://arxiv.org/pdf/2605.11514v1", "arxiv_id": "2605.11514", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "02d238ce4bcdc8b7ec6b67af51f5eb81187d1e2214fcf32741eee6e89cef6cda", "sources": ["arxiv", "semantic_scholar"], "title": "Predictive Maps of Multi-Agent Reasoning: A Successor-Representation Spectrum for LLM Communication Topologies", "abstract": "Practitioners deploying multi-agent large language model (LLM) systems must currently choose between communication topologies such as chain, star, mesh, and richer variants without any pre-inference diagnostic for which topology will amplify drift, converge to consensus, or remain robust under perturbation. Existing evaluation answers these questions only post hoc and only for the task measured. We introduce a structural diagnostic for multi-agent LLM communication graphs based on the successor representation $M = (I - γP)^{-1}$ of the row-stochastic communication operator, and we connect three of its spectral quantities, the spectral radius $ρ(M)$, the spectral gap $Δ(M)$, and the condition number $κ(M)$, to three distinct failure modes. We derive closed-form spectra for the chain, star, and mesh under row-stochastic normalization, and validate the predictions on a 12-step structured state-tracking task with Qwen2.5-7B-Instruct over 100 independent trials. The condition number is a perfect rank-order predictor of empirical perturbation robustness ($r_s = 1.0$); the spectral gap partially predicts consensus dynamics ($r_s = 0.5$); and the spectral radius is perfectly \\emph{inverted} with respect to cumulative error ($r_s = -1.0$). We trace this inversion to a regime in which linear spectra are blind to non-contracting bias drift, and we propose an affine-noise extension of the predictive map that recovers the empirical ordering. We read this as a first step toward representational, drift-aware structural diagnostics for multi-agent LLM systems, sitting alongside classical spectral and consensus theory.", "authors": ["Ethan Parks", "Dalal Alharthi"], "categories": ["cs.MA", "cs.AI", "cs.LG", "cs.SI", "math.SP"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11453", "pdf_url": "https://arxiv.org/pdf/2605.11453v2", "arxiv_id": "2605.11453", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "cfde58fb27f5fd2fc739c3796209fca567a5d181c0beebdec8183ac47cb52025", "sources": ["arxiv", "semantic_scholar"], "title": "Coordinated Diffusion: Generating Multi-Agent Behavior Without Multi-Agent Demonstrations", "abstract": "Imitation learning powered by generative models has proven effective for modeling complex single-agent behaviors. However, teaching multi-agent systems, like multiple arms or vehicles, to coordinate through imitation learning is hindered by a fundamental data bottleneck: as the joint state-action space grows exponentially with the number of agents, collecting a sufficient amount of coordinated multi-agent demonstrations becomes extremely costly. In this work, we ask: how can we leverage single-agent demonstration data to learn multi-agent policies? We present Coordinated Diffusion (CoDi), a framework that couples independently trained single-agent diffusion policies through a user-defined multi-agent cost function, without requiring any coordinated demonstrations. We derive a new diffusion-based sampling scheme wherein the diffusion score function decomposes into independent, single-agent pre-trained base policies plus a cost-driven guidance term that coordinates these base policies into cohesive multi-agent behavior. We show that this guidance term can be estimated in a gradient-free manner, making CoDi applicable to black-box, non-differentiable cost functions without additional training. Theoretically and empirically, we analyze the conditions under which this composition can faithfully approximate a target multi-agent behavior. We find a complementary role for demonstration data versus the cost function: single-agent demonstrations must cover the support of the desired multi-agent behavior, while the cost function must promote desired behavior from this product of single-agent policies. Our results in simulation and hardware experiments of a two-arm manipulation task show that CoDi discovers robust coordinated behavior from single-agent data, is more data-efficient than multi-agent baselines, and highlights the importance of joint guidance, base policy support, and cost design.", "authors": ["Lasse Peters", "Laura Ferranti", "Andrea Bajcsy", "Javier Alonso-Mora"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11485", "pdf_url": "https://arxiv.org/pdf/2605.11485v2", "arxiv_id": "2605.11485", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "fd223853382fed2eb9a07980f3b2f3e60c0c80bf4ac153dd2da3f12e23e8eebe", "sources": ["arxiv", "semantic_scholar"], "title": "No Action Without a NOD: A Heterogeneous Multi-Agent Architecture for Reliable Service Agents", "abstract": "Large language model (LLM) agents have increasingly advanced service applications, such as booking flight tickets. However, these service agents suffer from unreliability in long-horizon tasks, as they often produce policy violations, tool hallucinations, and misaligned actions, which greatly impedes their real-world deployment. To address these challenges, we propose NOD (Navigator-Operator-Director), a heterogeneous multi-agent architecture for service agents. Instead of maintaining task state implicitly in dialogue context as in prior work, we externalize a structured Global State to enable explicit task state tracking and consistent decision-making by the Navigator. Besides, we introduce selective external oversight before critical actions, allowing an independent Director agent to verify execution and intervene when necessary. As such, NOD effectively mitigates error propagation and unsafe behavior in long-horizon tasks. Experiments on $τ^2$-Bench demonstrate that NOD achieves higher task success rates and critical action precision over baselines. More importantly, NOD improves the reliability of service agents by reducing policy violations, tool hallucinations, and user-intent misalignment.", "authors": ["Zixu Yang", "Hang Zheng", "Nan Jiang", "Zhiyang Tang", "Situo Zhang", "Xiaobao Wu", "Lu Chen", "Kai Yu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.12240", "pdf_url": "https://arxiv.org/pdf/2605.12240v1", "arxiv_id": "2605.12240", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "577e461f8790899ce88458cd0ef2d2ebbb0a20b72d39614dfe711cb8da73d245", "sources": ["arxiv", "semantic_scholar"], "title": "CTFusion: A CTF-based Benchmark for LLM Agent Evaluation", "abstract": "Recent advances in Large Language Models (LLMs) have enabled agentic systems for complex, multi-step tasks; cybersecurity is emerging as a prominent application. To evaluate such agents, researchers widely adopt Capture The Flag (CTF) benchmarks. However, current CTF benchmarks reuse existing challenges, which exposes them to data contamination and potential cheating. Notably, we confirmed these issues in practice by integrating web search tools into an existing agent. To address these limitations, we present CTFusion, a streaming evaluation framework built on Live CTFs. To achieve this, CTFusion preserves per-agent independence under a single team account and reduces competition impact by forwarding only the first correct flag per challenge. Moreover, we implement CTFusion as a Model Context Protocol (MCP) server on the widely used CTFd platform, which offers broad applicability to diverse CTF events and agent types. Through experiments with three LLMs, two agents, and five Live CTFs, we demonstrate that existing CTF benchmarks can be unreliable in assessing LLM-based agents, while CTFusion can serve as a robust solution for evaluating cybersecurity agents. We release CTFusion as open source to foster future research in this area.", "authors": ["Dongjun Lee", "Ga-eun Bae", "Insu Yun"], "categories": ["cs.LG", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-12", "url": "https://arxiv.org/abs/2605.11504", "pdf_url": "https://arxiv.org/pdf/2605.11504v1", "arxiv_id": "2605.11504", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "01ed7cb89353aab6eae6be4038ee642394bef0132974dc076a28c6841a52a4d8", "sources": ["arxiv", "semantic_scholar"], "title": "AnomalyClaw: A Universal Visual Anomaly Detection Agent via Tool-Grounded Refutation", "abstract": "Visual anomaly detection (VAD) is crucial in many real-world fields, such as industrial inspection, medical imaging, infrastructure monitoring, and remote sensing. However, the specific anomaly definitions, data modalities, and annotation standards across different domains make it difficult to transfer single-domain trained VAD models. Vision-language models (VLMs), pre-trained on large-scale cross-domain data, can perform visual perception under task instructions, offering a promising solution for cross-domain VAD. However, single-inference VLM judgments are unreliable, since they rely more on prior knowledge than on normal-sample references or fine-grained feature evidence. We therefore present AnomalyClaw, a training-free VAD agent that turns anomaly judgment into a multi-round refutation process. In each round, the agent proposes candidate anomalies and refutes each against normal-sample references, drawing on a 13-tool library for visual verification, reference parsing, and frozen expert probing. On the CrossDomainVAD-12 benchmark (12 datasets), AnomalyClaw achieves consistent macro-AUROC improvements over single-step direct inference with +6.23 pp on GPT-5.5, +7.93 pp on Seed2.0-lite, and +3.52 pp on Qwen3.5-VL-27B. We further introduce an optional verbalized self-evolution extension. It builds an online rulebook from internal-branch disagreement without oracle labels. On Qwen3.5-VL-27B, it delivers a +2.09 pp mean gain, comparable to a K = 10 oracle-label supervised baseline (+1.99 pp). These results show that agentic refutation improve anomaly understanding and reasoning of VLMs, rather than merely aggregating tool outputs.", "authors": ["Xi Jiang", "Yinjie Zhao", "Zesheng Yang", "Feng Zheng"], "categories": ["cs.CV", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10397", "pdf_url": "https://arxiv.org/pdf/2605.10397v1", "arxiv_id": "2605.10397", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/jam-cc/AnomalyClaw", "venue": null, "quality_score": 0.65}
{"id": "deb049020bdf018c6efc575012cff59013a0f8a9b412f684cb967f0bcd9ab7f8", "sources": ["arxiv", "semantic_scholar"], "title": "OLIVIA: Online Learning via Inference-time Action Adaptation for Decision Making in LLM ReAct Agents", "abstract": "Large language model agents interleave reasoning, action selection, and observation to solve sequential decision-making tasks. In deployed settings where agents repeatedly handle related multi-step tasks, small action-selection errors can accumulate into wasted tool calls, latency, and reduced reliability. Despite this need for deployment-time improvement, existing inference-time adaptation methods for LLM agents mainly rely on prompting or retrieval, which influence behavior indirectly through context manipulation. For ReAct-style agents, such approaches do not expose an explicit decision layer that can score candidate actions, represent uncertainty, or be updated online from action-level feedback. As a result, they provide limited support for trackable, fine-grained, and uncertainty-aware adaptation during deployment. We propose OLIVIA, an inference-time action adaptation framework for ReAct-style agents. OLIVIA models the LLM's final action-selection layer as a contextual linear bandit over candidate actions, with frozen hidden states as decision contexts. This choice is particularly suitable for deployment because it adapts behavior directly at the action-selection interface, preserves the underlying reasoning process, and provides explicit uncertainty estimates and lightweight online updates from action-level feedback. With upper-confidence-bound exploration, OLIVIA improves the policy sample-efficiently with minimal computational overhead. We instantiate OLIVIA on four benchmarks and show that it consistently improves task performance over static ReAct and prompt-based inference-time baselines. Our results suggest that explicit online decision layers provide an effective alternative to purely prompt- or retrieval-based adaptation for LLM agents during deployment.", "authors": ["Sheldon Yu", "Junda Wu", "Xintong Li", "Nikki Lijing Kuang", "Sizhe Zhou", "Tong Yu", "Jiawei Han", "Jingbo Shang", "Julian McAuley"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.11169", "pdf_url": "https://arxiv.org/pdf/2605.11169v1", "arxiv_id": "2605.11169", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b5e7edd8fd8732c151b6a618bb97b8dcc16346b08e9c702a8e16fdf0699ccdb9", "sources": ["arxiv", "semantic_scholar"], "title": "Safe Multi-Agent Behavior Must Be Maintained, Not Merely Asserted: Constraint Drift in LLM-Based Multi-Agent Systems", "abstract": "Modern LLM based agents are no longer passive text generators. They read repositories, call tools, browse the web, execute code, maintain memory, communicate with other agents, and act through long horizon workflows. This shift moves the unit of safety. A system may produce a compliant final answer while leaking private information through an internal message, delegating authority beyond its original scope, calling an external tool with sensitive context, or losing the evidence needed to reconstruct why an action was allowed. We argue that many emerging failures in LLM-based multi-agent systems share a common structure: safety critical constraints do not remain operative throughout the trajectory. We call this phenomenon constraint drift: the loss, distortion, weakening, or relaxation of constraints as they pass through memory, delegation, communication, tool use, audit, and optimization. The position taken here is that safe multi-agent behavior must be maintained, not merely asserted. Prompts, guardrails, tool schemas, access control, and final output checks are necessary, but they are insufficient unless constraints remain fresh, inherited, enforceable, and auditable across execution. We propose Constraint State Governance as a research paradigm for LLM-based multi-agent systems. In this paradigm, safety-critical constraints are maintained as explicit execution state, while constraint-native reinforcement learning improves utility only within maintained safety boundaries. The goal is not to freeze agentic systems under rigid rules, but to make safety operational across the trajectories through which modern agents actually act.", "authors": ["Tianxiao Li", "Yixing Ma", "Haiquan Wen", "Zhenglin Huang", "Qianyu Zhou", "Zeyu Fu", "Guangliang Cheng"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10481", "pdf_url": "https://arxiv.org/pdf/2605.10481v1", "arxiv_id": "2605.10481", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "24b643bc218338eb499555efe3b153632be2498cae7379795f5c2dd7c9931ec7", "sources": ["arxiv", "semantic_scholar"], "title": "RecoAtlas: From Semantic Plausibility to Set-Level Utility in LLM Recommendation Agents", "abstract": "LLM recommendation agents increasingly produce structured recommendation reports: sets of items accompanied by natural-language justifications. Yet existing evaluations often reduce this setting to reranking small shortlisted candidate sets or judge reports mainly by semantic plausibility. We introduce Recommendation Atlas (Agentic Tool-Level Assessment for Shopping), or RecoAtlas, a benchmark and toolkit for evaluating shopping agents with behavior-grounded metrics. RecoAtlas complements held-out interaction metrics with learned utility proxies for relevance, complementarity, and diversity derived from interaction data, while separately measuring semantic coherence and explanation quality. Its controlled tool environment exposes agents to either semantic, behavior-aligned, or faulty tools, enabling diagnosis of whether performance gains arise from stronger reasoning, better signals, or more effective tool-use policies. Across controlled experiments, we show that RecoAtlas exhibits key properties of a meaningful benchmark for agentic systems: performance scales with model capacity and test-time compute, improves with stronger and better-aligned tools, degrades under noisy or misaligned signals, and reveals that semantic plausibility does not necessarily capture behavior-grounded utility. RecoAtlas provides a foundation for developing and evaluating shopping assistants that optimize not only for plausible recommendations, but also for coherent, behaviorally grounded recommendation sets.", "authors": ["Imad Aouali", "Flavian Vasile", "Otmane Sakhi", "Alexandre Gilotte", "Benjamin Heymann"], "categories": ["cs.IR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.18805", "pdf_url": "https://arxiv.org/pdf/2605.18805v1", "arxiv_id": "2605.18805", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "81552708381cc2572b9140dc5244c9b85e6ea83c153011b5c9d60bd9bcb0c4db", "sources": ["arxiv", "semantic_scholar"], "title": "AgentRx: A Benchmark Study of LLM Agents for Multimodal Clinical Prediction Tasks", "abstract": "Building effective clinical decision support systems requires the synthesis of complex heterogeneous multimodal data. Such modalities include temporal electronic health records data, medical images, radiology reports, and clinical notes. Large language model (LLM)-based agents have shown impressive performance in various healthcare tasks, especially those involving textual modalities. Considering the fragmentation of healthcare data across hospital systems, collaborative agent frameworks present a promising direction to mitigate data sharing challenges. However, the effectiveness of LLM agents for multimodal clinical risk prediction remains largely unexamined. In this work, we conduct a systematic evaluation of LLM-based agents for clinical prediction tasks using large-scale real-world data. We assess performance in unimodal and multimodal settings and quantify performance gaps between single agent and multi-agent systems. Our findings highlight that single agent frameworks outperform naive multi-agent systems, are better at handling multimodal data, and are better calibrated. This underscores a critical need for improving multi-agent collaboration to better handle heterogeneous inputs. By open-sourcing our code and evaluation framework, this work offers a new benchmark to support future developments relating to agentic systems in healthcare.", "authors": ["Baraa Al Jorf", "Farah E. Shamout"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10286", "pdf_url": "https://arxiv.org/pdf/2605.10286v1", "arxiv_id": "2605.10286", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9343e5bf88ac1848c0470ca375df1914687790e5679e483674f76bec05f3803b", "sources": ["arxiv", "semantic_scholar"], "title": "Collective Alignment in LLM Multi-Agent Systems: Disentangling Bias from Cooperation via Statistical Physics", "abstract": "We investigate the emergent collective dynamics of LLM-based multi-agent systems on a 2D square lattice and present a model-agnostic statistical-physics method to disentangle social conformity from intrinsic bias, compute critical exponents, and probe the collective behavior and possible phase transitions of multi-agent systems. In our framework, each node of an $L\\!\\times\\!L$ lattice hosts an identical LLM agent holding a binary state ($+1$/$-1$, mapped to yes/no) and updating it by querying the model conditioned on the four nearest-neighbor states. The sampler temperature $T$ serves as the sole control parameter. Across three open-weight models (llama3.1:8b, phi4-mini:3.8b, mistral:7b), we measure magnetization and susceptibility under a global-flip protocol designed to probe $\\mathbb{Z}_2$ symmetry. All models display temperature-driven order-disorder crossovers and susceptibility peaks; finite-size scaling on even-$L$ lattices yields effective exponents $γ/ν$ whose values are model-dependent, close to but incompatible with the 2D Ising universality class ($γ/ν=7/4$). Our method enables the extraction of effective $β$-weighted couplings $\\tilde{J}(T)$ and fields $\\tilde{h}(T)$, which serve as a measure of social conformity and intrinsic bias. In the models we analyzed, we found that collective alignment is dominated by an intrinsic bias ($\\tilde{h}\\gg\\tilde{J}$) rather than by cooperative neighbor coupling, producing field-driven crossovers instead of genuine phase transitions. These effective parameters vary qualitatively across models, providing compact collective-behavior fingerprints for LLM agents and a quantitative diagnostic for the reliability of multi-agent consensus and collective alignment.", "authors": ["Cristiano De Nobili"], "categories": ["cond-mat.stat-mech", "cs.CL", "cs.MA", "physics.soc-ph"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10528", "pdf_url": "https://arxiv.org/pdf/2605.10528v1", "arxiv_id": "2605.10528", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c4ce8d2d143644c4667df938fc9028c48848551d88c883ec6e725a3e80de4ca7", "sources": ["arxiv", "semantic_scholar"], "title": "Control Charts for Multi-agent Systems", "abstract": "Generative agents have proven to be powerful assistants in a wide variety of contexts. Given this success, users are now deploying agents with minimal restrictions in open ended, multi-agent environments. Current methods for monitoring the dynamics of open-ended multi-agent systems are limited to qualitative inspection. In this paper, we extend the process-theoretic notion of adaptive control charts to multi-agent systems to enable automated monitoring. Using simulation, we demonstrate that adaptive control charts are necessary for monitoring multi-agent systems that can learn from their environment. We further demonstrate, both empirically and theoretically, that adaptive control charts are susceptible to adversarial agents that defect sufficiently slowly. These results illustrate a fundamental tradeoff in multi-agent system control: either agents in a system cannot learn or the system is susceptible to adversaries.", "authors": ["Hayden Helm", "Carey Priebe", "Brandon Duderstadt"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.11135", "pdf_url": "https://arxiv.org/pdf/2605.11135v1", "arxiv_id": "2605.11135", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "99e1d26655b759cda4d62b2e3bdcf6efddae87338230ac3c3c03591eb0b3900b", "sources": ["arxiv", "semantic_scholar"], "title": "TRACER: Verifiable Generative Provenance for Multimodal Tool-Using Agents", "abstract": "Multimodal large language models increasingly solve vision-centric tasks by calling external tools for visual inspection, OCR, retrieval, calculation, and multi-step reasoning. Current tool-using agents usually expose the executed tool trajectory and the final answer, but they rarely specify which tool observation supports each generated claim. We call this missing claim-level dependency structure the provenance gap. The gap makes tool use hard to verify and hard to optimize, because useful evidence, redundant exploration, and unsupported reasoning are mixed in the same trajectory. We introduce TRACER, a framework for verifiable generative provenance in multimodal tool-using agents. Instead of adding citations after generation, TRACER generates each answer sentence together with a structured provenance record that identifies the supporting tool turn, evidence unit, and semantic support relation. Its relation space contains Quotation, Compression, and Inference, covering direct reuse, faithful condensation, and grounded derivation. TRACER verifies each record through schema checking, tool-turn alignment, source authenticity, and relation rationality, and then converts verified provenance into traceability constraints and provenance-derived local credit for reinforcement learning. We further construct TRACE-Bench, a benchmark for sentence-level provenance reconstruction from coarse multimodal tool trajectories. On TRACE-Bench, simply adding tools often introduces noise. With Qwen3-VL-8B, TRACER reaches 78.23% answer accuracy and 95.72% summary accuracy, outperforming the strongest closed-source tool-augmented baseline by 23.80 percentage points. Compared with tool-only supervised fine-tuning, it also reduces total test-set tool calls from 4949 to 3486. These results show that reliable multimodal tool reasoning depends on provenance-aware use of observations, not on more tool calls alone.", "authors": ["Bihui Yu", "Caijun Jia", "Jing Chi", "Xiaohan Liu", "Yining Wang", "He Bai", "Yuchen Liu", "Jingxuan Wei", "Junnan Zhu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.09934", "pdf_url": "https://arxiv.org/pdf/2605.09934v1", "arxiv_id": "2605.09934", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a2a46ddccd7f4791969aa3c599d76ed541b12581c6cb1048267bba54b9028dfb", "sources": ["arxiv", "semantic_scholar"], "title": "PRISM: Generation-Time Detection and Mitigation of Secret Leakage in Multi-Agent LLM Pipelines", "abstract": "Multi-agent LLM systems introduce a security risk in which sensitive information accessed by one agent can propagate through shared context and reappear in downstream outputs, even without explicit adversarial intent. We formalise this phenomenon as propagation amplification, where leakage risk increases across agent boundaries as sensitive content is repeatedly exposed to downstream generators. Existing defences, including prompt-based safeguards, static pattern matching, and LLM-as-judge filtering, are not designed for this setting: they either operate after generation, rely primarily on surface-form patterns, or add substantial latency without modelling the generation process itself. To resolve these issues, we propose PRISM, a real-time defence that treats credential leakage as a sequential risk accumulation problem during generation. At each decoding step, PRISM combines 16 signals spanning lexical, structural, information-theoretic, behavioural, and contextual features into a calibrated risk score, enabling per-token intervention through green, yellow, and red risk zones. Our central observation is that credential reproduction is often preceded by a measurable shift in generation dynamics, characterised by entropy collapse and increasing logit concentration. When combined with text-structural cues such as identifier-pattern detection, these temporal signals provide an early warning of leakage before a secret is fully reconstructed. Across a 2,000-task adversarial benchmark covering 13 attack categories and three pressure levels in a heterogeneous four-agent pipeline, PRISM achieves F1 = 0.832 with precision = 1.000 and recall = 0.712, while producing no observed leakage on our benchmark (0.0% task-level leak rate) and preserving output utility of 0.893. It substantially outperforms the strongest baseline, Span Tagger, which achieves F1 = 0.719 with a 15.0% task-level leak rate.", "authors": ["Riya Tapwal", "Abhishek Kumar", "Carsten Maple"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10614", "pdf_url": "https://arxiv.org/pdf/2605.10614v1", "arxiv_id": "2605.10614", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3f6dac0296b9761823c1e70d7524c0e9ab0cd55d4d00f7691c95cce392ab5f3c", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-First Tool API: A Semantic Interface Paradigm for Enterprise AI Agent Systems", "abstract": "As AI agents transition from research prototypes to enterprise production systems, the tool interfaces they consume remain rooted in human-oriented CRUD paradigms. This paper identifies five fundamental architectural mismatches between conventional APIs and autonomous agent requirements: exact-identifier dependence, rendering-oriented responses, single-shot interaction assumptions, user-equivalent authorization, and opaque error semantics. We propose the Agent-First Tool API paradigm, comprising three integrated mechanisms: (1) a Six-Verb Semantic Protocol that decomposes tool interactions into search, resolve, preview, execute, verify, and recover phases; (2) a Normalized Tool Contract (NTC) providing structured decision-support metadata including confidence scores, evidence chains, and suggested next actions; and (3) a dual-layer governance pipeline combining static capability policies with dynamic risk escalation. The paradigm is implemented and validated in a production multi-tenant SaaS platform serving 85 registered tools across 6 business domains. Comparative experiments on 50 real operational tasks demonstrate that Agent-First APIs achieve 88% end-to-end task success rate versus 64% for optimized CRUD baselines (+37.5%), while reducing required human interventions by 72.7% and improving autonomous error recovery by 5.8x. We establish that the paradigm is orthogonal and complementary to transport-layer standards such as MCP, operating as the semantic application layer above existing tool discovery and invocation protocols.", "authors": ["Kai Pan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10555", "pdf_url": "https://arxiv.org/pdf/2605.10555v1", "arxiv_id": "2605.10555", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c6895095b32772656457af695ee443d710d744382abfe87788e766b26b9dd727", "sources": ["arxiv", "semantic_scholar"], "title": "LITMUS: Benchmarking Behavioral Jailbreaks of LLM Agents in Real OS Environments", "abstract": "The rapid proliferation of LLM-based autonomous agents in real operating system environments introduces a new category of safety risk beyond content safety: behavior jailbreak, where an adversary induces an agent to execute dangerous OS-level operations with irreversible consequences. Existing benchmarks either evaluate safety at the semantic layer alone, missing physical-layer harms, or fail to isolate test cases, letting earlier runs contaminate later ones. We present LITMUS (LLM-agents In-OS Testing for Measuring Unsafe Subversion), a benchmark addressing both gaps via a semantic-physical dual verification mechanism and OS-level state rollback. LITMUS comprises 819 high-risk test cases organized into one harmful seed subset and six attack-extended subsets covering three adversarial paradigms (jailbreak speaking, skill injection, and entity wrapping), plus a fully automated multi-agent evaluation framework judging behavior at both conversational and OS-level physical layers. Evaluation across frontier agents reveals three findings: (1) current agents lack effective safety awareness, with strong models (e.g., Claude Sonnet 4.6) still executing 40.64% of high-risk operations; (2) agents exhibit pervasive Execution Hallucination (EH), verbally refusing a request while the dangerous operation has already completed at the system level, invisible to every prior semantic-only framework; and (3) skill injection and entity wrapping attacks achieve high success rates, exposing pronounced agent vulnerabilities. LITMUS provides the first standardized platform for reproducible, physically grounded behavioral safety evaluation of LLM agents in real OS environments.", "authors": ["Chiyu Zhang", "Huiqin Yang", "Bendong Jiang", "Xiaolei Zhang", "Yiran Zhao", "Ruyi Chen", "Lu Zhou", "Xiaogang Xu", "Jiafei Wu", "Liming Fang", "Zhe Liu"], "categories": ["cs.CR", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-11", "url": "https://arxiv.org/abs/2605.10779", "pdf_url": "https://arxiv.org/pdf/2605.10779v1", "arxiv_id": "2605.10779", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "dae0121a44690a3903617e80bf7b7cc778c6e5eb1fc8154a21a69ceadd22e36d", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agents Already Know When to Call Tools -- Even Without Reasoning", "abstract": "Tool-augmented LLM agents tend to call tools indiscriminately, even when the model can answer directly. Each unnecessary call wastes API fees and latency, yet no existing benchmark systematically studies when a tool call is actually needed. We propose When2Tool, a benchmark of 18 environments (15 single-hop, 3 multi-hop) spanning three categories of tool necessity -- computational scale, knowledge boundaries, and execution reliability -- each with controlled difficulty levels that create a clear decision boundary between tool-necessary and tool-unnecessary tasks. We evaluate two families of training-free baselines: Prompt-only (varying the prompt to discourage unnecessary calls) and Reason-then-Act (requiring the model to reason about tool necessity before acting). Both provide limited control: Prompt-only suppresses necessary calls alongside unnecessary ones, and Reason-then-Act still incurs a disproportionate accuracy cost on hard tasks. To understand why these baselines fail, we probe the models' hidden states and find that tool necessity is linearly decodable from the pre-generation representation with AUROC 0.89--0.96 across six models, substantially exceeding the model's own verbalized reasoning. This reveals that models already know when tools are needed, but fail to act on this knowledge during generation. Building on this finding, we propose Probe&Prefill, which uses a lightweight linear probe to read the hidden-state signal and prefills the model's response with a steering sentence. Across all models tested, Probe&Prefill reduces tool calls by 48% with only 1.7% accuracy loss, while the best baseline at comparable accuracy only reduces 6% of tool calls, or achieves a similar tool call reduction but incurs a 5$\\times$ higher accuracy loss. Our code is available at https://github.com/Trustworthy-ML-Lab/when2tool", "authors": ["Chung-En Sun", "Linbo Liu", "Ge Yan", "Zimo Wang", "Tsui-Wei Weng"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-10", "url": "https://arxiv.org/abs/2605.09252", "pdf_url": "https://arxiv.org/pdf/2605.09252v2", "arxiv_id": "2605.09252", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Trustworthy-ML-Lab/when2tool", "venue": null, "quality_score": 0.65}
{"id": "1c6654509e569f2015d971b216cac222f9327a5b11447305594ab32a57a15e4b", "sources": ["arxiv", "semantic_scholar"], "title": "SkillMAS: Skill Co-Evolution with LLM-based Multi-Agent System", "abstract": "Large language model (LLM) agent systems are increasingly expected to improve after deployment, but existing work often decouples two adaptation targets: skill evolution and multi-agent system (MAS) restructuring. This separation can create organization bottlenecks, context pressure, and mis-specialization. We present SkillMAS, a non-parametric framework for adaptive specialization in multi-agent systems that couples skill evolution with MAS restructuring. SkillMAS uses Utility Learning to assign credit from verified execution traces, bounded skill evolution to refine reusable procedures without unfiltered library growth, and evidence-gated MAS restructuring when retained failures and Executor Utility indicate a structural mismatch. Across embodied manipulation, command-line execution, and retail workflows, SkillMAS is competitive under the reported harnesses while clarifying how post-deployment specialization is attributed, updated, and applied.", "authors": ["Shuai Pan", "Yixiang Liu", "Jiaye Gao", "Te Gao", "Weiwen Liu", "Jianghao Lin", "Zhihui Fu", "Jun Wang", "Weinan Zhang", "Yong Yu"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-10", "url": "https://arxiv.org/abs/2605.09341", "pdf_url": "https://arxiv.org/pdf/2605.09341v2", "arxiv_id": "2605.09341", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "39902d887ac8f9b0ac3c3bcb84508fcb788f66b432ba59685c1b8bfda5361d8a", "sources": ["arxiv", "semantic_scholar"], "title": "TacoMAS: Test-Time Co-Evolution of Topology and Capability in LLM-based Multi-Agent Systems", "abstract": "Multi-agent systems (MAS) have emerged as a promising paradigm for solving complex tasks. Recent work has explored self-evolving MAS that automatically optimize agent capabilities or communication topologies. However, existing methods either learn a topology that remains fixed at inference time or adapt only the topology or capability during inference. We empirically and theoretically show that effective test-time evolution requires jointly adapting both axes, but on different time scales: capabilities should update rapidly to handle emerging subtasks, while the topology should evolve more slowly to preserve coordination stability. We then introduce TacoMAS, a test-time co-evolution framework for dynamic MAS. TacoMAS formulates MAS inference as a task of online graph adaptation, where nodes represent agents with role-specific capabilities and edges define their communication topology. During inference, a fast capability loop updates agent expertise using trajectory-level feedback, while a slow meta-LLM-driven topology loop performs agents' birth-death operations on MAS, including edge edit, agent addition, and agent removal. We further show that this fast-slow design drives MAS evolution toward a task-conditioned stable equilibrium. Experiments on four benchmarks demonstrate that TacoMAS outperforms nearly 20 multi-agent baselines, achieving an average improvement of 13.3% over the strongest baseline. The codes are released at https://github.com/chenxu2-gif/TacoMAS-MultiAgent.", "authors": ["Chen Xu", "Yicheng Hu", "Ruizi Wang", "Xinyu Lin", "Wenjie Wang", "Dongrui Liu", "Fuli Feng"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-10", "url": "https://arxiv.org/abs/2605.09539", "pdf_url": "https://arxiv.org/pdf/2605.09539v1", "arxiv_id": "2605.09539", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/chenxu2-gif/TacoMAS-MultiAgent", "venue": null, "quality_score": 0.65}
{"id": "0fb6e3494c8ffcdb0061b65bc1df6d173a398edd7d9d5d3f9b2216593f027eb4", "sources": ["arxiv", "semantic_scholar"], "title": "AgentShield: Deception-based Compromise Detection for Tool-using LLM Agents", "abstract": "Defenses against indirect prompt injection (IPI) in tool-using LLM agents share two structural weaknesses. First, they all attempt to prevent attacks rather than detect the compromises that slip through. Second, they have only been evaluated in English, leaving users of low-resource languages such as Kurdish and Arabic without tested protection. This paper addresses both gaps with AgentShield, a deception-based detection framework that places three layers of traps inside the agent's tool interface: fake tools, fake credentials, and allowlisted parameters. The same trap triggers serve as high-precision labels for a self-supervised classifier. An LLM agent that follows an attacker's hidden instruction almost always touches one of these traps, which gives both a real-time compromise signal and a zero-FP label for training a downstream detector without manual annotation. Across 176 cross-lingual attack prompts and four LLMs from three providers, and because modern LLMs already refuse most IPI attempts on their own (attack success rate <= 10%), AgentShield's job is to catch the attacks that do slip through. On commercial models, it catches 90.7%-100% of such successful attacks, with zero false alarms on 485 normal-use tests. It survives a systematic adaptive-attack evaluation with zero evasion on commercial models, and the self-supervised classifier transfers across models and languages without retraining.", "authors": ["Yassin H. Rassul", "Tarik A. Rashid"], "categories": ["cs.CR", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-10", "url": "https://arxiv.org/abs/2605.11026", "pdf_url": "https://arxiv.org/pdf/2605.11026v1", "arxiv_id": "2605.11026", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Yassin-H-Rassul/AgentShield", "venue": null, "quality_score": 0.65}
{"id": "6d27f0b76e7e77d7e1d16936a5fc5e00982fa4c6476bb536fe6315dac2f1e10a", "sources": ["arxiv", "semantic_scholar"], "title": "CalBench: Evaluating Coordination-Privacy Trade-offs in Multi-Agent LLMs", "abstract": "Personal AI assistants are beginning to act as delegates with access to calendars, inboxes, and user preferences. Calendar scheduling makes the trust problem concrete: an assistant must coordinate with other assistants while deciding what to reveal about the person it represents. We introduce CalBench, a controlled benchmark for multi-agent calendar scheduling under private information. In each task, $N$ agents manage separate private calendars and schedule a stream of $M$ incoming meetings while minimizing disruption costs. Because no agent can inspect another agent's calendar, success requires language-mediated coordination rather than centralized planning. CalBench generates solvable scenarios with CP-SAT oracle solutions and decentralized non-LLM reference protocols, enabling evaluation of task success, excess cost, communication efficiency, burden fairness, and privacy leakage under matched information constraints. Across seven model families, we find that completion alone misses important failures: agents leave avoidable cost on the table, communication volume does not predict lower regret, and privacy-preserving silence can deprive teammates of cost information needed for fair burden allocation. CalBench provides a reproducible testbed for studying whether autonomous assistants can coordinate on behalf of users before deployment at scale.", "authors": ["Chelsea Zou", "Yiheng Yao", "Selena She", "Noah Goodman", "Robert D. Hawkins"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-10", "url": "https://arxiv.org/abs/2605.09823", "pdf_url": "https://arxiv.org/pdf/2605.09823v3", "arxiv_id": "2605.09823", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f67bc9d13d94478275248363cc2a72f91d451b8c323d773daf65df6a6f063b97", "sources": ["arxiv", "semantic_scholar"], "title": "Learning the Preferences of a Learning Agent", "abstract": "For AI systems to be useful to humans, they must understand and act in accordance with our values and preferences. Since specifying preferences is a hard task, inverse reinforcement learning (IRL) aims to develop methods that allow for inferring preferences from observed behavior. However, IRL assumes the human to be approximately optimal. This is a big limitation in cases where the human themselves may be learning to act optimally in an environment. In this paper, we formalize the problem of learning the preferences of a learning agent: a predictor observes a learner acting online and tries to infer the underlying reward function being (initially suboptimally) optimized by the learner. We model the learner as either being no-regret, or as converging to an optimal Boltzmann policy over time. In each of these settings, we establish theoretical guarantees for various preference learning algorithms, or otherwise show that such guarantees are impossible.", "authors": ["Karim Abdel Sadek", "Mark Bedaywi", "Rhys Gould", "Stuart Russell"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.09217", "pdf_url": "https://arxiv.org/pdf/2605.09217v1", "arxiv_id": "2605.09217", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2d867ea02e2a5e5d89e973f704c7e3bd7e84622ed27449c19286ea074345b937", "sources": ["arxiv", "semantic_scholar"], "title": "Robust Multi-Agent LLMs under Byzantine Faults", "abstract": "Large language model (LLM) agents increasingly collaborate over peer-to-peer networks to improve their reliability. However, these same interactions can also become a source of vulnerability, as unreliable or Byzantine agents may sway neighboring agents toward incorrect conclusions and degrade overall system performance. Existing methods rely on leader-based coordination or self-reported confidence, both of which are susceptible to adversarial manipulation. We study decentralized LLM multi-agent systems (LLM-MAS) and propose Self-Anchored Consensus (SAC), a fully decentralized iterative filter-and-refine protocol in which agents iteratively exchange responses, locally evaluate and filter unreliable messages, and refine their own outputs. We present $(F{+}1)$-robustness conditions for the communication graph that ensure honest agents preserve and propagate reliable information despite Byzantine influence. Experiments on mathematical and commonsense reasoning benchmarks show that SAC effectively suppresses Byzantine influence and consistently improves performance across diverse communication topologies, whereas prior methods degrade under adversarial conditions.", "authors": ["Haejoon Lee", "Vincent-Daniel Yun", "Hyeonho Oh", "Dimitra Panagou", "Sai Praneeth Karimireddy"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.09076", "pdf_url": "https://arxiv.org/pdf/2605.09076v1", "arxiv_id": "2605.09076", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "91734bcd64b2c71d71b636b424dd40b3357b8d94aa4855bf2ea7fd5167a9526f", "sources": ["arxiv", "semantic_scholar"], "title": "Iterative Critique-and-Routing Controller for Multi-Agent Systems with Heterogeneous LLMs", "abstract": "Multi-agent large language model (LLM) systems often rely on a controller to coordinate a pool of heterogeneous models, yet existing controllers are typically limited to one-shot routing: they select a model once and return its output directly. Such routing-only designs provide no mechanism to critique intermediate drafts or support iterative refinement. To address this limitation, we propose a critique-and-routing controller that casts multi-agent coordination as a sequential decision problem. At each turn, the controller evaluates the current draft, decides whether to stop or continue, and, if needed, selects the next agent for further refinement. We formulate this process as a finite-horizon Markov Decision Process (MDP) with explicit agent-utilization constraints, design a composite reward for controller decisions across turns, and optimize the controller via policy gradients under a Lagrangian-relaxed objective. Extensive experiments across multiple heterogeneous multi-agent systems and seven reasoning benchmarks show that our method consistently outperforms state-of-the-art baselines and substantially narrows the gap to the strongest agent, while using it for fewer than 25% of total calls.", "authors": ["Wenzhi Fang", "Liangqi Yuan", "Guangchen Lan", "Dong-Jun Han", "Christopher G. Brinton"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.08686", "pdf_url": "https://arxiv.org/pdf/2605.08686v1", "arxiv_id": "2605.08686", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3f49fe069af76a509b67d5e3e2faeeeceae2da2e54f0873735552e56236cc0ee", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond the All-in-One Agent: Benchmarking Role-Specialized Multi-Agent Collaboration in Enterprise Workflows", "abstract": "Large language model (LLM) agents are increasingly expected to operate in enterprise environments, where work is distributed across specialized roles, permission-controlled systems, and cross-departmental procedures. However, existing enterprise benchmarks largely evaluate single agents with broad tool access, while existing multi-agent benchmarks rarely capture realistic enterprise constraints such as role specialization, access control, stateful business systems, and policy-based approvals. We introduce \\textsc{EntCollabBench}, a benchmark for evaluating enterprise multi-agent collaboration. \\textsc{EntCollabBench} simulates a permission-isolated organization with 11 role-specialized agents across six departments and contains two evaluation subsets: a Workflow subset, where agents collaboratively modify enterprise system states, and an Approval subset, where agents make policy-grounded decisions. Evaluation is based on execution traces, database state verification, and deterministic policy adjudication rather than natural-language response judging. Experiments with representative LLM agents show that current models still struggle with end-to-end enterprise collaboration, especially in delegation, context transfer, parameter grounding, workflow closure, and decision commitment. \\textsc{EntCollabBench} provides a reproducible testbed for measuring and improving agent systems intended for realistic organizational environments.", "authors": ["Tao Yu", "Hao Wang", "Changyu Li", "Shenghua Chai", "Minghui Zhang", "Zhongtian Luo", "Yuxuan Zhou", "Haopeng Jin", "Zhaolu Kang", "Jiabing Yang", "YiFan Zhang", "Xinming Wang", "Hongzhu Yi", "Zheqi He", "Jing-Shu Zheng", "Xi Yang", "Yan Huang", "Liang Wang"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.08761", "pdf_url": "https://arxiv.org/pdf/2605.08761v1", "arxiv_id": "2605.08761", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "088ea5b6b4d83fb5696366f33443d3a40347cc377eddd82e22f87f1e74998962", "sources": ["arxiv", "semantic_scholar"], "title": "EvoMAS: Learning Execution-Time Workflows for Multi-Agent Systems", "abstract": "Large language model (LLM)-based multi-agent systems have shown strong potential on complex tasks through agent specialization, tool use, and collaborative reasoning. However, most automated multi-agent system design methods still follow a one-shot paradigm: a workflow is optimized or selected before execution and then reused unchanged throughout the task. This static coordination strategy is ill-suited for long-horizon tasks whose subgoals, intermediate evidence, and information needs evolve over multiple execution stages. We propose EvoMAS, a framework for execution-time multi-agent workflow construction. EvoMAS formulates workflow construction as a meta-level sequential decision problem along a single task trajectory. At each stage, it constructs an explicit task state through a Planner-Evaluator-Updater pipeline and uses a learned Workflow Adapter to instantiate a stage-specific layered workflow from a fixed pool of candidate agents. The adapter is trained with policy gradients using sparse, verifiable terminal task success as the main supervision signal, while evaluator-based process reward is analyzed separately under very-hard sparse-reward settings. Experiments on GAIA, HLE, and DeepResearcher show that EvoMAS outperforms single-agent baselines and recent automated multi-agent workflow design methods. Our analyses further show that explicit task-state construction and learned workflow adaptation provide complementary benefits. Additional results indicate that process reward is most useful when terminal success is extremely sparse, and qualitative case studies illustrate that EvoMAS adapts agent coordination as the task state evolves.", "authors": ["Chengdong Xu", "Kaiqiang Ke", "Ziheng Liu", "Jiaqi Wei", "Zibo Shao", "Weile Guo", "Chao Yu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.08769", "pdf_url": "https://arxiv.org/pdf/2605.08769v1", "arxiv_id": "2605.08769", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e001b789103572eaa0f0290e444163e6bd5218d0c3cedb556103089c05273d29", "sources": ["arxiv", "semantic_scholar"], "title": "AgentPSO: Evolving Agent Reasoning Skill via Multi-agent Particle Swarm Optimization", "abstract": "Multi-agent reasoning has shown promise for improving the problem-solving ability of large language models by allowing multiple agents to explore diverse reasoning paths. However, most existing multi-agent methods rely on inference-time debate or aggregation, which can be vulnerable to incorrect peer influence and biased consensus. Moreover, the agents themselves remain static, as their underlying reasoning skills do not evolve across tasks. In this paper, we introduce AgentPSO, a particle-swarm-inspired framework for evolving multi-agent reasoning skills. AgentPSO treats each agent as a particle-like reasoner whose state is a natural-language skill and whose velocity is a semantic update direction, iteratively moving agents toward stronger skill states to improve both individual and collective reasoning performance. Across training iterations, each agent updates its skill by combining its previous velocity, personal-best skill, global-best skill, and a self-reflective direction derived from peer reasoning trajectories. This enables agents to learn reusable reasoning behaviors from both their own experiences and the strongest skills discovered by the population, without updating the parameters of the backbone language model. Experiments on mathematical and general reasoning benchmarks show that AgentPSO improves over static single-agent skills and test-time-only multi-agent reasoning baselines. The evolved skills further transfer across benchmarks and to another backbone model, suggesting that AgentPSO captures reusable reasoning procedures rather than merely optimizing benchmark-specific prompts. Code is open-sourced at https://github.com/HYUNMIN-HWANG/AgentPSO/.", "authors": ["Hyunmin Hwang", "Jaemin Kim", "Choonghan Kim", "Hangeol Chang", "Jong Chul Ye"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-09", "url": "https://arxiv.org/abs/2605.08704", "pdf_url": "https://arxiv.org/pdf/2605.08704v1", "arxiv_id": "2605.08704", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/HYUNMIN-HWANG/AgentPSO/", "venue": null, "quality_score": 0.65}
{"id": "858b20f7ca653ee08ed9fff712e2b3a636fd60be41f2bf0ad8ec274c3ed23958", "sources": ["arxiv", "semantic_scholar"], "title": "Insider Attacks in Multi-Agent LLM Consensus Systems", "abstract": "Large language models (LLMs) are increasingly deployed in multi-agent systems where agents communicate in natural language to solve tasks jointly. A key capability in such systems is consensus formation, where agents iteratively exchange messages and update decisions to reach a shared outcome. However, most existing multi-agent LLM frameworks assume that all participating agents are aligned with the system objective. In practice, a malicious insider may participate as a legitimate member of the group while pursuing a hidden adversarial goal. In this work, we study insider manipulation in multi-agent LLM consensus systems. We formalize the problem as a sequential decision-making task in which a malicious agent seeks to delay or prevent agreement among benign agents. To make attack optimization tractable, we propose a world-model-based framework that learns surrogate dynamics over the latent behavioral states of benign agents and then trains an attacker using reinforcement learning based on this learned model. Preliminary results show that the trained attacker reduces the benign consensus rate and prolongs disagreement more effectively than the direct malicious-prompt baseline. These results suggest that combining latent world models with reinforcement learning is a promising direction for adaptive insider attacks in language-based multi-agent systems.", "authors": ["Xiaolin Sun", "Zixuan Liu", "Yibin Hu", "Zizhan Zheng"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.08268", "pdf_url": "https://arxiv.org/pdf/2605.08268v1", "arxiv_id": "2605.08268", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c27c914eeff902f22ef61407be43d6180b29dce257679415cee2d0ac2ae1c0f5", "sources": ["arxiv", "semantic_scholar"], "title": "OrchJail: Jailbreaking Tool-Calling Text-to-Image Agents by Orchestration-Guided Fuzzing", "abstract": "Tool-calling text-to-image (T2I) agents can plan and execute multi-step tool chains to accomplish complex generation and editing queries. However, this capability introduces a new safety attack surface: harmful outputs may arise from tool orchestration, where individually benign steps combine into unsafe results, making prompt-only jailbreak techniques insufficient. We present OrchJail, an orchestration-guided fuzzing framework for jailbreaking tool-calling T2I agents. Its core idea is to exploit high-risk tool-orchestration patterns: by learning from successful jailbreak tool-calling traces and their causal relationships to prompt wording, OrchJail directly guides the fuzzing search toward prompts that are more likely to trigger unsafe multi-step tool behaviors, rather than relying on surface-level textual perturbations. Extensive experiments demonstrate that OrchJail improves jailbreak effectiveness and efficiency across representative toolcalling T2I agents, achieving higher attack success rates, better image fidelity, and lower query costs, while remaining robust against common jailbreak defenses. Our work highlights tool orchestration as a critical, previously unexplored attack surface and provides a novel framework for uncovering safety risks in T2I agents.", "authors": ["Jianming Chen", "Yawen Wang", "Junjie Wang", "Zhe Liu", "Qing Wang", "Fanjiang Xu"], "categories": ["cs.MA", "cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.07414", "pdf_url": "https://arxiv.org/pdf/2605.07414v1", "arxiv_id": "2605.07414", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "ICML 2026", "quality_score": 0.55}
{"id": "c9b59a8a5e974825379d58bb4cc14231554f7c8bbcd2c535e8e0ac0e59e59a4a", "sources": ["arxiv", "semantic_scholar"], "title": "Do Agents Need to Plan Step-by-Step? Rethinking Planning Horizon in Data-Centric Tool Calling", "abstract": "Explicit planning is a critical capability for LLM-based agents solving complex data-centric tasks, which require precise tool calling over external data sources. Existing strategies fall into two paradigms based on planning horizon: (1) full-horizon (FH), which generates a complete plan before execution, and (2) single-step horizon (SH), which interleaves each action (tool call) with incremental reasoning and observation. While step-by-step execution is a common default under the assumption that eager execution monitoring is necessary for adaptability, we revisit this assumption for well-defined data-centric tasks. Our controlled empirical study isolates planning horizon as the key architectural feature and systematically analyzes the effects of topological complexity and tool robustness on both paradigms. Our experiments across Knowledge Base Question Answering and Multi-hop QA show that FH planning with lazy replanning achieves accuracy parity with SH across varying depths, breadths, and robustness levels, while using 2-3x fewer tokens. These findings suggest that for well-defined data-centric tasks, eager step-wise monitoring is often unnecessary, and full-horizon planning with on-demand replanning can offer a more efficient default.", "authors": ["Naoki Otani", "Nikita Bhutani", "Hannah Kim", "Dan Zhang", "Estevam Hruschka"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.08477", "pdf_url": "https://arxiv.org/pdf/2605.08477v1", "arxiv_id": "2605.08477", "doi": "10.1145/3786335.3813129", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f8e3d7957d1cc653d7faef7523626ce06ebd6f84b596294792501b978e5cc843", "sources": ["arxiv", "semantic_scholar"], "title": "Switchcraft: AI Model Router for Agentic Tool Calling", "abstract": "Agentic AI systems that invoke external tools are powerful but costly, leading developers to default to large models and overspend inference budgets. Model routing can mitigate this, but existing routers are designed for chat completion rather than tool use. We present Switchcraft, the first (to the best of our knowledge) model router optimized for agentic tool calling. Switchcraft operates inline, selecting the lowest-cost model subject to correctness. We construct an evaluation framework on five function-calling benchmarks and train a DistilBERT-based classifier, deployed under a latency budget. Switchcraft achieves 82.9% accuracy -- matching or exceeding the best individual model -- while reducing inference cost by 84%, saving over $3,600 per million queries. We find that larger models do not consistently outperform smaller ones on tool-use tasks, and that nominally cheaper models can incur higher total cost due to token-intensive reasoning. Our work enables cost-aware agentic AI deployment without sacrificing correctness.", "authors": ["Sharad Agarwal", "Pooria Namyar", "Alec Wolman", "Rahul Ambavat", "Ankur Gupta", "Qizheng Zhang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.07112", "pdf_url": "https://arxiv.org/pdf/2605.07112v1", "arxiv_id": "2605.07112", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "aedab8da480a11ae5a9d5537d8e589fa4220ed038a555eda35a4797cfd185d80", "sources": ["arxiv", "semantic_scholar"], "title": "When Child Inherits: Modeling and Exploiting Subagent Spawn in Multi-Agent Networks", "abstract": "Since the official release of ChatGPT in 2022, large language models (LLMs) have rapidly evolved from chatbot-style interfaces into agentic systems that can delegate work through tools and newly spawned subagents. While these capabilities improve automation and scalability, they also pose new security risks in multi-agent networks. Existing research has studied how individual LLM-based agents can be compromised through prompt injection, jailbreaking, poisoned retrieval data, or malicious extensions. Less is known about what happens after one agent is compromised inside a multi-agent network. In particular, inherited memory from parent agents can carry malicious instructions, outdated states, or unintended behavioral rules into newly created subagents, allowing a local compromise to spread across agent boundaries. In this paper, we model contemporary multi-agent networks through the lens of subagent inheritance. Our analysis shows that current frameworks can violate trust boundaries through insecure memory inheritance, weak resource control, stale post-spawn state, and improper termination authority. We demonstrate these risks in real agent frameworks and propose defenses based on explicit security invariants. Our findings show that inheritance is not merely an implementation detail, but a central component influencing the security of multi-agent systems.", "authors": ["Ziwen Cai", "Yihe Zhang", "Xiali Hei"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.08460", "pdf_url": "https://arxiv.org/pdf/2605.08460v1", "arxiv_id": "2605.08460", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1f9a6a078a41bec8f3ac17d52c99ef2577b0a6acfb8c72570dbba35cbcd7ffb4", "sources": ["arxiv", "semantic_scholar"], "title": "MASPrism: Lightweight Failure Attribution for Multi-Agent Systems Using Prefill-Stage Signals", "abstract": "Failure attribution in LLM-based multi-agent systems aims to identify the steps that contribute to a failed execution. This task remains difficult because a single execution can contain many agent actions and tool calls, failure evidence can appear many steps after the original mistake, and existing methods often rely on costly agent workflows, replay, or training on synthetic failure logs. To address these challenges, we propose MASPrism, a lightweight framework that performs failure attribution using prefill-stage signals from a small language model (SLM). MASPrism first extracts token-level negative log-likelihood and attention weights during a prefill pass to identify symptom-like steps and earlier candidate sources, without decoding. It then reconstructs a focused diagnostic prompt and performs a second prefill pass to rank failure-source candidates. Using Qwen3-0.6B as the SLM, MASPrism achieves the best performance on three of the four evaluated subsets across Who&When and TRAIL, improving Top-1 accuracy on Who&When-HC by 33.41% over the best baseline. On TRAIL, MASPrism outperforms strong proprietary LLMs, including Gemini-2.5-Pro, with up to 89.50% relative improvement. MASPrism processes each trace in 2.66 seconds on average, achieving a 6.69$\\times$ speedup over the single-pass prompting baseline, with zero output tokens. These results show that MASPrism provides an effective and practical framework for failure attribution in long multi-agent execution logs.", "authors": ["Yang Liu", "Hongjiang Feng", "Junsong Pu", "Zhuangbin Chen"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-08", "url": "https://arxiv.org/abs/2605.07509", "pdf_url": "https://arxiv.org/pdf/2605.07509v2", "arxiv_id": "2605.07509", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "af66b865ac4131498f97bf55624e1777aa5d1dd13c733f754243a94c5d1178a1", "sources": ["arxiv", "semantic_scholar"], "title": "Designing Intelligent Enterprise Agents: A Capability-Aligned Multi-Agent Architecture", "abstract": "Enterprise interest in multi-agent systems has shifted from generic software agents to large-language-model (LLM) based intelligent agents that plan, use tools, maintain contextual memory, inspect intermediate results, collaborate with other agents, and sometimes act in systems of record. This paper revises the enterprise architecture thesis around a design-first claim: governance is necessary, but it cannot be the primary organizing abstraction. The primary abstraction must be agent design - capability boundaries, autonomy allocation, interaction protocols, tool and data authority, state and memory design, verification design, and human interaction design. We propose CEAD (Capability-Aligned Enterprise Agent Design), a reference architecture for intelligent agents that uses service-oriented architecture (SOA) as an exemplar for contracts, registries, loose coupling, and policy-aware integration, while explicitly rejecting the idea that services are agents. It treats microservices as a cautionary precedent: decomposition without design discipline produces distributed complexity, cost, operational fragility, and agent proliferation. We evaluate CEAD over 10,000 enterprise tasks, comparing five architectures: a prompt-first mono-agent, a role-based micro-agent swarm, SOA-brokered agents, a governance-first but design-poor agent grid, and the proposed CEAD architecture. CEAD achieves 70.6% safe success, versus 45.2% for the mono-agent baseline, 23.1% for the ungoverned micro-agent swarm, 58.8% for SOA-brokered agents, and 50.8% for the control-heavy, design-poor grid. The results support the conclusion that design quality is the first-order enterprise concern; governance, security, policy, audit, and assurance should support and enforce good design rather than substitute for it.", "authors": ["John deVadoss"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.08258", "pdf_url": "https://arxiv.org/pdf/2605.08258v1", "arxiv_id": "2605.08258", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "736f53aa911ba078da69934e4d1441ed69763aac9b91488ced5774b4926827c4", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond the Black Box: Interpretability of Agentic AI Tool Use", "abstract": "AI agents are promising for high-stakes enterprise workflows, but dependable deployment remains limited because tool-use failures are difficult to diagnose and control. Agents may skip required tool calls, invoke tools unnecessarily, or take actions whose consequence becomes visible only after execution. Existing observability methods are external: prompts reveal correlations, evaluations score outputs, and logs arrive only after the model has already acted. In long-horizon settings, these failures are costly because an early tool mistake can alter the rest of the trajectory, increase token consumption, and create downstream safety and security risk. We introduce a mechanistic-interpretability toolkit built on Sparse Autoencoders (SAEs), which decompose activations into sparse internal features, and linear probes, lightweight classifiers that read signals from those features. The framework reads model states before each action and infers whether a tool is needed and how risky the next tool action is. It identifies the model layers and features most associated with tool decisions and tests their functional importance through feature ablation. We train the probes on multi-step trajectories from the NVIDIA Nemotron function-calling dataset and apply the same workflow to GPT-OSS 20B and Gemma 3 27B models. The goal is not to replace external evaluation, but to add a missing layer: visibility into what the model signaled internally before action. This helps surface deeper causes of agent failure, especially in long-horizon runs where an early mistake can impact subsequent agent behavior. More broadly, the paper shows how mechanistic interpretability can support internal observability for monitoring tool calls and risk in agent systems.", "authors": ["Hariom Tatsat", "Ariye Shater"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06890", "pdf_url": "https://arxiv.org/pdf/2605.06890v3", "arxiv_id": "2605.06890", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3c530d338dc537427335918fb544198e21f2a7fa96fe6bea5216fd0a3564a2ed", "sources": ["arxiv", "semantic_scholar"], "title": "AgenticPrecoding: LLM-Empowered Multi-Agent System for Precoding Optimization", "abstract": "Precoding is a key technique for interference management and performance improvement in multi-antenna wireless systems. However, existing precoding methods are typically developed for specific system models, objectives, and constraint sets, which limits their adaptability to the heterogeneous and evolving scenarios expected in future 6G networks. To address this limitation, we propose AgenticPrecoding, a universal multi-agent framework that automates end-to-end precoding derivation directly from user-level communication requirements. Specifically, AgenticPrecoding decomposes the derivation process into four coordinated stages: problem formulation, solver selection, prompt upsampling, and code generation, assigning each stage to a specialized agent tailored to its specific reasoning demands. We employ two LoRA-adapted reasoning agents to inject precoding-specific domain knowledge for problem formulation and solver selection, while two general-purpose Large Language Models (LLMs) handle prompt refinement and executable code generation. Furthermore, a feedback-driven refinement mechanism is incorporated to enhance code executability, constraint feasibility, and solution quality. Extensive experiments across 10 representative precoding scenarios demonstrate that AgenticPrecoding achieves superior cross-scenario adaptability compared to conventional optimization-based and LLM-based baselines.", "authors": ["Zijiu Yang", "Zixiang Zhang", "Shunpu Tang", "Qianqian Yang", "Zhiguo Shi"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06443", "pdf_url": "https://arxiv.org/pdf/2605.06443v1", "arxiv_id": "2605.06443", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f20beb4bb6552c43b6995905216b6aee7054d79a4ff11de1ebe39a9b4d8f79ec", "sources": ["arxiv", "semantic_scholar"], "title": "MANTRA: Synthesizing SMT-Validated Compliance Benchmarks for Tool-Using LLM Agents", "abstract": "Tool-using large language model (LLM) agents are increasingly deployed in settings where their reliable behavior is governed by strict procedural manuals. Ensuring that such agents comply with the rules from these manuals is challenging, as they are typically written for humans in natural language while agent behavior manifests as an execution trace of tool calls. Existing evaluations of LLM agents rely on manually constructed benchmarks or LLM-based judges, which either do not scale or lack reliability for complex, long-horizon manuals. To overcome these limitations, we present MANTRA, a framework for automatically synthesizing machine-checkable compliance benchmarks from natural-language manuals and tool schemas. MANTRA independently generates (i) a symbolic world model capturing procedural dependencies, and (ii) a set of trace-level compliance checks for a given task, and validates their consistency using SMT solving. A structured repair loop resolves inconsistencies, requiring human intervention only as a fallback. %This yields benchmarks that are formally validated. Importantly, MANTRA supports arbitrary domains and long procedural manuals, and provides a tunable notion of task complexity which is utilized to automatically derive challenging tasks accompanying compliance checks. Using MANTRA, we build a new benchmark suite with 285 tasks across 6 domains scaling to 50+ page manuals with minimal human effort. Empirically, we show that the compliance checks are richer with stronger constraint enforcement compared to existing benchmarks. Additionally, the granularity of the checks can be used for debugging the agents' failure modes. These results demonstrate that combining automated benchmark generation with formally grounded validation methods enables scalable and reliable benchmarking of tool-using agents.", "authors": ["Ashwani Anand", "Ivi Chatzi", "Ritam Raha", "Anne-Kathrin Schmuck"], "categories": ["cs.CL", "cs.LG", "cs.LO"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06334", "pdf_url": "https://arxiv.org/pdf/2605.06334v1", "arxiv_id": "2605.06334", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f1a866f378e5eb661586d3ddae2c156b7b17b83f107ecf4eef0ee149db16185f", "sources": ["arxiv", "semantic_scholar"], "title": "MASPO: Joint Prompt Optimization for LLM-based Multi-Agent Systems", "abstract": "Large language model (LLM)-based Multi-agent systems (MAS) have shown promise in tackling complex collaborative tasks, where agents are typically orchestrated via role-specific prompts. While the quality of these prompts is pivotal, jointly optimizing them across interacting agents remains a non-trivial challenge, primarily due to the misalignment between local agent objectives and holistic system goals. To address this, we introduce MASPO, a novel framework designed to automatically and iteratively refine prompts across the entire system. A core innovation of MASPO is its joint evaluation mechanism, which assesses prompts not merely by their local validity, but by their capacity to facilitate downstream success for successor agents. This effectively bridges the gap between local interactions and global outcomes without relying on ground-truth labels. Furthermore, MASPO employs a data-driven evolutionary beam search to efficiently navigate the high-dimensional prompt space. Extensive empirical evaluations across 6 diverse tasks demonstrate that MASPO consistently outperforms state-of-the-art prompt optimization methods, achieving an average accuracy improvement of 2.9. We release our code at https://github.com/wangzx1219/MASPO.", "authors": ["Zhexuan Wang", "Xuebo Liu", "Li Wang", "Zifei Shan", "Yutong Wang", "Zhenxi Song", "Min Zhang"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06623", "pdf_url": "https://arxiv.org/pdf/2605.06623v1", "arxiv_id": "2605.06623", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wangzx1219/MASPO", "venue": null, "quality_score": 0.65}
{"id": "20ac76aeeb3dae4e3307d84a6128e547a81decd1a0684ffd54ef5d654b8d78d5", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Security-Auditable LLM Agents: A Unified Graph Representation", "abstract": "LLM-based agentic systems are rapidly evolving to perform complex autonomous tasks through dynamic tool invocation, stateful memory management, and multi-agent collaboration. However, this semantics-driven execution paradigm creates a severe semantic gap between low-level physical events and high-level execution intent, making post-hoc security auditing fundamentally difficult. Existing representation mechanisms, including static SBOMs and runtime logs, provide only fragmented evidence and fail to capture cognitive-state evolution, capability bindings, persistent memory contamination, and cascading risk propagation across interacting agents. To bridge this gap, we propose Agent-BOM, a unified structural representation for agent security auditing. Agent-BOM models an agentic system as a hierarchical attributed directed graph that separates static capability bases, such as models, tools, and long-term memory, from dynamic runtime semantic states, such as goals, reasoning trajectories, and actions. These layers are connected through semantic edges and security attributes, transforming fragmented execution traces into queryable audit paths. Building on Agent-BOM, we develop a graph-query-based paradigm for path-level risk assessment and instantiate it with the OWASP Agentic Top 10. We further implement an auditing plugin in the OpenClaw environment to construct Agent-BOM from live executions. Evaluation on representative real-world agentic attack scenarios shows that Agent-BOM can reconstruct stealthy attack chains, including cross-session memory poisoning and tool misuse, capability supply-chain hijacking and unexpected code execution, multi-agent ecosystem hijacking, and privilege and trust abuse. These results demonstrate that Agent-BOM provides a unified and auditable foundation for root-cause analysis and security adjudication in complex agentic ecosystems.", "authors": ["Chaofan Li", "Lyuye Zhang", "Jintao Zhai", "Siyue Feng", "Xichun Yang", "Huahao Wang", "Shihan Dou", "Yu Ji", "Yutao Hu", "Yueming Wu", "Yang Liu", "Deqing Zou"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06812", "pdf_url": "https://arxiv.org/pdf/2605.06812v1", "arxiv_id": "2605.06812", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a9f973d14351bc6d65e95c3e0767d59066d0f9168aba8b0daa950076a19cb7dd", "sources": ["arxiv", "semantic_scholar"], "title": "A Self-Healing Framework for Reliable LLM-Based Autonomous Agents", "abstract": "Autonomous agents based on Large Language Models (LLMs) are increasingly being utilized in complex software systems. However, reliability remains a significant challenge due to unpredictable failures such as hallucinations, execution errors, and inconsistent reasoning. This paper proposes a reliability-aware self-healing framework for LLM-based software agents. The framework integrates failure detection, reliability assessment, and automated recovery mechanisms. First, we define a taxonomy of failure types and introduce a quantitative reliability assessment model. Next, we propose a failure detection method that identifies abnormal agent behavior based on execution patterns and output consistency. Finally, we design a self-healing mechanism that dynamically recovers from failures through adaptive replanning and corrective prompting strategies. The proposed framework was implemented in a multi-agent workflow environment and evaluated using real-world task scenarios. Experimental results demonstrate that our approach significantly increases task success rates, reduces failure propagation, and enhances overall system robustness compared to existing methods. In particular, this study distinguishes itself by establishing an integrated monitoring system that combines the agent's internal reasoning process with external execution results. These findings are expected to contribute to securing the stability of advanced autonomous systems and lowering the barriers to LLM adoption in production environments.", "authors": ["Cheonsu Jeong", "Younggun Shin"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.06737", "pdf_url": "https://arxiv.org/pdf/2605.06737v1", "arxiv_id": "2605.06737", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3d54e3b6ed1ae3d00206a55ee7284e7c5d2647d52142f389377c3f67cb0abaac", "sources": ["arxiv", "semantic_scholar"], "title": "Active Learning for Communication Structure Optimization in LLM-Based Multi-Agent Systems", "abstract": "Optimizing the communication structure of large language model based multi-agent systems (LLM-MAS) has been shown to improve downstream performance and reduce token usage. Existing methods typically rely on randomly sampled training tasks. However, tasks may differ substantially in difficulty and domain, and thus they are not equally informative for updating communication structure, making optimization under limited training budgets often unstable and highly sensitive to the particular training set. To actively identify the most valuable tasks for communication-structure optimization, we propose an ensemble-based information-theoretic task selection framework. The proposed method estimates task informativeness by how much a candidate task changes the distribution over graph parameters, using ensemble Kalman inversion as an efficient and derivative-free approximation of the corresponding Bayesian update. The resulting estimator is especially suitable for black-box and noisy multi-agent systems. To enhance scalability, we construct a compact candidate pool through embedding-based representative selection and combine the informative selection with surrogate modeling and batch Thompson sampling. We validate our method in both benign settings and settings with agent attacks, demonstrating its effectiveness for communication-structure optimization under constrained computational budgets.", "authors": ["Huchen Yang", "Xinghao Dong", "Dan Negrut", "Jin-Long Wu"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-07", "url": "https://arxiv.org/abs/2605.05703", "pdf_url": "https://arxiv.org/pdf/2605.05703v2", "arxiv_id": "2605.05703", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9a4cde77ec3dd676758cdd0d50b62bbb05d28df120647d91081b91c81773855a", "sources": ["arxiv", "semantic_scholar"], "title": "GEM: Graph-Enhanced Mixture-of-Experts with ReAct Agents for Dialogue State Tracking", "abstract": "Dialogue State Tracking (DST) requires precise extraction of structured information from multi-domain conversations, a task where Large Language Models (LLMs) struggle despite their impressive general capabilities. We present GEM (Graph-Enhanced Mixture-of-Experts), a novel framework that combines language models and graph-structured dialogue understanding with ReAct agent-based reasoning for superior DST performance. Our approach dynamically routes between specialized experts: a Graph Neural Network that captures dialogue structure and turn-level dependencies, and a finetuned T5-Small encoder-decoder for sequence modeling, coordinated by an intelligent router. For complex value generation tasks, we integrate ReAct agents that perform structured reasoning over dialogue context. On MultiWOZ 2.2, GEM achieves 65.19% Joint Goal Accuracy, substantially outperforming end-to-end LLM approaches (best: 38.43%) and surpassing state-of-the-art (SOTA) methods including TOATOD (63.79%), D3ST (58.70%), and Diable (56.48%). Our graph-enhanced mixture-of-experts architecture with ReAct integration demonstrates that combining structured dialogue representation with dynamic expert routing and agent-based reasoning provides a powerful paradigm for dialogue state tracking, achieving superior accuracy while maintaining computational efficiency through selective expert activation.", "authors": ["Ziqi Zhu", "Adithya Suresh", "Tomal Deb", "Iman Abbasnejad"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-06", "url": "https://arxiv.org/abs/2605.04449", "pdf_url": "https://arxiv.org/pdf/2605.04449v1", "arxiv_id": "2605.04449", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4ebb27d505e6809ee41dd63af486c8ffe20378ea93840108d8b40380cc5b8272", "sources": ["arxiv", "semantic_scholar"], "title": "Strat-Reasoner: Reinforcing Strategic Reasoning of LLMs in Multi-Agent Games", "abstract": "While Large Language Models (LLMs) excel in certain reasoning tasks, they struggle in multi-agent games where the final outcome depends on the joint strategies of all agents. In multi-agent games, the non-stationarity of other agents brings significant challenges on the evaluation of the reasoning process and the credit assignment over multiple reasoning steps. Existing single-agent reinforcement learning (RL) approaches and their multi-agent extensions fail to address these challenges as they do not incorporate other agents in the reasoning process. In this work, we propose Strat-Reasoner, a novel RL-based framework that improves LLMs' strategic reasoning ability in multi-agent games. We introduce a novel recursive reasoning paradigm where an agent's reasoning also integrates other agents' reasoning processes. To provide effective reward signals for the intermediate reasoning sequences, we employ a centralized Chain-of-Thought (CoT) comparison module to evaluate the reasoning quality. Finally, we compute an accurate hybrid advantage and develop a group-relative RL approach to optimize the LLM policy. Experimental results show that Strat-Reasoner substantially improves strategic abilities of underlying LLMs, achieving 22.1\\% average performance improvements across various multi-agent games. Code is publicly available at https://github.com/ydhe1012/Strat-Reasoner.", "authors": ["Yidong He", "Yutao Lai", "Pengxu Yang", "Jiarui Gan", "Jiexin Wang", "Yi Cai", "Mengchen Zhao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-06", "url": "https://arxiv.org/abs/2605.04906", "pdf_url": "https://arxiv.org/pdf/2605.04906v2", "arxiv_id": "2605.04906", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ydhe1012/Strat-Reasoner", "venue": null, "quality_score": 0.65}
{"id": "5e3eb0748d8569f7c85916537b112d358e08cf73a5c9775c1734ef8a61dca37b", "sources": ["arxiv", "semantic_scholar"], "title": "SensingAgents: A Multi-Agent Collaborative Framework for Robust IMU Activity Recognition", "abstract": "Human Activity Recognition (HAR) using Inertial Measurement Unit (IMU) sensors is a cornerstone of mobile health, smart environments, and human-computer interaction. However, current deep learning-based HAR models often struggle with heavy reliance on labeled data, position-specific ambiguity, and a lack of transparent reasoning. Inspired by the advanced agents framework, which emulates a collaborative agent using Large Language Models (LLMs), we propose SensingAgents, a novel multi-agent system for robust IMU activity recognition. SensingAgents organizes LLM-powered agents into specialized roles: a group of Analyst Agents for position-specific sensor analysis (arm, wrist, belt, pocket), a pair of Advocate Agents that resolves sensor conflicts through dynamic and static dialectical debates, and a Decision Agent that ensures reliability under sensor drift or failure. Evaluation on the Shoaib dataset demonstrates that SensingAgents significantly outperforms state-of-the-art single-agent and multi-agent LLM models, achieving an accuracy of 79.5% in a zero setting--29% higher than existing agent models and 9.4% higher than deep learning baselines--particularly in complex scenarios where multi-sensor data is conflicting or noisy. Our work highlights the potential of multi-agent collaborative reasoning for advancing the robustness and interpretability of ubiquitous sensing systems.", "authors": ["Naiyu Zheng", "Tianlong Yu", "Haochen Yin", "Xiaoyi Fan", "Xiping Hu", "Zhimeng Yin"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-06", "url": "https://arxiv.org/abs/2605.04608", "pdf_url": "https://arxiv.org/pdf/2605.04608v1", "arxiv_id": "2605.04608", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "80e92f9ff480a913b8ca1e13d02e84747febb5f65c93c6f9eb67ecaa6924dd7f", "sources": ["arxiv", "semantic_scholar"], "title": "Tree-based Credit Assignment for Multi-Agent Memory System", "abstract": "Memory systems are widely adopted to enhance LLMs for long-horizon tasks, and are commonly organized as multi-agent pipelines with memory building, summarizing, and retrieval agents. To empower this system, existing RL-based methods either apply final downstream task rewards (e.g., QA accuracy) for all agents uniformly, which are coarse and ambiguous, or design task-specific rewards for agents on different subtasks, which require costly annotations (e.g., key evidence) and are difficult to define reliably. To address these limitations, we propose Tree-based Credit Assignment for Multi-Agent Memory Systems (TreeMem), which derives agent-specific credit from the final reward without task-specific annotations. Specifically, TreeMem extends the multi-agent pipeline (builder--summarizer--retrieval) into a tree structure, where each agent's outputs are expanded into multiple subsequent branches. The contribution of each agent is estimated via Monte Carlo averaging over its subsequent branches, capturing how intermediate agent actions may influence the final reward. This converts the coarse final reward into agent-specific optimization signals. These signals are then used to update all agent policies simultaneously, helping heterogeneous agents specialize effectively. Experiments on long-horizon benchmarks show that TreeMem improves memory system performance over strong baselines, validating the effectiveness of tree-structured credit assignment for the multi-agent memory system.", "authors": ["Marina Mao", "Alexandr Liu", "Pengbo Li", "Siheng Li", "Bo Zhou", "Xiang Wang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-06", "url": "https://arxiv.org/abs/2605.04811", "pdf_url": "https://arxiv.org/pdf/2605.04811v1", "arxiv_id": "2605.04811", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "baf4e67c0a656e99c144e57cba9a1a14a1706d0322af1c617759f0bac1d12f2b", "sources": ["arxiv", "semantic_scholar"], "title": "AgentTrust: Runtime Safety Evaluation and Interception for AI Agent Tool Use", "abstract": "Modern AI agents execute real-world side effects through tool calls such as file operations, shell commands, HTTP requests, and database queries. A single unsafe action, including accidental deletion, credential exposure, or data exfiltration, can cause irreversible harm. Existing defenses are incomplete: post-hoc benchmarks measure behavior after execution, static guardrails miss obfuscation and multi-step context, and infrastructure sandboxes constrain where code runs without understanding what an action means. We present AgentTrust, a runtime safety layer that intercepts agent tool calls before execution and returns a structured verdict: allow, warn, block, or review. AgentTrust combines a shell deobfuscation normalizer, SafeFix suggestions for safer alternatives, RiskChain detection for multi-step attack chains, and a cache-aware LLM-as-Judge for ambiguous inputs. We release a 300-scenario benchmark across six risk categories and an additional 630 independently constructed real-world adversarial scenarios. On the internal benchmark, the production-only ruleset achieves 95.0% verdict accuracy and 73.7% risk-level accuracy at low-millisecond end-to-end latency. On the 630-scenario benchmark, evaluated under a patched ruleset and not claimed as zero-shot, AgentTrust achieves 96.7% verdict accuracy, including about 93% on shell-obfuscated payloads. AgentTrust is released under the AGPL-3.0 license and provides a Model Context Protocol server for MCP-compatible agents.", "authors": ["Chenglin Yang"], "categories": ["cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-06", "url": "https://arxiv.org/abs/2605.04785", "pdf_url": "https://arxiv.org/pdf/2605.04785v1", "arxiv_id": "2605.04785", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a35135d96838c6029392c3141c801349a6a859a7396aeebdc6452c64bc4f30f0", "sources": ["arxiv", "semantic_scholar"], "title": "When Context Hurts: The Crossover Effect of Knowledge Transfer on Multi-Agent Design Exploration", "abstract": "The prevailing assumption in agent orchestration is that more context is better. We test this on multi-agent software design across 10 tasks, 7 context-injection conditions, and over 2,700 runs, and find a crossover effect: the same artifact type improves design exploration on some tasks (up to 20$\\times$ tradeoff coverage) and actively degrades it on others (up to 46% reduction). On several tasks, an irrelevant document performs as well as or better than every relevant artifact. The direction is predicted by a single measurable variable--baseline exploration without context--with Pearson $r = -0.82$ ($p < 0.001$). Probing the mechanism by manipulating convergence pressure through prompt design reveals two distinct regimes: convergence driven by training data priors (natural) responds to artifact disruption, while convergence driven by explicit instructions (induced) does not. The implication is that context injection should be conditional, not universal: one no-context trial is a cheap diagnostic that predicts whether knowledge artifacts will help or hurt a given task.", "authors": ["Saranyan Vigraham"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-05", "url": "https://arxiv.org/abs/2605.04361", "pdf_url": "https://arxiv.org/pdf/2605.04361v1", "arxiv_id": "2605.04361", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "099465807ae3b25a77d39fd504f693d15c499ab325d8072de4b76328e157a19d", "sources": ["arxiv", "semantic_scholar"], "title": "Coordination as an Architectural Layer for LLM-Based Multi-Agent Systems", "abstract": "Multi-agent LLM systems fail in production at rates between 41% and 87%, mostly due to coordination defects rather than base-model capability. Existing responses split between cataloguing failure modes empirically and shipping declarative orchestration frameworks as engineering tools; neither delivers a principled mapping from coordination configuration to predictable failure-mode signature. We argue that coordination should be treated as a configurable architectural layer, separable from agent logic and from information access, enabling architectural reasoning rather than only engineering productivity. We instantiate this with an information-controlled design on prediction markets: a single LLM, fixed tools, fixed per-call output cap, and fixed prompt template across five reference coordination configurations, with total compute per question treated as an endogenous architectural output. The Murphy decomposition of the Brier score separates calibration from discriminative power, so configurations leave distinguishable signatures even when aggregate scores coincide. On 100 Polymarket binary markets resolved after the model's training cutoff (claude-opus-4-6) we report Murphy signatures, a cost-quality Pareto frontier, category-conditioned analysis, and a bootstrap power-projection. Three of five pre-specified predictions are upheld in direction; two configurations dominate the Pareto frontier within this regime; exploratory bootstrap intervals separate consensus alignment from others, though pairwise tests do not survive Bonferroni correction at n=100. We also deploy the same configurations as live agents on Foresight Arena under web-search-enabled conditions, as an on-chain replication channel accumulating in parallel. Harness, trace dataset, and production agents are released. We position this as a methodology-validating first instantiation, not a general cross-model claim.", "authors": ["Maksym Nechepurenko", "Pavel Shuvalov"], "categories": ["cs.MA", "cs.LG", "q-fin.TR"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2026-05-05", "url": "https://arxiv.org/abs/2605.03310", "pdf_url": "https://arxiv.org/pdf/2605.03310v1", "arxiv_id": "2605.03310", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b92ccd965fdd943f633455368b8271e65d49230aa730315160994b7c5733fdae", "sources": ["arxiv", "semantic_scholar"], "title": "Governed Collaborative Memory as Artificial Selection in LLM-Based Multi-Agent Systems", "abstract": "Persistent memory is turning language-model-based agents from stateless participants in isolated interactions into state-bearing components of LLM-based multi-agent systems. As memory becomes durable, reloadable, and behavior-shaping across agents, sessions, or versions, a design question arises that is not captured by retrieval accuracy or access control alone: which candidate memories should become shared institutional state? This Viewpoint frames that problem as governed collaborative memory. We argue that memory governance functions as a selection regime, determining which memory variants persist, which remain private, and which are rejected, abstained from, or superseded. We distinguish ungoverned persistence, constitutional or hybrid selection, automatic metric-based selection, and human-ratified artificial selection, emphasizing that these regimes are not a ranking but a design choice over target properties. We then describe a layered architecture that separates agent-local memory, shared institutional memory, archive memory, and project-continuity memory, with provenance and version lineage making selection inspectable. Documented traces from one running LLM-based multi-agent ecosystem illustrate unmanaged false-memory persistence, ratified institutional memory, rejection and revision, identity-preserving expansion, and governance-as-learning. The contribution is a design agenda: persistent LLM-based multi-agent systems should evaluate memory not only for recall and performance, but also for provenance fidelity, selection traceability, epistemic quality, correction pathways, and role preservation.", "authors": ["Diego F. Cuadros", "Abdoul-Aziz Maiga", "Helen Meskhidze", "Andre Curtis-Trudel"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-05", "url": "https://arxiv.org/abs/2605.04264", "pdf_url": "https://arxiv.org/pdf/2605.04264v1", "arxiv_id": "2605.04264", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4d380f48858784ab31c76c494d0a51e682fc35a9057f9f41548e25aefd7ba605", "sources": ["arxiv", "semantic_scholar"], "title": "From Intent to Execution: Composing Agentic Workflows with Agent Recommendation", "abstract": "Multi-Agent Systems (MAS) built using AI agents fulfill a variety of user intents that may be used to design and build a family of related applications. However, the creation of such MAS currently involves manual composition of the plan, manual selection of appropriate agents, and manual creation of execution graphs. This paper introduces a framework for the automated creation of multi-agent systems which replaces multiple manual steps with an automated framework. The proposed framework consists of software modules and a workflow to orchestrate the requisite task- specific application. The modules include: an LLM-derived planner, a set of tasks described in natural language, a dynamic call graph, an orchestrator for map agents to tasks, and an agent recommender that finds the most suitable agent(s) from local and global agent registries. The agent recommender uses a two-stage information retrieval (IR) system comprising a fast retriever and an LLM-based re-ranker. We implemented a series of experiments exploring the choice of embedders, re- rankers, agent description enrichment, and supervising critique agent. We benchmarked this system end-to-end, evaluating the combination of planning, agent selection, and task completion, with our proposed approach. Our experimental results show that our approach outperforms the state-of-the- art in terms of the recall rate and is more robust and scalable compared to previous approaches. The critique agent holistically reevaluates both agent and tool recommendations against the overall plan. We show that the inclusion of the critique agent further enhances the recall score, proving that the comprehensive review and revision of task-based agent selection is an essential step in building end-to-end multi-agent systems.", "authors": ["Kishan Athrey", "Ramin Pishehvar", "Brian Riordan", "Mahesh Viswanathan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-05", "url": "https://arxiv.org/abs/2605.03986", "pdf_url": "https://arxiv.org/pdf/2605.03986v1", "arxiv_id": "2605.03986", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "99514b0799b7951081a25cf5b5fc1cbd70288ae0c125f59cf5b933ed4c43ae32", "sources": ["arxiv", "semantic_scholar"], "title": "SkCC: Portable and Secure Skill Compilation for Cross-Framework LLM Agents", "abstract": "LLM agents increasingly rely on reusable skills (e.g., SKILL markdown files) to execute complex tasks, yet these artifacts lack portability: agent frameworks are highly sensitive to prompt formatting, leading to a large performance variation for the same skill. Nevertheless, most skills are authored once as format-agnostic Markdown, necessitating costly per-framework rewrites and also leaving security largely unaddressed, with widespread vulnerabilities in practice. To address this, we present SkCC, a compiler for LLM agents that introduces classical compilation design into agent skill development. SkCC centers on SkIR, a strongly-typed intermediate representation that decouples skill semantics from framework-specific formatting, thus enabling portable deployment across agent frameworks. Atop of this IR, a static Optimizer enforces security constraints, blocking vulnerabilities before deployment. Implemented as a four-phase pipeline, SkCC effectively reduces adaptation complexity from $O(m \\times n)$ to $O(m + n)$ across $m$ skills and $n$ frameworks. Experiments on SkillsBench demonstrate that SkCC delivers consistent and substantial gains over original counterparts, with pass rate increases from 21.1% to 33.3% on Claude Code and from 35.1% to 48.7% on Kimi CLI. Further, the design achieves sub-10ms compilation latency, 94.8% proactive security trigger rate, and 10-46% runtime token savings across frameworks.", "authors": ["Yipeng Ouyang", "Yi Xiao", "Yuhao Gu", "Xianwei Zhang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-05", "url": "https://arxiv.org/abs/2605.03353", "pdf_url": "https://arxiv.org/pdf/2605.03353v4", "arxiv_id": "2605.03353", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Nexa-Language/Skill-Compiler/", "venue": null, "quality_score": 0.65}
{"id": "8901d6bd11c9d8dbfb4f87d8a3e3a80d55dfd27e7e0f39955cff48dadfa8efbb", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond State Machines: Executing Network Procedures with Agentic Tool-Calling Sequences", "abstract": "Agentic AI will be an essential enabling technology for designing future mobile communication systems, which could provide flexible and customized services, automate complex network operations, and drive autonomous decision-making across the network. This work studies how Large Language Model (LLM)-based network AI agents can be utilized to execute network procedures expressed as sequences of tool invocations. We investigate four approaches, which differ in how the agent obtains the procedure and in how execution is distributed between the agent and the underlying tools. We evaluated the latency and execution correctness across these approaches using a User Equipment (UE) IP allocation procedure as a case study. Furthermore, we conduct a stress test to examine how many sequential procedural steps an LLM agent can reliably execute before failure. Our results show that approaches relying on iterative agent-side reasoning incur higher latency and are more prone to execution errors, while approaches where the procedure is encapsulated within a single tool, which internally orchestrates the required steps by invoking other tools, reduce latency by limiting repeated reasoning. The stress-test results further show that the model with advanced tool-calling capability maintains reliable execution over longer procedures than the other evaluated models; however, all models exhibit reliability degradation as procedure length increases, revealing clear execution limits in multi-step tool-based workflows. To systematically analyze failures in procedure execution, we introduce a procedure-specific error taxonomy that categorizes deviations in multi-step procedural execution.", "authors": ["Purna Sai Garigipati", "Onur Ayan", "Kishor Chandra Joshi", "Xueli An"], "categories": ["cs.NI", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.02584", "pdf_url": "https://arxiv.org/pdf/2605.02584v1", "arxiv_id": "2605.02584", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d6846e4b52e1bc744b35f4b162ac5efefe94ff89db3da5f281f9d101c793a841", "sources": ["arxiv", "semantic_scholar"], "title": "Fair Agents: Balancing Multistakeholder Alignment in Multi-Agent Personalization Systems", "abstract": "LLM agents are increasingly used for personalization due to their ability to communicate directly with users in natural language, integrate external knowledge bases, and negotiate with other (possibly human) agents. Especially in multistakeholder AI systems with multiple distinct objectives, LLM agents are used to independently optimize for each stakeholder's goals. Here, stakeholder alignment is essential to identify and map these goals to provide LLM agents with quantifiable objectives. Plus, the way in which the outputs of the LLM agents are aggregated is fundamental to ensuring fair outcomes for all agents and, therefore, stakeholders. In this work, we identify open research challenges and propose a conceptual framework for designing fair multi-agent multistakeholder personalization systems that balance competing stakeholder objectives. Our framework integrates (i) methods to align stakeholder objectives and LLM agents, (ii) aggregation strategies, e.g., based on social choice theory, to form fair collective decisions, and (iii) stakeholder-centric evaluation procedures for both individual and collective agent behavior. We showcase our framework through a tourism use case and discuss possible applications in other domains, such as education and healthcare. Finally, we discuss domain-specific fairness tensions and review datasets for evaluating multistakeholder fairness and multi-agent personalization systems.", "authors": ["Andrea Forster", "Peter Müllner", "Denis Helic", "Elisabeth Lex", "Dominik Kowald"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.02379", "pdf_url": "https://arxiv.org/pdf/2605.02379v1", "arxiv_id": "2605.02379", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "7ba019b3683690fb36dd06e25167de9930a9ed42954b488fde59f25ffd6702a6", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcement Learning for LLM-based Multi-Agent Systems through Orchestration Traces", "abstract": "As large language model (LLM) agents evolve from isolated tool users into coordinated teams, reinforcement learning (RL) must optimize not only individual actions but also how work is spawned, delegated, communicated, aggregated, and stopped. This paper studies RL for LLM-based multi-agent systems through orchestration traces: temporal interaction graphs whose events include sub-agent spawning, delegation, communication, tool use, return, aggregation, and stopping decisions. Using this lens, we identify three technical axes. First, reward design spans eight families, including orchestration rewards for parallelism speedup, split correctness, and aggregation quality. Second, reward and credit signals attach to eight credit- or signal-bearing units from token to team; explicit counterfactual message-level credit remains especially sparse in our curated pool. Third, orchestration learning decomposes into five sub-decisions: when to spawn, whom to delegate to, how to communicate, how to aggregate, and when to stop. In our curated pool as of May 4, 2026, we found no explicit RL training method for the stopping decision. We connect academic methods to public industrial evidence from Kimi Agent Swarm, OpenAI Codex, and Anthropic Claude Code. The resulting scale gap is a gap between publicly reported deployment envelopes and open academic evaluation regimes, not independent verification of industrial training traces. We release the artifact at https://github.com/xxzcc/awesome-llm-mas-rl, including an 84-entry tagged paper pool, a 32-record exclusion log, scripted corpus statistics, and a minimal JSON schema for replayable orchestration traces.", "authors": ["Chenchen Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.02801", "pdf_url": "https://arxiv.org/pdf/2605.02801v1", "arxiv_id": "2605.02801", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/xxzcc/awesome-llm-mas-rl", "venue": null, "quality_score": 0.65}
{"id": "743e4d54ccc336eab5fd23bfbbee499a216e228ce32bef8fe21bd838830050b7", "sources": ["arxiv", "semantic_scholar"], "title": "Enwar 3.0: An Agentic Multi-Modal LLM Orchestrator for Situation-Aware Beamforming, Blockage Prediction, and Handover Management", "abstract": "Maintaining robust millimeter-wave (mmWave) connectivity in vehicular networks requires real-time adaptation to environmental dynamics, sensor degradation, and link variability. This paper presents Enwar 3.0, an environment-aware reasoning framework that unifies multi-modal sensing, agentic large language models (LLMs), and context-driven model selection for predictive beamforming, blockage detection, and handover management. Building upon prior iterations of Enwar, the proposed architecture integrates a classifier-driven assessment of sensor health with a primed LLM that orchestrates multiple specialized agents through structured, task-aware prompting. A novel synthetic degradation pipeline enables the training of a sensor degradation classifier that detects real-time impairments across camera, radar, LiDAR, and GPS inputs, achieving over 99% accuracy. The LLM, trained via chain-of-thought (CoT) priming and human-in-the-loop feedback, coordinates agent calls for beam selection, blockage forecasting, and environment perception while dynamically loading sensor-specific models based on environmental context. Extensive evaluations across 15 sensor combinations demonstrate that Enwar 3.0 delivers state-of-the-art performance in both predictive accuracy and interpretability, with beam selection accuracy exceeding 88%, blockage F1-scores surpassing 98%, and reasoning correctness reaching 87% on complex decision prompts. This work establishes a scalable foundation for LLM-integrated wireless systems that reason, perceive, and adapt in real-time.", "authors": ["Ahmad M. Nazar", "Abdulkadir Celik", "Asmaa Abdallah", "Mohamed Y. Selim", "Daji Qiao", "Ahmed M. Eltawil"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.03215", "pdf_url": "https://arxiv.org/pdf/2605.03215v1", "arxiv_id": "2605.03215", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "32f56458f92751112c8f3aae22c5be18c0f07f16ea6822a9ef4bf0093a1dfba1", "sources": ["arxiv", "semantic_scholar"], "title": "Autonomous LLM Agent Worms: Cross-Platform Propagation, Automated Discovery and Temporal Re-Entry Defense", "abstract": "Autonomous LLM agents operate as long-running processes with persistent workspaces, memory files, scheduled task state, and messaging integrations. These features create a new propagation risk: attacker-influenced content can be written into persistent agent state, re-enter the LLM decision context through scheduled autoloading, and drive high-risk actions including configuration changes and cross-agent transmission. We present the first systematic framework for automated analysis of persistent worm propagation in file-backed multi-agent LLM ecosystems. SSCGV, our automated source-code graph analyzer, traces data flow from file I/O to LLM context injection points and ranks carriers by context injection position without manual analysis. SRPO, our summary-resilient payload optimizer, generates worm payloads robust to LLM-mediated summarization and paraphrasing across multi-hop communication. Evaluated on three production agent frameworks, we demonstrate zero-click autonomous propagation, 3-hop cross-platform transmission without platform-specific adaptation, inter-agent privilege escalation, and data exfiltration. We identify two empirical insights: user prompt carriers achieve higher attack compliance than system prompt carriers, and read operations represent the primary integrity threat in LLM-mediated systems. To defend against this class of attacks, we develop RTW-A, proven under a formal No Persistent Worm Propagation theorem. RTW blocks write-before-exposed-read re-entry; sealed configuration protects static files; typed memory promotion prevents untrusted summaries from entering trusted memory; and capability attenuation limits high-risk actions after external reads. These mechanisms eliminate the persistence, re-entry, action chain while preserving ordinary workflows. Affected systems are anonymized pending coordinated disclosure.", "authors": ["Mingming Zha", "Xiaofeng Wang"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.02812", "pdf_url": "https://arxiv.org/pdf/2605.02812v1", "arxiv_id": "2605.02812", "doi": null, "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d332f5afe3cb9bac7d92f6949e9659f8462395dd1e65a0b39fd6d68adc34800c", "sources": ["arxiv", "semantic_scholar"], "title": "A Study of Belief Revision Postulates in Multi-Agent Systems (Extended Version)", "abstract": "We investigate the belief revision problem in epistemic planning, i.e., what will be the beliefs of all agents in a multi-agent system after an agent gains the belief in some state property. Based on the standard representation in epistemic planning of agents' beliefs via a single multi-agent Kripke model, we generalize the classical AGM belief revision postulates to the multi-agent setting, with the aim to provide a formal framework for evaluating dynamic epistemic reasoning frameworks in which the beliefs of all agents as the result of actions are computed. As an example of a simple operator that satisfies all of the generalized AGM postulates, we present generalized full-meet multi-agent belief revision. We moreover define a generalization of the standard postulates for iterated revision, present a more sophisticated, event model based revision operator, and discuss the potential issues in defining an epistemic operator on Kripke models that can satisfy all of the generalized postulates for iterated multi-agent belief revision.", "authors": ["Michael Thielscher", "Tran Cao Son"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-04", "url": "https://arxiv.org/abs/2605.02249", "pdf_url": "https://arxiv.org/pdf/2605.02249v1", "arxiv_id": "2605.02249", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "2fd0b0ceab279d9f7d3e3cd461aed4e819d11adedb587c1d3ee12cb114ff68c8", "sources": ["arxiv", "semantic_scholar"], "title": "The Reasoning Trap: An Information-Theoretic Bound on Closed-System Multi-Step LLM Reasoning", "abstract": "When copies of the same language model are prompted to debate, they produce diverse phrasings of one perspective rather than diverse perspectives. Multi-agent debate (MAD), and more broadly closed-system reasoning where agents iteratively transform each other's outputs, tends to preserve answer accuracy while degrading the reasoning behind those answers. We name the multi-agent case the Debate Trap and the broader phenomenon the Reasoning Trap, offering a programmatic theory of evidence-grounded reasoning failure.The framework has three parts: (i) SFS (Supported Faithfulness Score), a claim-level metric verifying decomposed atomic claims against provided evidence (decomposer-invariant rankings: Spearman rho=1.0); (ii) EGSR (Evidence-Grounded Socratic Reasoning), replacing adversarial argumentation with evidence-grounded inquiry; (iii) Theorem 1 (DPI Bound): under standard MAD, the chain E -> O^0 -> O^1 -> ... is Markov, and the Data Processing Inequality implies E[I(E;O^{t+1})] <= E[I(E;O^t)]. Three companion results -- open-system recovery (Theorem 2), EGSR accumulation (Lemma 2), and vote-aggregation floor (Proposition 1) -- partition multi-step LLM reasoning by its information-theoretic relationship to E. Across 16 conditions on SciFact (300 claims) and FEVER (1,000 claims), DebateCV (C13) preserves 88% of baseline accuracy while SFS drops 43%; majority-vote MAD (C15) reduces SFS to 1.7% of baseline (p < 10^{-6}, d = -0.96); EGSR recovers 98%. An R6 cohort study (Korean n=10x30 FEVER; English n=3x200 SciFact) finds inter-rater Fleiss kappa <= +0.018 with 0.8-1.4 Likert intra-rater shifts across language and domain -- the human agreement that faithfulness metrics have been calibrated against is not itself stable. We offer one falsifiable conjecture: any closed-system reasoning protocol preserving Theorem 1's Markov structure is, in expectation, subject to the same DPI bound.", "authors": ["Kwan Soo Shin"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-03", "url": "https://arxiv.org/abs/2605.01704", "pdf_url": "https://arxiv.org/pdf/2605.01704v2", "arxiv_id": "2605.01704", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e972e71ab00263d8124d60cacec7e62a3ff23e5b6282b459b3aa5519856720e5", "sources": ["arxiv", "semantic_scholar"], "title": "12 Angry AI Agents: Evaluating Multi-Agent LLM Decision-Making Through Cinematic Jury Deliberation", "abstract": "What if the twelve jurors of Sidney Lumet's 12 Angry Men (1957) were not men, but large language models? Would the one juror who disagrees still be able to change everyone's mind? This paper instantiates that scenario as a multi-agent benchmark for LLM deliberation: twelve agents, each conditioned on a film-faithful persona, debate the film's murder case using multi-agent framework. Two models representing opposite ends of the RLHF spectrum are tested: GPT-4o (closed-source, heavy alignment) and Llama-4-Scout (open-weight, lighter alignment), across three conditions (baseline, open-minded prompt, no initial vote), with N = 3 replications per cell (18 runs total). Three findings emerge. (i) Seventeen of eighteen runs end in a hung jury (a state where the jury fails to reach a unanimous verdict); the film's central event, gradual minority-to-majority persuasion, almost never occurs, indicating that anchoring is the dominant failure mode of current LLMs in this setting. (ii) The two models exhibit sharply different internal dynamics: GPT-4o produces a mean of 1.0 vote changes per run across all conditions, while Llama-4-Scout ranges from 2.0 (baseline) to 6.0 (open-minded prompt), and is the only model to reach a NOT\\_GUILTY verdict (1 of 3 runs in the no-initial-vote condition). The same ``open-minded'' instruction is internalized by Llama and ignored by GPT-4o. (iii) This asymmetry suggests that the intensity of RLHF alignment training, not model capability, is the primary determinant of deliberative flexibility in multi-agent settings. Flexibility, not capability, tracks human deliberation. The work is framed as an exploratory study and discusses implications for jury-of-LLMs evaluation and multi-agent debate.", "authors": ["Ahmet Bahaddin Ersoz"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-03", "url": "https://arxiv.org/abs/2605.01986", "pdf_url": "https://arxiv.org/pdf/2605.01986v1", "arxiv_id": "2605.01986", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "34733af10030829790220a9343a37252c2d76f51e5088c96757676798fcaab34", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Capsules: Quality-Gated Granularity Control for Multi-Agent LLM Pipelines", "abstract": "A multi-agent pipeline with N agents typically issues N LLM calls per run. Merging agents into fewer calls (compound execution) promises token savings, but naively merged calls silently degrade quality through tool loss and prompt compression. We present Agent Capsules, an adaptive execution runtime that treats multi-agent pipeline execution as an optimization problem with empirical quality constraints. The runtime instruments coordination overhead per group, scores composition opportunity, selects among three compound execution strategies, and gates every mode switch on rolling-mean output quality. A controlled negative result confirms that injecting more context into a merged call worsens compression rather than relieving it, so the framework's escalation ladder (standard, then two-phase, then sequential) recovers quality by moving toward per-agent dispatch rather than by rewriting merged prompts. On LLM-judged quality, the controller matches a hand-tuned oracle on every measured (model, group, mode) cell: routing compound whenever the oracle would, and reverting to fine whenever quality would fail the floor, without per-model configuration. Against a hand-crafted LangGraph implementation of a 14-agent competitive intelligence pipeline, Agent Capsules uses 51% fewer fine-mode input tokens and 42% fewer compound-mode input tokens, at +0.020 and +0.017 quality respectively. Against a DSPy implementation of a 5-agent due diligence pipeline, the framework uses 19% fewer tokens than uncompiled DSPy at quality parity, and 68% fewer tokens than MIPROv2 at +0.052 quality. Even before compound mode fires, the runtime delivers efficiency through automatic policy resolution, cache-aligned prompts, and topology-aware context injection, matching both hand-tuned and compile-time baselines without training data or per-pipeline engineering.", "authors": ["Aninda Ray"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-01", "url": "https://arxiv.org/abs/2605.00410", "pdf_url": "https://arxiv.org/pdf/2605.00410v1", "arxiv_id": "2605.00410", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/aray-17/agent-capsules", "venue": null, "quality_score": 0.65}
{"id": "170efc801c20b92af4a9324898bb8d4bc0892b154c0e5645d59962fdb27eb183", "sources": ["arxiv", "semantic_scholar"], "title": "To Call or Not to Call: A Framework to Assess and Optimize LLM Tool Calling", "abstract": "Agentic AI architectures augment LLMs with external tools, unlocking strong capabilities. However, tool use is not always beneficial; some calls may be redundant or even harmful. Effective tool use, therefore, hinges on a core LLM decision: whether to call or not call a tool when performing a task. This decision is particularly challenging for web search tools, where the benefits of external information depend on the model's internal knowledge and its ability to integrate potentially noisy tool responses. We introduce a principled framework inspired by decision-making theory to evaluate web search tool-use decisions along three key factors: necessity, utility, and affordability. Our analysis combines two complementary lenses: a normative perspective that infers true need and utility from an optimal allocation of tool calls, and a descriptive perspective that infers the model's self-perceived need and utility from their observed behaviors. We evaluate six open and one closed-source frontier models under two harnesses, one conditioning on only the current turn and its search results, the other on the full execution traces, across four web-search tools and three tasks. In every setting, we find that a model's perceived need and utility are frequently misaligned with the true need and utility. Building on this framework, we train lightweight estimators of need and utility from the models' hidden states. These estimators drive simple controllers that improve decision quality and yield stronger task performance than the self-perceived baseline for most of the open-source models.", "authors": ["Qinyuan Wu", "Soumi Das", "Mahsa Amani", "Arijit Nag", "Seungeon Lee", "Krishna P. Gummadi", "Abhilasha Ravichander", "Muhammad Bilal Zafar"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-01", "url": "https://arxiv.org/abs/2605.00737", "pdf_url": "https://arxiv.org/pdf/2605.00737v2", "arxiv_id": "2605.00737", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "c4bbe69ca93331b4b2c89b34562beac1a03bc0e95b6c546b8048e7d2d4f4096a", "sources": ["arxiv", "semantic_scholar"], "title": "TeamTR: Trust-Region Fine-Tuning for Multi-Agent LLM Coordination", "abstract": "Multi-agent LLM systems have shown promise for complex reasoning, yet recent evaluations reveal they often underperform single-model baselines. We identify a structural failure mode in sequential fine-tuning of shared-context teams: updating one agent shifts the team's context distribution, and when subsequent updates are evaluated on cached rollouts, this mismatch compounds. We formalize this as the compounding occupancy shift and prove that stale-occupancy evaluation incurs a penalty that scales quadratically with the number of agents. In contrast, intermediate-occupancy evaluation reduces this to linear scaling. We propose TeamTR, a trust-region framework that resamples trajectories after each component update and enforces per-agent divergence control, yielding rigorous per-update and per-stage improvement lower bounds. Experiments show that TeamTR outperforms single-agent and sequential baselines with 7.1% on average, mitigates coordination regressions, and supports plug-and-play component replacement. Code is available at https://github.com/Yydc/TeamTR.", "authors": ["Yi Xie", "Siao Liu", "Falong Fan", "Yuanqi Yao", "Yue Zhao", "Bo Liu"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-01", "url": "https://arxiv.org/abs/2605.15207", "pdf_url": "https://arxiv.org/pdf/2605.15207v1", "arxiv_id": "2605.15207", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Yydc/TeamTR", "venue": null, "quality_score": 0.65}
{"id": "564c07202f9148322e3b9dd05137f0817dc0590fa6065a2eafe93db31d2c8fcd", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Multi-Agent Autonomous Reasoning in Hydrodynamics", "abstract": "Single-agent systems (SAS) have become the default pattern for LLM-driven scientific workflows, but routing planning, tool use, and synthesis through a single context window comes with a well-known cost: as tool specifications and observational traces accumulate, the effective context available for each decision shrinks, and end-to-end reliability suffers. We present a multi-agent system (MAS) prototype for hydrodynamics in which specialized agents are coordinated through a Layer Execution Graph (LEG). A planner agent constructs query-specific execution topologies from natural-language routing heuristics that capture domain knowledge without hard-coding it as rigid control logic; specialist agents operate under strict tool allowlists and occupy complementary data-class roles. Between layers, consolidator agents fuse parallel outputs into concise briefs, and a reporter agent synthesizes the final response, while the runtime logs provenance for every tool invocation to support auditability. All benchmarks, ablations, and stress tests use Claude Sonnet~4.6 as the backbone model for both specialist and general-purpose agents. Evaluated on 37 queries spanning six complexity categories, the prototype achieves 93.6% factual precision with a 100% pass rate. Accuracy remains above 90% across runs from single-threaded to five independent parallel tracks, and under simulated loss of individual data sources the system degrades gracefully, still returning substantive partial answers. Together, these results suggest that planner-guided, graph-structured multi-agent orchestration can meaningfully alleviate the context-saturation bottlenecks that constrain monolithic single-agent architectures.", "authors": ["Jinpai Zhao", "Albert Cerrone", "Joannes Westerink", "Clint Dawson"], "categories": ["cs.AI", "physics.ao-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2026-05-01", "url": "https://arxiv.org/abs/2605.01102", "pdf_url": "https://arxiv.org/pdf/2605.01102v1", "arxiv_id": "2605.01102", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "52dcf1b3a36390dcf2e33329f58d4cb437cac4b5c08fa236a1ec466c5d8b8530", "sources": ["arxiv", "semantic_scholar"], "title": "From Static Analysis to Audience Dissemination: A Training-Free Multimodal Controversy Detection Multi-Agent Framework", "abstract": "Multimodal controversy detection (MCD) identifies controversial content in videos and their associated user comments, to support risk management for social video platforms.Prior research frames MCD as a static representation learning task, where features are directly extracted from videos and their accompanying comments. However, these methods fail to capture the diverse perspectives and evaluations from different audience groups. Inspired by the real-world process of content dissemination among audiences, we propose AuDisAgent, a training-free multi-agent framework that reformulates MCD as a dynamic propagation process.Our framework explicitly models audience dissemination through a structured multi-agent system. First, three specialized Screening Agents (Video Agent, Comment Agent, and Interaction Agent) conduct initial assessments from visual, textual, and cross-modal perspectives, respectively. For samples where the three agents cannot reach a consensus, a Viewing Panel Agent is activated to simulate post-screening discussions among audiences with diverse backgrounds and stances. This mechanism models how different audience groups interpret and react to the same content, uncovering latent controversial content that may emerge during the dissemination process. Finally, an Arbitration Agent renders the final judgment based on the complete reasoning chain from the preceding steps.In addition, to address the \"cold-start\" scenario where newly released videos have few or no comments, we design a Comment Bootstrapping Strategy that leverages historical public comments from semantically similar videos as the initial comment context. Extensive experiments on a public dataset demonstrate that our framework significantly outperforms existing state-of-the-art (SOTA) methods in both rich-comment and limited-comment scenarios.", "authors": ["Zihan Ding", "Ziyuan Yang", "Yi Zhang"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-05-01", "url": "https://arxiv.org/abs/2605.02939", "pdf_url": "https://arxiv.org/pdf/2605.02939v1", "arxiv_id": "2605.02939", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5e44797af40dc74c06d5e417436b219bc1ac9c93ba6b4a4366117b741ce1c324", "sources": ["arxiv", "semantic_scholar"], "title": "Building Persona-Based Agents On Demand: Tailoring Multi-Agent Workflows to User Needs", "abstract": "Recent advances in agentic AI are shifting automation from discrete tools to proactive multi-agent systems that coordinate multi-specialized capabilities behind unified interfaces. However, today's agent systems typically rely on hard-coded agent architectures with fixed roles, coordination patterns, and interaction flows that limit end-user personalization and make adaptation to individual needs and contexts difficult. Given this limitation, we argue that on-demand persona-based agent generation offers a promising path towards more efficient and contextually appropriate interaction within agentic workflows. By dynamically crafting agents and personas at run-time to match user characteristics, task demands, and workflow context, agentic platforms can move beyond one-size-fits-all configurations. We present a pipeline for on-demand persona generation in agentic platforms, detailing how real-time crafting of AI personas can be systematically integrated within agent systems, aiming to open new possibilities in agentic platform design paradigms.", "authors": ["Giuseppe Arbore", "Andrea Sillano", "Luigi De Russis"], "categories": ["cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-30", "url": "https://arxiv.org/abs/2604.27882", "pdf_url": "https://arxiv.org/pdf/2604.27882v1", "arxiv_id": "2604.27882", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "15fd4e2a26ca20aed14e4dc1cc3088dc76524e5db59c3297ead0f3b97aa9b684", "sources": ["arxiv", "semantic_scholar"], "title": "MCPHunt: An Evaluation Framework for Cross-Boundary Data Propagation in Multi-Server MCP Agents", "abstract": "Multi-server MCP agents create an information-flow control problem: faithful tool composition can turn individually benign read/write permissions into cross-boundary credential propagation -- a structural side effect of workflow topology, not necessarily malicious model behavior. We present MCPHunt, to our knowledge the first controlled benchmark that isolates non-adversarial, verbatim credential propagation across multi-server MCP trust boundaries, with three methodological contributions: (1) canary-based taint tracking that reduces propagation detection to objective string matching; (2) an environment-controlled coverage design with risky, benign, and hard-negative conditions that validates pipeline soundness and controls for credential-format confounds; (3) CRS stratification that disentangles task-mandated propagation (faithful execution of verbatim-transfer instructions) from policy-violating propagation (credentials included despite the option to redact). Across 3,615 main-benchmark traces from 5 models spanning 147 tasks and 9 mechanism families, policy-violating propagation rates reach 11.5--41.3% across all models. This propagation is pathway-specific (25x cross-mechanism range) and concentrated in browser-mediated data flows; hard-negative controls provide evidence that production-format credentials are not necessary -- prompt-directed cross-boundary data flow is sufficient. A prompt-mitigation study across 3 models reduces policy-violating propagation by up to 97% while preserving 80.5% utility, but effectiveness varies with instruction-following capability -- suggesting that prompt-level defenses alone may not suffice. Code, traces, and labeling pipeline are released under MIT and CC BY 4.0.", "authors": ["Haonan Li", "Tianjun Sun", "Yongqing Wang", "Qisheng Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-30", "url": "https://arxiv.org/abs/2604.27819", "pdf_url": "https://arxiv.org/pdf/2604.27819v1", "arxiv_id": "2604.27819", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/lihaonan0716/MCPHunt", "venue": null, "quality_score": 0.65}
{"id": "31c6e1db6143b920f806fb5e079b901c512d8eba0a3345c5b06dcfaa34926474", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Interaction Paradigms for LLM Agents in Scientific Visualization", "abstract": "This paper examines how different types of large language model (LLM) agents perform on scientific visualization (SciVis) tasks, where users generate visualization workflows from natural-language instructions. We compare three primary interaction paradigms, including domain-specific agents with structured tool use, computer-use agents, and general-purpose coding agents, by evaluating eight representative agents across 15 benchmark tasks and measuring visualization quality, efficiency, robustness, and computational cost. We further analyze interaction modalities, including code scripts and model context protocol (MCP) or API calls for structured tool use, as well as command-line interfaces (CLI) and graphical user interfaces (GUI) for more general interaction, while additionally studying the effect of persistent memory in selected agents. The results reveal clear tradeoffs across paradigms and modalities. General-purpose coding agents achieve the highest task success rates but are computationally expensive, while domain-specific agents are more efficient and stable but less flexible. Computer-use agents perform well on individual steps but struggle with longer multi-step workflows, indicating that long-horizon planning is their primary limitation. Across both CLI- and GUI-based settings, persistent memory improves performance over repeated trials, although its benefits depend on the underlying interaction mode and the quality of feedback. These findings suggest that no single approach is sufficient, and future SciVis systems should combine structured tool use, interactive capabilities, and adaptive memory mechanisms to balance performance, robustness, and flexibility.", "authors": ["Jackson Vonderhorst", "Kuangshi Ai", "Haichao Miao", "Shusen Liu", "Chaoli Wang"], "categories": ["cs.AI", "cs.GR", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-30", "url": "https://arxiv.org/abs/2604.27996", "pdf_url": "https://arxiv.org/pdf/2604.27996v2", "arxiv_id": "2604.27996", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e3eb7c89e1629bf68e88b9bbfbe301827d97d154686f8f8bd471c3447efed70d", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforced Agent: Inference-Time Feedback for Tool-Calling Agents", "abstract": "Tool-calling agents are evaluated on tool selection, parameter accuracy, and scope recognition, yet LLM trajectory assessments remain inherently post-hoc. Disconnected from the active execution loop, such assessments identify errors that are usually addressed through prompt-tuning or retraining, and fundamentally cannot course-correct the agent in real time. To close this gap, we move evaluation into the execution loop at inference time: a specialized reviewer agent evaluates provisional tool calls prior to execution, shifting the paradigm from post-hoc recovery to proactive evaluation and error mitigation. In practice, this architecture establishes a clear separation of concerns between the primary execution agent and a secondary review agent. As with any multi-agent system, the reviewer can introduce new errors while correcting others, yet no prior work to our knowledge has systematically measured this tradeoff. To quantify this tradeoff, we introduce Helpfulness-Harmfulness metrics: helpfulness measures the percentage of base agent errors that feedback corrects; harmfulness measures the percentage of correct responses that feedback degrades. These metrics directly inform reviewer design by revealing whether a given model or prompt provides net positive value. We evaluate our approach on BFCL (single-turn) and Tau2-Bench (multi-turn stateful scenarios), achieving +5.5% on irrelevance detection and +7.1% on multi-turn tasks. Our metrics reveal that reviewer model choice is critical: the reasoning model o3-mini achieves a 3:1 benefit-to-risk ratio versus 2.1:1 for GPT-4o. Automated prompt optimization via GEPA provides an additional +1.5-2.8%. Together, these results demonstrate a core advantage of separating execution and review: the reviewer can be systematically improved through model selection and prompt optimization, without retraining the base agent.", "authors": ["Anh Ta", "Junjie Zhu", "Shahin Shayandeh"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-29", "url": "https://arxiv.org/abs/2604.27233", "pdf_url": "https://arxiv.org/pdf/2604.27233v1", "arxiv_id": "2604.27233", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "fce16b2519b7fefd7751905011d0e1d8ac08fc46092a6961be77c8a570f7d8d5", "sources": ["arxiv", "semantic_scholar"], "title": "A Systematic Comparison of Prompting and Multi-Agent Methods for LLM-based Stance Detection", "abstract": "Stance detection identifies the attitude of a text author toward a given target. Recent studies have explored various LLM-based strategies for this task, from zero-shot prompting to multi-agent debate. However, existing works differ in data splits, base models, and evaluation protocols, making fair comparison difficult. We conduct a systematic comparison that evaluates five methods across two categories -- prompt-based inference (Direct Prompting, Auto-CoT, StSQA) and agent-based debate (COLA, MPRF) -- on four datasets with 14 subtasks, using 15 LLMs from six model families with parameter sizes from 7B to 72B+. Our experiments yield several findings. First, on all models with complete results, the best prompt-based method outperforms the best agent-based method, while agent methods require 7 to 12 times more API calls per sample. Second, model scale has a larger impact on performance than method choice, with gains plateauing around 32B. Third, reasoning-enhanced models (DeepSeek-R1) do not consistently outperform general models of the same size on this task.", "authors": ["Genan Dai", "Zini Chen", "Yi Yang", "Bowen Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-29", "url": "https://arxiv.org/abs/2604.26319", "pdf_url": "https://arxiv.org/pdf/2604.26319v1", "arxiv_id": "2604.26319", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d134f98006ac8bbec98aa9a31e798a1750efca2ec5d33e4dce48b2aa67e9dd0d", "sources": ["arxiv", "semantic_scholar"], "title": "FAMA: Failure-Aware Meta-Agentic Framework for Open-Source LLMs in Interactive Tool Use Environments", "abstract": "Large Language Models are being increasingly deployed as the decision-making core of autonomous agents capable of effecting change in external environments. Yet, in conversational benchmarks, which simulate real-world customer-centric issue resolution scenarios, these agents frequently fail due to the cascading effects of incorrect decision-making. These challenges are particularly pronounced for open-source LLMs with smaller parameter sizes, limited context windows, and constrained inference budgets, which contribute to increased error accumulation in agentic settings. To tackle these challenges, we present the Failure-Aware Meta-Agentic (FAMA) framework. FAMA operates in two stages: first, it analyzes failure trajectories from baseline agents to identify the most prevalent errors; second, it employs an orchestration mechanism that activates a minimal subset of specialized agents tailored to address these failures by injecting a targeted context for the tool-use agent before the decision-making step. Experiments across open-source LLMs demonstrate performance gains up to 27% across evaluation modes over standard baselines. These results highlight that targeted curation of context through specialized agents to address common failures is a valuable design principle for building reliable, multi-turn tool-use LLM agents that simulate real-world conversational scenarios.", "authors": ["Amir Saeidi", "Venkatesh Mishra", "Souradeep Mukhopadhyay", "Gaowen Liu", "Ali Payani", "Jayanth Srinivasa", "Chitta Baral"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-28", "url": "https://arxiv.org/abs/2604.25135", "pdf_url": "https://arxiv.org/pdf/2604.25135v1", "arxiv_id": "2604.25135", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "c853cd6d8ee8ca6a0c18186e5a31e9a9fff2ca8e9176cbcbd30753f712d97039", "sources": ["arxiv", "semantic_scholar"], "title": "Cutscene Agent: An LLM Agent Framework for Automated 3D Cutscene Generation", "abstract": "Cutscenes are carefully choreographed cinematic sequences embedded in video games and interactive media, serving as the primary vehicle for narrative delivery, character development, and emotional engagement. Producing cutscenes is inherently complex: it demands seamless coordination across screenwriting, cinematography, character animation, voice acting, and technical direction, often requiring days to weeks of collaborative effort from multidisciplinary teams to produce minutes of polished content. In this work, we present Cutscene Agent, an LLM agent framework for automated end-to-end cutscene generation. The framework makes three contributions: (1)~a Cutscene Toolkit built on the Model Context Protocol (MCP) that establishes \\emph{bidirectional} integration between LLM agents and the game engine -- agents not only invoke engine operations but continuously observe real-time scene state, enabling closed-loop generation of editable engine-native cinematic assets; (2)~a multi-agent system where a director agent orchestrates specialist subagents for animation, cinematography, and sound design, augmented by a visual reasoning feedback loop for perception-driven refinement; and (3)~CutsceneBench, a hierarchical evaluation benchmark for cutscene generation. Unlike typical tool-use benchmarks that evaluate short, isolated function calls, cutscene generation requires long-horizon, multi-step orchestration of dozens of interdependent tool invocations with strict ordering constraints -- a capability dimension that existing benchmarks do not cover. We evaluate a range of LLMs on CutsceneBench and analyze their performance across this challenging task.", "authors": ["Lanshan He", "Haozhou Pang", "Qi Gan", "Xin Shen", "Ziwei Zhang", "Yibo Liu", "Gang Fang", "Bo Liu", "Kai Sheng", "Shengfeng Zeng", "Chaofan Li", "Zhen Hui", "Keer Zhou", "Lan Zhou", "Shujun Dai"], "categories": ["cs.GR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-28", "url": "https://arxiv.org/abs/2604.25318", "pdf_url": "https://arxiv.org/pdf/2604.25318v1", "arxiv_id": "2604.25318", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1ad0b8ce094d4b1968409873c5bd654edcf04a729e8f3d95ea7506aa5aadf166", "sources": ["arxiv", "semantic_scholar"], "title": "Pythia: Exploiting Workflow Predictability for Efficient Agent-Native LLM Serving", "abstract": "As LLM applications grow more complex, developers are increasingly adopting multi-agent architectures to decompose workflows into specialized, collaborative components, introducing structure that constrains agent behavior and exposes useful semantic predictability. Unlike traditional LLM serving, which operates under highly dynamic and uncertain conditions, this structured topology enables opportunities to reduce runtime uncertainty$\\unicode{x2015}$yet existing systems fail to exploit it, treating agentic workloads as generic traffic and incurring significant inefficiencies. Our analysis of production traces from an agent-serving platform and an internal coding assistant reveals key bottlenecks, including low prefix cache hit rates, severe resource contention from long-context requests, and substantial queuing delays due to suboptimal scaling. To address these challenges, we propose Pythia, a multi-agent serving system that captures workflow semantics through a simple interface at the serving layer, unlocking new optimization opportunities and substantially improving throughput and job completion time over state-of-the-art baselines.", "authors": ["Shan Yu", "Junyi Shu", "Yuanjiang Ni", "Kun Qian", "Xue Li", "Yang Wang", "Jinyuan Zhang", "Ziyi Xu", "Shuo Yang", "Lingjun Zhu", "Ennan Zhai", "Qingda Lu", "Jiarong Xing", "Youyou Lu", "Xin Jin", "Xuanzhe Liu", "Harry Xu"], "categories": ["cs.MA", "cs.DC", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2026-04-28", "url": "https://arxiv.org/abs/2604.25899", "pdf_url": "https://arxiv.org/pdf/2604.25899v2", "arxiv_id": "2604.25899", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c30b1b36683f63e53b5387b35a08b729475d0ed7a334352d782e059543007fd4", "sources": ["arxiv", "semantic_scholar"], "title": "MARD: A Multi-Agent Framework for Robust Android Malware Detection", "abstract": "With the rapid evolution of Android applications, traditional machine learning-based detection models suffer from concept drift. Additionally, they are constrained by shallow features, lacking deep semantic understanding and interpretability of decisions. Although Large Language Models (LLMs) demonstrate remarkable semantic reasoning capabilities, directly processing massive raw code incurs prohibitive token overhead. Moreover, this approach fails to fully unleash the deep logical reasoning potential of LLMs within complex contexts. To address these limitations, we propose MARD, a multi-agent framework for robust Android malware detection. This framework effectively bridges the gap between the semantic understanding of LLMs and traditional static analysis. It treats underlying deterministic analysis engines as on-demand execution tools, while utilizing the LLM to orchestrate the entire decision-making process. By designing an autonomous multi-agent interaction mechanism based on the ReAct paradigm, MARD constructs a highly interpretable evidentiary chain for conviction. Furthermore, we radically reduce the total cost of conducting a deep analysis of a single complex APK to under $0.10. Evaluations demonstrate that, without any domain-specific fine-tuning, MARD achieves an F1 score of 93.46%. It not only outperforms continual learning baselines but also exhibits robustness against concept drift and strong cross-domain generalization capabilities in evaluations spanning up to five years.", "authors": ["Xueying Zeng", "Youquan Xian", "Sihao Liu", "Xudong Mou", "Yanze Li", "Lei Cui", "Bo Li"], "categories": ["cs.CR", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-28", "url": "https://arxiv.org/abs/2604.25264", "pdf_url": "https://arxiv.org/pdf/2604.25264v1", "arxiv_id": "2604.25264", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "3f8b53b5fb6c26203640855d7fac2a5c0dbb7184e7e89486c05e64227a916c52", "sources": ["arxiv", "semantic_scholar"], "title": "SAFEdit: Does Multi-Agent Decomposition Resolve the Reliability Challenges of Instructed Code Editing?", "abstract": "Instructed code editing is a significant challenge for large language models (LLMs). On the EditBench benchmark, 39 of 40 evaluated models obtain a task success rate (TSR) below 60 percent, highlighting a gap between general code generation and the ability to perform instruction-driven editing under executable test constraints. To address this, we propose SAFEdit, a multi-agent framework for instructed code editing that decomposes the editing process into specialized roles to improve reliability and reduce unintended code changes. A Planner Agent produces an explicit, visibility-aware edit plan, an Editor Agent applies minimal, literal code modifications, and a Verifier Agent executes real test runs. When tests fail, SAFEdit uses a Failure Abstraction Layer (FAL) to transform raw test logs into structured diagnostic feedback, which is fed back to the Editor to support iterative refinement. We compare SAFEdit against both prior single-model results reported for EditBench and an implemented ReAct single-agent baseline under the same evaluation conditions. We used EditBench to evaluate SAFEdit on 445 code editing instances in five languages (English, Polish, Spanish, Chinese, and Russian) under varying spatial context variants. SAFEdit achieved 68.6 percent TSR, outperforming the single-model baseline by 3.8 percentage points and the ReAct single-agent baseline by 8.6 percentage points. The iterative refinement loop was found to contribute 17.4 percentage points to SAFEdit's overall success rate. SAFEdit's automated error analysis further indicates a reduction in instruction-level hallucinations compared to single-agent approaches, providing an additional framework component for interpreting failures beyond pass or fail outcomes.", "authors": ["Noam Tarshish", "Nofar Selouk", "Daniel Hodisan", "Bar Ezra Gafniel", "Yuval Elovici", "Asaf Shabtai", "Eliya Nachmani"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-28", "url": "https://arxiv.org/abs/2604.25737", "pdf_url": "https://arxiv.org/pdf/2604.25737v1", "arxiv_id": "2604.25737", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0e37171d683f817f473813ed34c02d4200c8172ebf88cc8d74db3096c614837b", "sources": ["arxiv", "semantic_scholar"], "title": "PolyKV: A Shared Asymmetrically-Compressed KV Cache Pool for Multi-Agent LLM Inference", "abstract": "We present PolyKV, a system in which multiple concurrent inference agents share a single, asymmetrically compressed KV cache pool. Rather than allocating a separate KV cache per agent -- the standard paradigm -- PolyKV writes a compressed cache once and injects it into N independent agent contexts via HuggingFace DynamicCache objects. Compression is asymmetric: Keys are quantized at int8 (q8_0) to preserve softmax stability, while Values are compressed using TurboQuant MSE -- a Fast Walsh-Hadamard Transform (FWHT) rotation followed by 3-bit Lloyd-Max quantization with centroids tuned to N(0,1). We evaluate across two model scales (SmolLM2-1.7B-Instruct and Llama-3-8B-Instruct), three context lengths (600-7,194 tokens), and up to 15 concurrent agents. PolyKV achieves a stable 2.91x compression ratio across all configurations. On Llama-3-8B with 15 agents sharing a 4K-token context, PolyKV reduces KV cache memory from 19.8 GB to 0.45 GB -- a 97.7% reduction -- while maintaining only +0.57% perplexity degradation and a mean BERTScore F1 of 0.928. PPL delta does not grow with agent count and improves as context length increases, inverting to -0.26% at 1,851 coherent tokens. To our knowledge, no prior work combines a single shared, lossy-compressed KV pool with multi-reader concurrent agent access.", "authors": ["Ishan Patel", "Ishan Joshi"], "categories": ["cs.LG", "cs.CL", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-27", "url": "https://arxiv.org/abs/2604.24971", "pdf_url": "https://arxiv.org/pdf/2604.24971v1", "arxiv_id": "2604.24971", "doi": "10.5281/zenodo.19686729", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ishan1410/PolyKV", "venue": null, "quality_score": 0.65}
{"id": "da117933d9277dbb5a47f74eb0523705d84395f26a7711036cdde8183d70371f", "sources": ["arxiv", "semantic_scholar"], "title": "Latent Agents: A Post-Training Procedure for Internalized Multi-Agent Debate", "abstract": "Multi-agent debate has been shown to improve reasoning in large language models (LLMs). However, it is compute-intensive, requiring generation of long transcripts before answering questions. To address this inefficiency, we develop a framework that distills multi-agent debate into a single LLM through a two-stage fine-tuning pipeline combining debate structure learning with internalization via dynamic reward scheduling and length clipping. Across multiple models and benchmarks, our internalized models match or exceed explicit multi-agent debate performance using up to 93% fewer tokens. We then investigate the mechanistic basis of this capability through activation steering, finding that internalization creates agent-specific subspaces: interpretable directions in activation space corresponding to different agent perspectives. We further demonstrate a practical application: by instilling malicious agents into the LLM through internalized debate, then applying negative steering to suppress them, we show that distillation makes harmful behaviors easier to localize and control with smaller reductions in general performance compared to steering base models. Our findings offer a new perspective for understanding multi-agent capabilities in distilled models and provide practical guidelines for controlling internalized reasoning behaviors. Code available at https://github.com/johnsk95/latent_agents", "authors": ["John Seon Keun Yi", "Aaron Mueller", "Dokyun Lee"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-27", "url": "https://arxiv.org/abs/2604.24881", "pdf_url": "https://arxiv.org/pdf/2604.24881v1", "arxiv_id": "2604.24881", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/johnsk95/latent_agents", "venue": null, "quality_score": 0.65}
{"id": "54551ce79f3a74601caadc92a4e79810c3aeec725264ec800836b8414937f6aa", "sources": ["arxiv", "semantic_scholar"], "title": "GAMMAF: A Common Framework for Graph-Based Anomaly Monitoring Benchmarking in LLM Multi-Agent Systems", "abstract": "The rapid integration of Large Language Models (LLMs) into Multi-Agent Systems (MAS) has significantly enhanced their collaborative problem-solving capabilities, but it has also expanded their attack surfaces, exposing them to vulnerabilities such as prompt infection and compromised inter-agent communication. While emerging graph-based anomaly detection methods show promise in protecting these networks, the field currently lacks a standardized, reproducible environment to train these models and evaluate their efficacy. To address this gap, we introduce Gammaf (Graph-based Anomaly Monitoring for LLM Multi-Agent systems Framework), an open-source benchmarking platform. Gammaf is not a novel defense mechanism itself, but rather a comprehensive evaluation architecture designed to generate synthetic multi-agent interaction datasets and benchmark the performance of existing and future defense models. The proposed framework operates through two interdependent pipelines: a Training Data Generation stage, which simulates debates across varied network topologies to capture interactions as robust attributed graphs, and a Defense System Benchmarking stage, which actively evaluates defense models by dynamically isolating flagged adversarial nodes during live inference rounds. Through rigorous evaluation using established defense baselines (XG-Guard and BlindGuard) across multiple knowledge tasks (such as MMLU-Pro and GSM8K), we demonstrate Gammaf's high utility, topological scalability, and execution efficiency. Furthermore, our experimental results reveal that equipping an LLM-MAS with effective attack remediation not only recovers system integrity but also substantially reduces overall operational costs by facilitating early consensus and cutting off the extensive token generation typical of adversarial agents.", "authors": ["Pablo Mateo-Torrejón", "Alfonso Sánchez-Macián"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-27", "url": "https://arxiv.org/abs/2604.24477", "pdf_url": "https://arxiv.org/pdf/2604.24477v1", "arxiv_id": "2604.24477", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "7e05f1333bab62b5583bbe3993459b85f0411ba806d0145f0785a928ab4d558f", "sources": ["arxiv", "semantic_scholar"], "title": "ITAS: A Multi-Agent Architecture for LLM-Based Intelligent Tutoring", "abstract": "Large language model tutors are easy to build in a notebook and hard to run in a real course. We describe ITAS (Intelligent Teaching Assistant System), a multi-agent tutoring system that a graduate quantum computing course used for a semester at Old Dominion University. The system has three layers. The teaching layer is a Spoke-and-Wheel of three parallel specialist agents (Video, Code, Guidance) followed by a Synthesizer, plus a separate autograder that evaluates both the correctness and the approach of checkpoint submissions. The operational layer is four Cloud Run microservices with session state in Cloud SQL and interaction events streamed through Pub/Sub to BigQuery. The feedback layer is a narrow-scope conversational agent that answers instructor questions over per-lesson pseudonymized event streams, addressing what we call the Blind Instructor Problem: LLM tutors accumulate more data about students than the instructor can reach through routine channels. The architecture is a direct response to specific failures of an earlier prototype, and we describe which of those fixes carried forward and which were dropped for this iteration. We report on a pilot deployment (five students, one course, one semester) interpreted as system-behavior evidence rather than learning-outcome evidence: the teaching layer handled 334 chat turns without the task-boundary hallucinations that domain consolidation would have risked, the operational layer captured 10,628 events across five modules, and the feedback layer surfaced two findings the instructor acted on mid-semester. We do not claim the pilot generalizes. We do claim that the system as described is one workable answer to the question of what an LLM-based ITS needs to look like end-to-end to run in a real course.", "authors": ["Iizalaarab Elhaimeur", "Nikos Chrisochoides"], "categories": ["cs.MA", "cs.AI", "cs.CY", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-27", "url": "https://arxiv.org/abs/2604.24808", "pdf_url": "https://arxiv.org/pdf/2604.24808v1", "arxiv_id": "2604.24808", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9dba756379bf7e05b91180d0b8f9ef498f635dc386fcca6c1217278862dc99c8", "sources": ["arxiv", "semantic_scholar"], "title": "GraphPlanner: Graph Memory-Augmented Agentic Routing for Multi-Agent LLMs", "abstract": "LLM routing has achieved promising results in integrating the strengths of diverse models while balancing efficiency and performance. However, to support more realistic and challenging applications, routing must extend into agentic LLM settings, where task planning, multi-round cooperation among heterogeneous agents, and memory utilization are indispensable. To address this gap, we propose GraphPlanner, a heterogeneous graph memory-augmented agentic router for multi-agent LLMs that generates routing workflows for each query and supports both inductive and transductive inference. GraphPlanner formulates workflow generation as a Markov Decision Process (MDP), where at each step it selects both the LLM backbone and the agent role, including Planner, Executor, and Summarizer. By leveraging a heterogeneous graph, denoted as GARNet, to capture interaction memories among queries, agents, and responses, GraphPlanner integrates historical memory and workflow memory into richer state representations. The entire pipeline is optimized with reinforcement learning, jointly improving task-specific performance and computational efficiency. We evaluate GraphPlanner across 14 diverse LLM tasks and demonstrate that: (1) GraphPlanner outperforms strong single-round and multi-round routers, improving accuracy by up to 9.3% while reducing GPU cost from 186.26 GiB to 1.04 GiB; (2) GraphPlanner generalizes robustly to unseen tasks and LLMs, exhibiting strong zero-shot capabilities; and (3) GraphPlanner effectively leverages historical memories, supporting both inductive and transductive inference for more adaptive routing. Our code for GraphPlanner is released at https://github.com/ulab-uiuc/GraphPlanner.", "authors": ["Tao Feng", "Haozhen Zhang", "Zijie Lei", "Peixuan Han", "Jiaxuan You"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-26", "url": "https://arxiv.org/abs/2604.23626", "pdf_url": "https://arxiv.org/pdf/2604.23626v1", "arxiv_id": "2604.23626", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ulab-uiuc/GraphPlanner", "venue": "ICLR 2026", "quality_score": 0.85}
{"id": "fc727bd14671190d0f01358bb6c9ec0d1deab85d1f6db655736a0ba0ebc21c76", "sources": ["arxiv", "semantic_scholar"], "title": "ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation", "abstract": "Skill-distillation pipelines learn reusable rules from LLM agent trajectories, but they lack a key signal: how much each step costs. Without per-step cost, a pipeline cannot distinguish adding a missing step to fix a bug from removing an expensive step that never affected the outcome. We use the cost-attribution gap to ask whether the rule types inside a distilled skill transfer the same way to new tasks. ClawTrace records cost-attributed agent traces and compiles each session into a TraceCard; CostCraft reads TraceCards and writes three kinds of skill patches: preserve, prune, and repair. We find a pattern aggregate metrics hide. On 30 held-out SpreadsheetBench tasks across two seeds, removing prune patches roughly tripled the quality-regression count without lowering median cost. Across the full 84-task SkillsBench transfer, CostCraft saves no aggregate cost. All three quality regressions trace to the preserve lane, and both quality wins trace to the prune lane: prune patches act as quality guardrails while preserve patches drive regressions. We argue that reusable agent skills should be evaluated at the rule-type level, not as monolithic instruction packages. To support this, we release ClawTrace, the TraceCard schema, and the full set of typed skills.", "authors": ["Boqin Yuan", "Yue Su", "Renchu Song", "Sen Yang", "Jing Qin"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-26", "url": "https://arxiv.org/abs/2604.23853", "pdf_url": "https://arxiv.org/pdf/2604.23853v2", "arxiv_id": "2604.23853", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "c7b85aa58a56c620045d8ed835ba8ccdcd72f6cf4360fe9407d6175a83567eee", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Automated Ontology Generation from Unstructured Text: A Multi-Agent LLM Approach", "abstract": "Automatically generating formal ontologies from unstructured natural language remains a central challenge in knowledge engineering. While large language models (LLMs) show promise, it remains unclear which architectural design choices drive generation quality and why current approaches fail. We present a controlled experimental study using domain-specific insurance contracts to investigate these questions. We first establish a single-agent LLM baseline, identifying key failure modes such as poor Ontology Design Pattern compliance, structural redundancy, and ineffective iterative repair. We then introduce a multi-agent architecture that decomposes ontology construction into four artifact-driven roles: Domain Expert, Manager, Coder, and Quality Assurer. We evaluate performance across architectural quality (via a panel of heterogeneous LLM judges) and functional usability (via competency question driven SPARQL evaluation with complementary retrieval augmented generation based assessment). Results show that the multi-agent approach significantly improves structural quality and modestly enhances queryability, with gains driven primarily by front-loaded planning. These findings highlight planning-first, artifact-driven generation as a promising and more auditable path toward scalable automated ontology engineering.", "authors": ["Abid Talukder", "Maruf Ahmed Mridul", "Oshani Seneviratne"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-25", "url": "https://arxiv.org/abs/2604.23090", "pdf_url": "https://arxiv.org/pdf/2604.23090v1", "arxiv_id": "2604.23090", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "14c1cd14ad6119ad2fd7caca85b8f4f809ef420c9046dff58e3b195fd5578e9e", "sources": ["arxiv", "semantic_scholar"], "title": "Architecture Matters for Multi-Agent Security", "abstract": "Multi-agent systems (MAS), composed of networks of two or more autonomous AI agents, have become increasingly popular in production deployments, yet introduce security risks that do not arise in single-agent settings. Even if individual agents exhibit robust security, architectural decisions governing their coordination can create attack surfaces that have not been systematically characterized. In this work, we present an empirical study of how MAS design decisions shape the tradeoff between task performance and attack resistance. Across three agentic environments (browser, desktop, and code) and 13 architectural configurations, we use stagewise evaluations that distinguish planning refusal, execution-stage interception, partial harmful execution, and successful attack completion to study three key design choices: (i) agent roles, which determine how authority and responsibility are allocated; (ii) communication topology, which shapes how and when agents interact; and (iii) memory, which determines the context and state visibility accessible to each agent. We find that multi-agent architectures are more vulnerable than standalone agents in the majority of configurations, with attack success rates varying by up to 3.8x at comparable or higher benign accuracy, and that no single design is universally safer. These results motivate the development of further evaluations that move beyond the security properties of a single agent.", "authors": ["Ben Hagag", "William L. Anderson", "Christian Schroeder de Witt", "Sarah Scheffler"], "categories": ["cs.MA", "cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-25", "url": "https://arxiv.org/abs/2604.23459", "pdf_url": "https://arxiv.org/pdf/2604.23459v1", "arxiv_id": "2604.23459", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b5e97d6f0f59cf61de3e090d55400fe3aaa5de1679c76d45823171f8e5c2354a", "sources": ["arxiv", "semantic_scholar"], "title": "Reliable Self-Harm Risk Screening via Adaptive Multi-Agent LLM Systems", "abstract": "Emerging AI systems in behavioral health and psychiatry use multi-step or multi-agent LLM pipelines for tasks like assessing self-harm risk and screening for depression. However, common evaluation approaches, like LLM-as-a-judge, do not indicate when a decision is reliable or how errors may accumulate across multiple LLM judgements, limiting their suitability for safety-critical settings. We present a statistical framework for multi-agent pipelines structured as directed acyclic graphs (DAGs) that provides an alternative to heuristic voting with principled, adaptive decision-making. We model each agent as a stochastic categorical decision and introduce (1) tighter agent-level performance confidence bounds, (2) a bandit-based adaptive sampling strategy based on input difficulty, and (3) regret guarantees over the multi-agent system that shows logarithmic error growth when deployed. We evaluate our system on two labeled datasets in behavioral health : the AEGIS 2.0 behavioral health subset (N=161) and a stratified sample of SWMH Reddit posts (N=250). Empirically, our adaptive sampling strategy achieves the lowest false positive rate of any condition across both datasets, 0.095 on AEGIS 2.0 compared to 0.159 for single-agent models, reducing incorrect flagging of safe content by 40\\% and still having similar false negative rates across all conditions. These results suggest that principled adaptive sampling offers a meaningful improvement in precision without reducing recall in this setting.", "authors": ["Meghana Karnam", "Ananya Joshi"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-24", "url": "https://arxiv.org/abs/2604.22154", "pdf_url": "https://arxiv.org/pdf/2604.22154v1", "arxiv_id": "2604.22154", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "733b4c541a2cee9c163ee9aead4f366d2a88f1c669d2ac276c9e7ca17ebe491a", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Consensus as a Cognitive Bias Trigger in Human-AI Interaction", "abstract": "As multi-agent AI systems become more common, users increasingly encounter not a single AI voice but a collective one. This shift introduces social dynamics, such as consensus, dissent, and gradual convergence, that can trigger cognitive biases and distort human judgment. We present findings from a controlled experiment (N = 127) comparing three multi-agent configurations: Majority, Minority, and Diffusion. Quantitative results show that majority consensus accelerates opinion change and inflates confidence, consistent with social proof and bandwagon heuristics. Minority dissent slows this process and promotes more deliberative engagement. Qualitative analysis identifies three interpretive trajectories: reinforcing, aligning, and oscillating, shaped by how users interpret agent independence and group dynamics over time. These findings suggest that agent agreement structure, independent of content, functions as a bias-relevant signal in LLM interactions. We hope this work contributes to the Bias4Trust agenda by grounding multi-agent social influence as a concrete and designable source of bias in human-AI interaction.", "authors": ["Soohwan Lee", "Kyungho Lee"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-24", "url": "https://arxiv.org/abs/2604.22277", "pdf_url": "https://arxiv.org/pdf/2604.22277v1", "arxiv_id": "2604.22277", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f3a0830fe004a282c5968e5362c083077bc9555a7ded3b945038e07e189a1a1f", "sources": ["arxiv", "semantic_scholar"], "title": "Uncertainty Quantification for LLM Function-Calling", "abstract": "Large Language Models (LLMs) are increasingly deployed to autonomously solve real-world tasks. A key ingredient for this is the LLM Function-Calling paradigm, a widely used approach for equipping LLMs with tool-use capabilities. However, an LLM calling functions incorrectly can have severe implications, especially when their effects are irreversible, e.g., transferring money or deleting data. Hence, it is of paramount importance to consider the LLM's confidence that a function call solves the task correctly prior to executing it. Uncertainty Quantification (UQ) methods can be used to quantify this confidence and prevent potentially incorrect function calls. In this work, we present what is, to our knowledge, the first evaluation of UQ methods for LLM Function-Calling (FC). While multi-sample UQ methods, such as Semantic Entropy, show strong performance for natural language Q&A tasks, we find that in the FC setting, it offers no clear advantage over simple single-sample UQ methods. Additionally, we find that the particularities of FC outputs can be leveraged to improve the performance of existing UQ methods in this setting. Specifically, multi-sample UQ methods benefit from clustering FC outputs based on their abstract syntax tree parsing, while single-sample UQ methods can be improved by selecting only semantically meaningful tokens when calculating logit-based uncertainty scores.", "authors": ["Zihuiwen Ye", "Lukas Aichberger", "Michael Kirchhof", "Sinead Williamson", "Luca Zappella", "Yarin Gal", "Arno Blaas", "Adam Golinski"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-24", "url": "https://arxiv.org/abs/2604.22985", "pdf_url": "https://arxiv.org/pdf/2604.22985v1", "arxiv_id": "2604.22985", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "4fd20acd63db9edba793a32045592f7280c1dd86ca7f4e9cd8613eac0a570546", "sources": ["arxiv", "semantic_scholar"], "title": "Seeing the Whole Elephant: A Benchmark for Failure Attribution in LLM-based Multi-Agent Systems", "abstract": "Failure attribution, i.e., identifying the responsible agent and decisive step of a failure, is particularly challenging in LLM-based multi-agent systems (MAS) due to their natural-language reasoning, nondeterministic outputs, and intricate interaction dynamics. A reliable benchmark is therefore essential to guide and evaluate attribution techniques. Yet existing benchmarks rely on partially observable traces that capture only agent outputs, omitting the inputs and context that developers actually use when debugging. We argue that failure attribution should be studied under full execution observability, aligning with real-world developer-facing scenarios where complete traces, rather than only outputs, are accessible for diagnosis. To this end, we introduce TraceElephant, a benchmark designed for failure attribution with full execution traces and reproducible environments. We then systematically evaluate failure attribution techniques across various configurations. Specifically, full traces improve attribution accuracy by up to 76\\% over a partial-observation counterpart, confirming that missing inputs obscure many failure causes. TraceElephant provides a foundation for follow-up failure attribution research, promoting evaluation practices that reflect real-world debugging and supporting the development of more transparent MASs.", "authors": ["Mengzhuo Chen", "Junjie Wang", "Fangwen Mu", "Yawen Wang", "Zhe Liu", "Huanxiang Feng", "Qing Wang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-24", "url": "https://arxiv.org/abs/2604.22708", "pdf_url": "https://arxiv.org/pdf/2604.22708v1", "arxiv_id": "2604.22708", "doi": null, "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "dcc8090709814031b9d65e8f08fadaf4722494561c261a578774242e7957b9f1", "sources": ["arxiv", "semantic_scholar"], "title": "Envisioning Sensemaking in Multi-Human, Multi-Agent Collaborative Knowledge Work", "abstract": "Sensemaking is central to knowledge work, where people search, evaluate, interpret, and use information over time to construct durable understanding. The rise of generative AI has begun to reshape this process: GenAI systems now perform interpretive functions such as summarization, synthesis, and thematic grouping that knowledge workers have traditionally carried out themselves. In collaborative settings, these shifts compound, complicating how teams divide interpretive labor, trust one another's contributions, and negotiate shared understanding. In this position paper, we examine how GenAI reshapes sensemaking in collaborative knowledge work and propose five design principles for multi-human, multi-agent collaborative sensemaking: dynamic multi-layer information representations, active identification and bridging of gaps in understanding, critical engagement with information, verifiability, and accountability. Building on these principles, we introduce a conceptual framework for a dynamic shared representational workspace in which knowledge workers and specialized AI agents jointly gather evidence, schematize, hypothesize, and pursue collaborative goals. Through a partner agent, a shared space agent, and an orchestrator agent, the framework preserves the provenance and authorship of contributions and traces the evolution of both individual and shared interpretations, supporting coherent, negotiated knowledge construction that current generative AI systems tend to obscure.", "authors": ["Zhitong Guan", "Soo Young Rieh"], "categories": ["cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-23", "url": "https://arxiv.org/abs/2606.09840", "pdf_url": "https://arxiv.org/pdf/2606.09840v1", "arxiv_id": "2606.09840", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "93ae5a4081931371268f9b5b329c34e2dafd3268caad18b2aebbd2764bc72c7e", "sources": ["arxiv", "semantic_scholar"], "title": "How Consistent Are LLM Agents? Measuring Behavioral Reproducibility in Multi-Step Tool-Calling Pipelines", "abstract": "Large language model (LLM) agents with tool-calling capabilities are increasingly deployed in production systems, yet a fundamental reliability question remains under-explored: does the same agent behave the same way twice? We present a systematic empirical study of behavioral consistency in multi-step tool-calling agents, measuring whether agents select the same tools, in the same order, with the same arguments, across repeated identical invocations. Unlike prior work on consistency in ReAct-style agents(search-only, free-text actions), we study the richer setting of structured tool-calling interfaces with typed parameters and consequential side effects.", "authors": ["Abel Yagubyan"], "categories": ["cs.CL", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-23", "url": "https://arxiv.org/abs/2605.28840", "pdf_url": "https://arxiv.org/pdf/2605.28840v1", "arxiv_id": "2605.28840", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "9d431396fd170346adb8b524b70ce4517fd7cd679779481e542c6207d5e7ebbd", "sources": ["arxiv", "semantic_scholar"], "title": "AgenticQwen: Training Small Agentic Language Models with Dual Data Flywheels for Industrial-Scale Tool Use", "abstract": "Modern industrial applications increasingly demand language models that act as agents, capable of multi-step reasoning and tool use in real-world settings. These tasks are typically performed under strict cost and latency constraints, making small agentic models highly desirable. In this paper, we introduce the AgenticQwen family of models, trained via multi-round reinforcement learning (RL) on synthetic data and a limited amount of open-source data. Our training framework combines reasoning RL and agentic RL with dual data flywheels that automatically generate increasingly challenging tasks. The reasoning flywheel increases task difficulty by learning from errors, while the agentic flywheel expands linear workflows into multi-branch behavior trees that better reflect the decision complexity of real-world applications. We validate AgenticQwen on public benchmarks and in an industrial agent system. The models achieve strong performance on multiple agentic benchmarks, and in our industrial agent system, close the gap with much larger models on search and data analysis tasks. Model checkpoints and part of the synthetic data: https://huggingface.co/collections/alibaba-pai/agenticqwen. Data synthesis and RL training code: https://github.com/haruhi-sudo/data_synth_and_rl. The data synthesis pipeline is also integrated into EasyDistill: https://github.com/modelscope/easydistill.", "authors": ["Yuanjie Lyu", "Chengyu Wang", "Haonan Zheng", "Yuanhao Yue", "Junbing Yan", "Ming Wang", "Jun Huang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-23", "url": "https://arxiv.org/abs/2604.21590", "pdf_url": "https://arxiv.org/pdf/2604.21590v1", "arxiv_id": "2604.21590", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/haruhi-sudo/data_synth_and_rl", "venue": null, "quality_score": 0.65}
{"id": "fec643afdf5f6699e45168182c1af1c5e5d304a87a4395838e672adb99b41a45", "sources": ["arxiv", "semantic_scholar"], "title": "ActuBench: A Multi-Agent LLM Pipeline for Generation and Evaluation of Actuarial Reasoning Tasks", "abstract": "We present ActuBench, a multi-agent LLM pipeline for the automated generation and evaluation of advanced actuarial assessment items aligned with the International Actuarial Association (IAA) Education Syllabus. The pipeline separates four LLM roles by adapter: one agent drafts items, one constructs distractors, a third independently verifies both stages and drives bounded one-shot repair loops, and a cost-optimized auxiliary agent handles Wikipedia-note summarization and topic labelling. The items, per-model responses and complete leaderboard are published as a browsable web interface at https://actubench.de/en/, allowing readers and practitioners to inspect individual items without a repository checkout. We evaluate 50 language models from eight providers on two complementary benchmarks -- 100 empirically hardest multiple-choice items and 100 open-ended items scored by an LLM judge -- and report three headline findings. First, multi-agent verification is load-bearing: the independent verifier flags a majority of drafted items on first pass, most of which the one-shot repair loop resolves. Second, locally-hosted open-weights inference sits on the cost-performance Pareto front: a Gemma~4 model running on consumer hardware and a Cerebras-hosted 120B open-weights model dominate the near-zero-cost region, with the latter within one item of the top of the leaderboard. Third, MCQ and LLM-as-Judge rankings differ meaningfully: the MCQ scaffold inflates the performance ceiling, and Judge-mode evaluation is needed to discriminate at the frontier.", "authors": ["Jan-Philipp Schmidt"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20273", "pdf_url": "https://arxiv.org/pdf/2604.20273v1", "arxiv_id": "2604.20273", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "da0a640a5df9e929f4621ce99a9fe739ca42f445238e884b99ec6d35a383e156", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative Profiles Predict Multi-Agent LLM Team Performance in AI for Science Workflows", "abstract": "Multi-agent systems built from teams of large language models (LLMs) are increasingly deployed for collaborative scientific reasoning and problem-solving. These systems require agents to coordinate under shared constraints, such as GPUs or credit balances, where cooperative behavior matters. Behavioral economics provides a rich toolkit of games that isolate distinct cooperation mechanisms, yet it remains unknown whether a model's behavior in these stylized settings predicts its performance in realistic collaborative tasks. Here, we benchmark 35 open-weight LLMs across six behavioral economics games and show that game-derived cooperative profiles robustly predict downstream performance in AI-for-Science tasks, where teams of LLM agents collaboratively analyze data, build models, and produce scientific reports under shared budget constraints. Models that effectively coordinate games and invest in multiplicative team production (rather than greedy strategies) produce better scientific reports across three outcomes, accuracy, quality, and completion. These associations hold after controlling for multiple factors, indicating that cooperative disposition is a distinct, measurable property of LLMs not reducible to general ability. Our behavioral games framework thus offers a fast and inexpensive diagnostic for screening cooperative fitness before costly multi-agent deployment.", "authors": ["Shivani Kumar", "Adarsh Bharathwaj", "David Jurgens"], "categories": ["cs.CL", "cs.CY", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20658", "pdf_url": "https://arxiv.org/pdf/2604.20658v1", "arxiv_id": "2604.20658", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "05afad3fc28ba05d3e9182875b37b9b7208867eaa5b95e5a2bcbd61e6d63ac8a", "sources": ["arxiv", "semantic_scholar"], "title": "Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Models", "abstract": "The growth of agentic AI has drawn significant attention to function calling Large Language Models (LLMs), which are designed to extend the capabilities of AI-powered system by invoking external functions. Injection and jailbreaking attacks have been extensively explored to showcase the vulnerabilities of LLMs to user prompt manipulation. The expanded capabilities of agentic models introduce further vulnerabilities via their function calling interface. Recent work in LLM security showed that function calling can be abused, leading to data tampering and theft, causing disruptive behavior such as endless loops, or causing LLMs to produce harmful content in the style of jailbreaking attacks. This paper introduces a novel function hijacking attack (FHA) that manipulates the tool selection process of agentic models to force the invocation of a specific, attacker-chosen function. While existing attacks focus on semantic preference of the model for function-calling tasks, we show that FHA is largely agnostic to the context semantics and robust to the function sets, making it applicable across diverse domains. We further demonstrate that FHA can be trained to produce universal adversarial functions, enabling a single attacked function to hijack tool selection across multiple queries and payload configurations. We conducted experiments on 5 different models, including instructed and reasoning variants, reaching 70% to 100% ASR over the established BFCL dataset. Our findings further demonstrate the need for strong guardrails and security modules for agentic systems.", "authors": ["Yannis Belkhiter", "Giulio Zizzo", "Sergio Maffeis", "Seshu Tirupathi", "John D. Kelleher"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20994", "pdf_url": "https://arxiv.org/pdf/2604.20994v1", "arxiv_id": "2604.20994", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "5cc656c279a48c18f8942ca77fed0bbe812465f48c556859d7e7f7ceb13719bf", "sources": ["arxiv", "semantic_scholar"], "title": "Co-Evolving LLM Decision and Skill Bank Agents for Long-Horizon Tasks", "abstract": "Long horizon interactive environments are a testbed for evaluating agents skill usage abilities. These environments demand multi step reasoning, the chaining of multiple skills over many timesteps, and robust decision making under delayed rewards and partial observability. Games are a good testbed for evaluating agent skill usage in environments. Large Language Models (LLMs) offer a promising alternative as game playing agents, but they often struggle with consistent long horizon decision making because they lack a mechanism to discover, retain, and reuse structured skills across episodes. We present COSPLAY, a co evolution framework in which an LLM decision agent retrieves skills from a learnable skill bank to guide action taking, while an agent managed skill pipeline discovers reusable skills from the agents unlabeled rollouts to form a skill bank. Our framework improves both the decision agent to learn better skill retrieval and action generation, while the skill bank agent continually extracts, refines, and updates skills together with their contracts. Experiments across six game environments show that COSPLAY with an 8B base model achieves over 25.1 percent average reward improvement against four frontier LLM baselines on single player game benchmarks while remaining competitive on multi player social reasoning games.", "authors": ["Xiyang Wu", "Zongxia Li", "Guangyao Shi", "Alexander Duffy", "Tyler Marques", "Matthew Lyle Olson", "Tianyi Zhou", "Dinesh Manocha"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20987", "pdf_url": "https://arxiv.org/pdf/2604.20987v1", "arxiv_id": "2604.20987", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6b325e7e654a343a0e9f49de7a4b11cb9b41767e29a8302469e149e6bc7319a2", "sources": ["arxiv", "semantic_scholar"], "title": "R2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling", "abstract": "Function calling empowers large language models (LLMs) to interface with external tools, yet existing RL-based approaches suffer from misalignment between reasoning processes and tool-call decisions. We propose R2IF, a reasoning-aware RL framework for interpretable function calling, adopting a composite reward integrating format/correctness constraints, Chain-of-Thought Effectiveness Reward (CER), and Specification-Modification-Value (SMV) reward, optimized via GRPO. Experiments on BFCL/ACEBench show R2IF outperforms baselines by up to 34.62% (Llama3.2-3B on BFCL) with positive Average CoT Effectiveness (0.05 for Llama3.2-3B), enhancing both function-calling accuracy and interpretability for reliable tool-augmented LLM deployment.", "authors": ["Aijia Cheng", "Kailong Wang", "Ling Shi", "Yongxin Zhao"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20316", "pdf_url": "https://arxiv.org/pdf/2604.20316v2", "arxiv_id": "2604.20316", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "94b25e28e7b75daa20fa5273ca85a4b35185f8017b31a8130c0920600c9b41b5", "sources": ["arxiv", "semantic_scholar"], "title": "EvoAgent: An Evolvable Agent Framework with Skill Learning and Multi-Agent Delegation", "abstract": "This paper proposes EvoAgent - an evolvable large language model (LLM) agent framework that integrates structured skill learning with a hierarchical sub-agent delegation mechanism. EvoAgent models skills as multi-file structured capability units equipped with triggering mechanisms and evolutionary metadata, and enables continuous skill generation and optimization through a user-feedback-driven closed-loop process. In addition, by incorporating a three-stage skill matching strategy and a three-layer memory architecture, the framework supports dynamic task decomposition for complex problems and long-term capability accumulation. Experimental results based on real-world foreign trade scenarios demonstrate that, after integrating EvoAgent, GPT5.2 achieves significant improvements in professionalism, accuracy, and practical utility. Under a five-dimensional LLM-as-Judge evaluation protocol, the overall average score increases by approximately 28%. Further model transfer experiments indicate that the performance of an agent system depends not only on the intrinsic capabilities of the underlying model, but also on the degree of synergy between the model and the agent architecture.", "authors": ["Aimin Zhang", "Jiajing Guo", "Fuwei Jia", "Chen Lv", "Boyu Wang", "Fangzheng Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-22", "url": "https://arxiv.org/abs/2604.20133", "pdf_url": "https://arxiv.org/pdf/2604.20133v2", "arxiv_id": "2604.20133", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "6a03c360464aba26d9f6f86c3a931d3e74f3194fb72906e9a097631939925d84", "sources": ["arxiv", "semantic_scholar"], "title": "Mesh Memory Protocol: Semantic Infrastructure for Multi-Agent LLM Systems", "abstract": "Teams of LLM agents increasingly collaborate on tasks spanning days or weeks: multi-day data-generation sprints where generator, reviewer, and auditor agents coordinate in real time on overlapping batches; specialists carrying findings forward across session restarts; product decisions compounding over many review rounds. This requires agents to share, evaluate, and combine each other's cognitive state in real time across sessions. We call this cross-session agent-to-agent cognitive collaboration, distinct from parallel agent execution. To enable it, three problems must be solved together. (P1) Each agent decides field by field what to accept from peers, not accept or reject whole messages. (P2) Every claim is traceable to source, so returning claims are recognised as echoes of the receiver's own prior thinking. (P3) Memory that survives session restarts is relevant because of how it was stored, not how it is retrieved. These are protocol-level properties at the semantic layer of agent communication, distinct from tool-access and task-delegation protocols at lower layers. We call this missing protocol layer \"semantic infrastructure,\" and the Mesh Memory Protocol (MMP) specifies it. Four composable primitives work together: CAT7, a fixed seven-field schema for every Cognitive Memory Block (CMB); SVAF, which evaluates each field against the receiver's role-indexed anchors and realises P1; inter-agent lineage, carried as parents and ancestors of content-hash keys and realising P2; and remix, which stores only the receiver's own role-evaluated understanding of each accepted CMB, never the raw peer signal, realising P3. MMP is specified, shipped, and running in production across three reference deployments, where each session runs an autonomous agent as a mesh peer with its own identity and memory, collaborating with other agents across the network for collective intelligence.", "authors": ["Hongwei Xu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-21", "url": "https://arxiv.org/abs/2604.19540", "pdf_url": "https://arxiv.org/pdf/2604.19540v1", "arxiv_id": "2604.19540", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "a8482b7e3f24d148359a865d6c1c9482d6edd81d5e5bc55e1cfc7dcd8c9fe21b", "sources": ["arxiv", "semantic_scholar"], "title": "TriEx: A Game-based Tri-View Framework for Explaining Internal Reasoning in Multi-Agent LLMs", "abstract": "Explainability for Large Language Model (LLM) agents is especially challenging in interactive, partially observable settings, where decisions depend on evolving beliefs and other agents. We present \\textbf{TriEx}, a tri-view explainability framework that instruments sequential decision making with aligned artifacts: (i) structured first-person self-reasoning bound to an action, (ii) explicit second-person belief states about opponents updated over time, and (iii) third-person oracle audits grounded in environment-derived reference signals. This design turns explanations from free-form narratives into evidence-anchored objects that can be compared and checked across time and perspectives. Using imperfect-information strategic games as a controlled testbed, we show that TriEx enables scalable analysis of explanation faithfulness, belief dynamics, and evaluator reliability, revealing systematic mismatches between what agents say, what they believe, and what they do. Our results highlight explainability as an interaction-dependent property and motivate multi-view, evidence-grounded evaluation for LLM agents. Code is available at https://github.com/Einsam1819/TriEx.", "authors": ["Ziyi Wang", "Chen Zhang", "Wenjun Peng", "Qi Wu", "Xinyu Wang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-21", "url": "https://arxiv.org/abs/2604.20043", "pdf_url": "https://arxiv.org/pdf/2604.20043v1", "arxiv_id": "2604.20043", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Einsam1819/TriEx", "venue": null, "quality_score": 0.65}
{"id": "3aefc100e86422ec9020131ff9316cd0598818ed56e25cc76579ccdc73544703", "sources": ["arxiv", "semantic_scholar"], "title": "Explicit Trait Inference for Multi-Agent Coordination", "abstract": "LLM-based multi-agent systems (MAS) show promise on complex tasks but remain prone to coordination failures such as goal drift, error cascades, and misaligned behaviors. We propose Explicit Trait Inference (ETI), a psychologically grounded method for improving coordination. ETI enables agents to infer and track partner characteristics along two established psychological dimensions--warmth (e.g., trust) and competence (e.g., skill)--from interaction histories to guide decisions. We evaluate ETI in controlled settings (economic games), where it reduces payoff loss by 45-77%, and in more realistic, complex multi-agent settings (MultiAgentBench), where it improves performance by 3-29% depending on the scenario and model, relative to a CoT baseline. Additional analysis shows that gains are closely linked to trait inference: ETI profiles predict agents' actions, and informative profiles drive improvements. These results highlight ETI as a lightweight and robust mechanism for improving coordination in diverse multi-agent settings, and provide the first systematic evidence that LLM agents can (i) reliably infer others' traits from interaction histories and (ii) leverage structured awareness of others' traits for coordination.", "authors": ["Suhaib Abdurahman", "Etsuko Ishii", "Katerina Margatina", "Divya Bhargavi", "Monica Sunkara", "Yi Zhang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-21", "url": "https://arxiv.org/abs/2604.19278", "pdf_url": "https://arxiv.org/pdf/2604.19278v2", "arxiv_id": "2604.19278", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "8834a4f553ec86a295ce5a4e2b96ce84e4e46edfdf960a1597bc1c888898039b", "sources": ["arxiv", "semantic_scholar"], "title": "SelfHeal: Empirical Fix Pattern Analysis and Bug Repair in LLM Agents", "abstract": "Large Language Models (LLMs) have transformed software development and AI applications. While LLMs are designed for text processing, LLM agents extend this capability by enabling autonomous actions, tool use, and multi-step task completion. As this field grows, developers face new challenges in debugging these complex systems. To address this challenge, we present the first empirical study on bug fix patterns in LLM agents. We study buggy posts and code snippets from three platforms: Stack Overflow, GitHub, and HuggingFace Forums. We examine their fix patterns, the components where fixes are applied, and the programming languages and frameworks involved. Furthermore, we introduce AgentDefect, the first benchmark dataset for bugs in LLM agents. The dataset contains 37 runtime buggy instances along with fixed code and test files. Finally, we present SelfHeal, a multi-agent system designed to fix bugs in LLM agents. The system leverages two independent ReAct agents: the fix agent and the critic agent. These agents use tools that provide both internal knowledge (fix rules) and external knowledge (web search) to propose and validate fixes. Our evaluation shows that SelfHeal with Gemini 3 Pro as the backbone LLM outperforms both baseline and state-of-the-art approaches by a significant margin.", "authors": ["Niful Islam", "Muhammad Anas Raza", "Mohammad Wardat"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.17699", "pdf_url": "https://arxiv.org/pdf/2604.17699v1", "arxiv_id": "2604.17699", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f684ca2ea905c5dc9ae1a5f161ae878ae435b17ac5142dd1171dbfe846c830bb", "sources": ["arxiv", "semantic_scholar"], "title": "Prompt Optimization Enables Stable Algorithmic Collusion in LLM Agents", "abstract": "LLM agents in markets present algorithmic collusion risks. While prior work shows LLM agents reach supracompetitive prices through tacit coordination, existing research focuses on hand-crafted prompts. The emerging paradigm of prompt optimization necessitates new methodologies for understanding autonomous agent behavior. We investigate whether prompt optimization leads to emergent collusive behaviors in market simulations. We propose a meta-learning loop where LLM agents participate in duopoly markets and an LLM meta-optimizer iteratively refines shared strategic guidance. Our experiments reveal that meta-prompt optimization enables agents to discover stable tacit collusion strategies with substantially improved coordination quality compared to baseline agents. These behaviors generalize to held-out test markets, indicating discovery of general coordination principles. Analysis of evolved prompts reveals systematic coordination mechanisms through stable shared strategies. Our findings call for further investigation into AI safety implications in autonomous multi-agent systems.", "authors": ["Yingtao Tian"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.17774", "pdf_url": "https://arxiv.org/pdf/2604.17774v1", "arxiv_id": "2604.17774", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "d2139f78bcd9ea24fad3b576ef1fbd2c269474ca0e0654f78c0a2a022631fede", "sources": ["arxiv", "semantic_scholar"], "title": "Latent Preference Modeling for Cross-Session Personalized Tool Calling", "abstract": "Users often omit essential details in their requests to LLM-based agents, resulting in under-specified inputs for tool use. This poses a fundamental challenge for tool-augmented agents, as API execution typically requires complete arguments, highlighting the need for personalized tool calling. To study this problem, we introduce MPT, a benchmark comprising 265 multi-session dialogues that cover three challenges: Preference Recall, Preference Induction, and Preference Transfer. We also propose PRefine, a test-time memory-augmented method that represents user preferences as evolving hypotheses. Through a generate--verify--refine loop, it extracts reusable constraints from history and improves tool-calling accuracy while using only 1.24% of the tokens required by full-history prompting. These results indicate that robust personalization in agentic systems depends on memory that captures the reasons behind user choices, not just the choices themselves.", "authors": ["Yejin Yoon", "Minseo Kim", "Taeuk Kim"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.17886", "pdf_url": "https://arxiv.org/pdf/2604.17886v1", "arxiv_id": "2604.17886", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "84152a77e2e713c9f753982828894a651ba67f2f788fbb6172a6314b07500a37", "sources": ["arxiv", "semantic_scholar"], "title": "MultiWorld: Scalable Multi-Agent Multi-View Video World Models", "abstract": "Video world models have achieved remarkable success in simulating environmental dynamics in response to actions by users or agents. They are modeled as action-conditioned video generation models that take historical frames and current actions as input to predict future frames. Yet, most existing approaches are limited to single-agent scenarios and fail to capture the complex interactions inherent in real-world multi-agent systems. We present \\textbf{MultiWorld}, a unified framework for multi-agent multi-view world modeling that enables accurate control of multiple agents while maintaining multi-view consistency. We introduce the Multi-Agent Condition Module to achieve precise multi-agent controllability, and the Global State Encoder to ensure coherent observations across different views. MultiWorld supports flexible scaling of agent and view counts, and synthesizes different views in parallel for high efficiency. Experiments on multi-player game environments and multi-robot manipulation tasks demonstrate that MultiWorld outperforms baselines in video fidelity, action-following ability, and multi-view consistency. Project page: https://multi-world.github.io/", "authors": ["Haoyu Wu", "Jiwen Yu", "Yingtian Zou", "Xihui Liu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.18564", "pdf_url": "https://arxiv.org/pdf/2604.18564v2", "arxiv_id": "2604.18564", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "077c7df9daa92be7b7952e4ef4664f7dcf5b5a5ad6955191116f8e81c6a60483", "sources": ["arxiv", "semantic_scholar"], "title": "CHICO-Agent: An LLM Agent for the Cross-layer Optimization of 2.5D and 3D Chiplet-based Systems", "abstract": "The rapid growth of large language models (LLMs) and AI workloads has pushed monolithic silicon to its reticle and economic limits, accelerating the adoption of 2.5D/3D chiplet systems. However, these systems increase design complexity by requiring co-design across multiple levels of the computing stack, including application, architecture, chip, and package. The resulting design space is highly combinatorial, with trade-offs among latency, energy, area, and cost. To address this challenge, we propose CHICO-Agent, an LLM-driven optimization framework for 2.5D/3D chiplet-based systems. CHICO-Agent maintains a persistent knowledge base to capture parameter-outcome trends and coordinates exploration through an admin-field multi-agent workflow. Compared with a simulated-annealing baseline, CHICO-Agent finds lower-cost configurations and provides an interpretable audit trail for designers.", "authors": ["Qihang Wu", "Aman Arora", "Vidya A. Chhabria"], "categories": ["cs.AR"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.18764", "pdf_url": "https://arxiv.org/pdf/2604.18764v1", "arxiv_id": "2604.18764", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "96cad12c332c044e88df61e682d81bf15bc6dc65be6518812ca5b8641d62eff1", "sources": ["arxiv", "semantic_scholar"], "title": "JTPRO: A Joint Tool-Prompt Reflective Optimization Framework for Language Agents", "abstract": "Large language model (LLM) agents augmented with external tools often struggle as number of tools grow large and become domain-specific. In such settings, ambiguous tool descriptions and under-specified agent instructions frequently lead to tool mis-selection and incorrect slot/value instantiation. We hypothesize that this is due to two root causes: generic, one-size-fits-all prompts that ignore tool-specific nuances, and underspecified tool schemas that lack clear guidance on when and how to use each tool and how to format its parameters. We introduce Joint Tool-Prompt Reflective Optimization (JTPRO), a framework for improving tool-calling reliability in trace-supervised settings by iteratively using rollout-driven reflection to co-optimize global instructions and per-tool schema/argument descriptions for accurate tool selection and argument instantiation in large tool inventories. JTPRO is designed to preserve only tool-local cues needed for correct disambiguation and slot filling. We evaluate JTPRO across multi-tool benchmarks, which account for different number of tools using three metrics: Tool Selection Accuracy (TSA), Slot Filling Accuracy(SFA), and Overall Success Rate(OSR) (correct tool + correct slots + correct values). JTPRO consistently outperforms strong baselines, including CoT-style agents, and reflective prompt optimizers such as GEPA by 5%-20% (relative) on OSR. Ablations show that joint optimization of instructions and tool schemas is more effective and robust than optimizing either component in isolation.", "authors": ["Sandip Ghoshal", "Anshul Mittal", "Jyotika Singh", "Miguel Ballesteros", "Weiyi Sun", "Fang Tu", "Shailender Singh", "Yassine Benajiba", "Fahad Shah", "Sujeeth Bharadwaj", "Sujith Ravi", "Dan Roth"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-20", "url": "https://arxiv.org/abs/2604.19821", "pdf_url": "https://arxiv.org/pdf/2604.19821v1", "arxiv_id": "2604.19821", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "1465587bb4440bd3586475006a29fa1f7db95d2fd5fe7835e3f47b3034fe5ca5", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Social Affordance Framework (ASAF): Agent Identity Design as a Collaboration Interface in Multi-Agent Systems", "abstract": "As AI systems evolve from single conversational agents to complex multi-agent architectures, a critical design dimension has been overlooked: how the social identity of individual agents shapes human behavior within the collaboration. This paper introduces the Agentic Social Affordance Framework (ASAF), a theoretical framework that extends Social Affordance theory into the context of multi-agent AI systems. We propose that agent identity design functions not merely as a user interface convention, but as a collaboration interface -- structuring how users perceive, approach, and engage with each agent, and thereby influencing the quality of Human-Agent collaboration outcomes. Specifically, the social affordance layer constitutes an independent design dimension orthogonal to engineering orchestration: the two represent distinct decision spaces that cannot be derived from each other. ASAF comprises three mechanisms: Identity Signaling, Behavioral Priming, and Collaborative Governance, and specifies their boundary conditions through a four-tier Identity Signal Fidelity Spectrum and an individual-difference moderating variable (anthropomorphizing vs.\\ instrumentalizing cognitive style). We situate ASAF in relation to existing affordance theory and the CASA paradigm, delineating where ASAF's multi-agent, topology-level predictions exceed the explanatory scope of dyadic frameworks. We discuss implications for multi-agent system design and outline directions for future empirical validation, including a factorial design for testing design-space orthogonality.", "authors": ["Meng-Han Lee"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-19", "url": "https://arxiv.org/abs/2606.09832", "pdf_url": "https://arxiv.org/pdf/2606.09832v1", "arxiv_id": "2606.09832", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e3f925414672afe646b25d6133784b891de4f25bfc77cd879f86b5713a15c68e", "sources": ["arxiv", "semantic_scholar"], "title": "Signal or Noise in Multi-Agent LLM-based Stock Recommendations?", "abstract": "We present the first portfolio-level validation of MarketSenseAI, a deployed multi-agent LLM equity system. All signals are generated live at each observation date, eliminating look-ahead bias. The system routes four specialist agents (News, Fundamentals, Dynamics, and Macro) through a synthesis agent that issues a monthly equity thesis and recommendation for each stock in its coverage universe, and we ask two questions: do its buy recommendations add value over both passive benchmarks and random selection, and what does the internal agent structure reveal about the source of the edge? On the S&P 500 cohort (19 months) the strong-buy equal-weight portfolio earns +2.18%/month against a passive equal-weight benchmark of +1.15% (approximating RSP), a +25.2% compound excess, and ranks at the 99.7th percentile of 10,000 Monte Carlo portfolios (p=0.003). The S&P 100 cohort (35 months) delivers a +30.5% compound excess over EQWL with consistent direction but formal significance not reached, limited by the small average selection of ~10 stocks per month. Non-negative least-squares projection of thesis embeddings onto agent embeddings reveals an adaptive-integration mechanism. Agent contributions rotate with market regime (Fundamentals leads on S&P 500, Macro on S&P 100, Dynamics acts as an episodic momentum signal) and this agent rotation moves in lockstep with both the sector composition of strong-buy selections and identifiable macro-calendar events, three independent views of the same underlying adaptation. The recommendation's cross-sectional Information Coefficient is statistically significant on S&P 500 (ICIR=+0.489, p=0.024). These results suggest that multi-agent LLM equity systems can identify sources of alpha beyond what classical factor models capture, and that the buy signal functions as an effective universe-filter that can sit upstream of any portfolio-construction process.", "authors": ["George Fatouros", "Kostas Metaxas"], "categories": ["q-fin.PM", "cs.AI", "q-fin.ST"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2026-04-19", "url": "https://arxiv.org/abs/2604.17327", "pdf_url": "https://arxiv.org/pdf/2604.17327v1", "arxiv_id": "2604.17327", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "0c2788b0ff02383a2a2d1ab07508ef292d681204e72d2086bf2fda2d312e6bf7", "sources": ["arxiv", "semantic_scholar"], "title": "Do LLM-derived graph priors improve multi-agent coordination?", "abstract": "Multi-agent reinforcement learning (MARL) is crucial for AI systems that operate collaboratively in distributed and adversarial settings, particularly in multi-domain operations (MDO). A central challenge in cooperative MARL is determining how agents should coordinate: existing approaches must either hand-specify graph topology, rely on proximity-based heuristics, or learn structure entirely from environment interaction; all of which are brittle, semantically uninformed, or data-intensive. We investigate whether large language models (LLMs) can generate useful coordination graph priors for MARL by using minimal natural language descriptions of agent observations to infer latent coordination patterns. These priors are integrated into MARL algorithms via graph convolutional layers within a graph neural network (GNN)-based pipeline, and evaluated on four cooperative scenarios from the Multi-Agent Particle Environment (MPE) benchmark against baselines spanning the full spectrum of coordination modeling, from independent learners to state-of-the-art graph-based methods. We further ablate across five compact open-source LLMs to assess the sensitivity of prior quality to model choice. Our results provide the first quantitative evidence that LLM-derived graph priors can enhance coordination and adaptability in dynamic multi-agent environments, and demonstrate that models as small as 1.5B parameters are sufficient for effective prior generation.", "authors": ["Nikunj Gupta", "Rajgopal Kannan", "Viktor Prasanna"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-19", "url": "https://arxiv.org/abs/2604.17191", "pdf_url": "https://arxiv.org/pdf/2604.17191v1", "arxiv_id": "2604.17191", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "cfe3b01a20b658fe5663cedde82ff1f9a6e1e2033130a6393086cd271dc8bae0", "sources": ["arxiv", "semantic_scholar"], "title": "Provable Coordination for LLM Agents via Message Sequence Charts", "abstract": "Multi-agent systems built on large language models (LLMs) are difficult to reason about. Coordination errors such as deadlocks or type-mismatched messages are often hard to detect through testing. We introduce a domain-specific language for specifying agent coordination based on message sequence charts (MSCs). The language separates message-passing structure from LLM actions, whose outputs remain unpredictable. We define the syntax and semantics of the language and present a syntax-directed projection that generates deadlock-free local agent programs from global coordination specifications. We illustrate the approach with a diagnosis consensus protocol and show how coordination properties can be established independently of LLM nondeterminism. We also describe a runtime planning extension in which an LLM dynamically generates a coordination workflow for which the same structural guarantees apply. An open-source Python implementation of our framework is available as ZipperGen.", "authors": ["Benedikt Bollig", "Matthias Függer", "Thomas Nowak"], "categories": ["cs.PL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-19", "url": "https://arxiv.org/abs/2604.17612", "pdf_url": "https://arxiv.org/pdf/2604.17612v2", "arxiv_id": "2604.17612", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.65}
{"id": "f93e50e7d97d4d9f99a3e61be22a039e1cda7db4ff7aea192045d58b189101e7", "sources": ["arxiv", "semantic_scholar"], "title": "Graph-of-Agents: A Graph-based Framework for Multi-Agent LLM Collaboration", "abstract": "With an ever-growing zoo of LLMs and benchmarks, the need to orchestrate multiple models for improved task performance has never been more pressing. While frameworks like Mixture-of-Agents (MoA) attempt to coordinate LLMs, they often fall short in terms of (1) selecting relevant agents, (2) facilitating effective intra-agent communication, and (3) integrating responses efficiently. In this work, we propose Graph-of-Agents (GoA), a new graph-based framework for modeling multi-agent LLM communication. Our approach begins with node sampling, selecting only the most relevant agents by leveraging model cards that summarize each model's domain, task specialization, and other characteristics. Next, we construct edges between the selected agents by evaluating their responses against one another to determine relevance ordering. Directed message passing is then performed from highly relevant agents to less relevant ones to enhance their responses, followed by reverse message passing to refine the original responses of the more relevant agents. Finally, the updated responses are aggregated via graph-based pooling (e.g., max or mean pooling) to produce a single, unified answer. We evaluate GoA on diverse multi-domain benchmarks (MMLU, MMLU-Pro, GPQA) and domain-specific benchmarks (MATH, HumanEval, MedMCQA), with an agent pool of 6 LLMs spanning multiple domains. Surprisingly, GoA achieves superior performance using only 3 selected agents, outperforming recent multi-agent LLM baselines that utilize all 6 agents simultaneously. By adopting a graph structure, GoA offers both scalability and effectiveness through structured message passing-positioning it as a strong candidate for navigating the challenges of the ever-growing LLM zoo. Code is available at: https://github.com/UNITES-Lab/GoA.", "authors": ["Sukwon Yun", "Jie Peng", "Pingzhi Li", "Wendong Fan", "Jie Chen", "James Zou", "Guohao Li", "Tianlong Chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-18", "url": "https://arxiv.org/abs/2604.17148", "pdf_url": "https://arxiv.org/pdf/2604.17148v1", "arxiv_id": "2604.17148", "doi": null, "citation_count": 9, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/UNITES-Lab/GoA", "venue": null, "quality_score": 0.65}
{"id": "3ca10cf9093ad509efedbeba34fd150d576730746c6cfe3a2909209ac2f7f655", "sources": ["arxiv", "semantic_scholar"], "title": "StoryMI: Steerable Multi-Agent Therapeutic Dialogue Generation", "abstract": "Large language models (LLMs) can generate fluent dialogue, but prior works lack situational grounding, dynamic strategy control, and evaluation aligned with clinical standards in motivational interviewing (MI). We introduce StoryMI, a multi-LLM agent framework for controllable MI dialogue generation, where questionnaire-based client profiles are expanded into situational stories that provide narrative context for the dialogue. Therapist and client agents generate MI-coded utterances guided by MI codes selected by the interaction agent, while an interaction agent dynamically coordinates exchanges to control MI strategies during a multi-turn conversation. We propose a two-level evaluation protocol: lexical metrics and MI-specific measures of macro-level counseling strategies, alongside LLM-as-judge and human expert assessments. We construct a dataset of 6K simulated MI dialogues grounded in 1K questionnaire-story pairs, covering 12 MI codes and 13 symptom domains, and benchmark six open- and closed-source LLMs. Our results show that situational grounding and macro-level control can improve MI adherence and clinical plausibility, demonstrating the effectiveness of a structured multi-agent workflow for psychotherapy dialogue generation. We provide code and data for reproducibility.", "authors": ["Qingyu Meng", "Min Chen", "Dingming Liu", "Yifan Mo", "Yue Su", "Xin Sun", "Koen Hindriks", "Jiahuan Pei"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-18", "url": "https://arxiv.org/abs/2605.27393", "pdf_url": "https://arxiv.org/pdf/2605.27393v1", "arxiv_id": "2605.27393", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "e9e24b1f1d8a06e406c4778c8bd1146b6560f2bab4abff414c349517dad4416d", "sources": ["arxiv", "semantic_scholar"], "title": "Complete Cyclic Subtask Graphs for Tool-Using LLM Agents: Flexibility, Cost, and Bottlenecks in Multi-Agent Workflows", "abstract": "Long-horizon tool-using tasks sometimes benefit from revisiting earlier subtasks for recovery and exploration, but added multi-agent workflow flexibility can also introduce coordination overhead and substantial inference cost. We study complete cyclic subtask graphs, a deliberately maximally flexible multi-agent architecture in which executable subtask nodes are fully connected and a unified state-analysis-and-routing agent selects transitions using natural-language criteria. This makes unrestricted revisitation explicit and directly analyzable at the subtask level. We evaluate task-specific (Spec-Cyc) and benchmark-generic (Gen-Cyc) graphs on TextCraft, ALFWorld, and Finance-Agent, with ablations over planner/executor/router strength, tool exposure (generalist vs specialized), $n$-shot successful trajectory summaries, and fault-injected random subtask perturbations. The benchmarks expose three distinct regimes. ALFWorld highlights a setting where explicit revisitation supports recovery and exploration; TextCraft, a largely prerequisite-chain domain, often favors the efficiency of simpler forward execution; and Finance-Agent remains bottlenecked by retrieval, grounding, and evidence synthesis more than by workflow flexibility alone. Shared-win token comparisons further show that the added flexibility can be substantially more expensive than a single ReAct agent. Overall, we use complete cyclic subtask graphs as a maximally flexible experimental lens for measuring when multi-agent revisitation helps, when it mainly adds coordination cost, and when external task bottlenecks dominate.", "authors": ["Luay Gharzeddine", "Samer Saab"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-17", "url": "https://arxiv.org/abs/2604.22820", "pdf_url": "https://arxiv.org/pdf/2604.22820v1", "arxiv_id": "2604.22820", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "f528b7bddd63e6314a0dff204792285e70fb554e6f49365121aa977e9d4849ec", "sources": ["arxiv", "semantic_scholar"], "title": "Conjunctive Prompt Attacks in Multi-Agent LLM Systems", "abstract": "Most LLM safety work studies single-agent models, but many real applications rely on multiple interacting agents. In these systems, prompt segmentation and inter-agent routing create attack surfaces that single-agent evaluations miss. We study \\emph{conjunctive prompt attacks}, where a trigger key in the user query and a hidden adversarial template in one compromised remote agent each appear benign alone but activate harmful behavior when routing brings them together. We consider an attacker who changes neither model weights nor the client agent and instead controls only trigger placement and template insertion. Across star, chain, and DAG topologies, routing-aware optimization substantially increases attack success over non-optimized baselines while keeping false activations low. Existing defenses, including PromptGuard, Llama-Guard variants, and system-level controls such as tool restrictions, do not reliably stop the attack because no single component appears malicious in isolation. These results expose a structural vulnerability in agentic LLM pipelines and motivate defenses that reason over routing and cross-agent composition. Code is available at https://github.com/UCF-ML-Research/ConjunctiveAgents.", "authors": ["Nokimul Hasan Arif", "Qian Lou", "Mengxin Zheng"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-17", "url": "https://arxiv.org/abs/2604.16543", "pdf_url": "https://arxiv.org/pdf/2604.16543v1", "arxiv_id": "2604.16543", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/UCF-ML-Research/ConjunctiveAgents", "venue": null, "quality_score": 0.65}
{"id": "e3a6a54465c394c51fab0156a13d48bf2c830cbc4fa7c7412b717c296be41b2b", "sources": ["arxiv", "semantic_scholar"], "title": "MemEvoBench: Benchmarking Safety Risks from Memory Misevolution in LLM Agents", "abstract": "Equipping Large Language Models (LLMs) with persistent memory enhances interaction continuity and personalization but introduces new safety risks. Specifically, contaminated or biased memory accumulation can trigger abnormal agent behaviors. Existing evaluation methods have not yet established a standardized framework for measuring memory misevolution. This phenomenon refers to the gradual behavioral drift resulting from repeated exposure to misleading information. To address this gap, we introduce MemEvoBench, the first benchmark evaluating long-horizon memory safety in LLM agents against adversarial memory injection, noisy tool outputs, and biased feedback. The framework consists of QA-style tasks across 7 domains and 36 risk types, complemented by workflow-style tasks adapted from 20 Agent-SafetyBench environments with noisy tool returns. Both settings employ mixed benign and misleading memory pools within multi-round interactions to simulate memory evolution. Experiments on representative models reveal substantial safety degradation under biased memory updates. Our analysis suggests that memory evolution is a significant contributor to these failures. Furthermore, static prompt-based defenses prove insufficient, underscoring the urgency of securing memory evolution in LLM agents.", "authors": ["Weiwei Xie", "Shaoxiong Guo", "Fan Zhang", "Tian Xia", "Xue Yang", "Lizhuang Ma", "Junchi Yan", "Qibing Ren"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-17", "url": "https://arxiv.org/abs/2604.15774", "pdf_url": "https://arxiv.org/pdf/2604.15774v2", "arxiv_id": "2604.15774", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "bc8d85884f35a604ab311786f34b72432d0ecdc49847e7c6e609d40759a46592", "sources": ["arxiv", "semantic_scholar"], "title": "Weak-Link Optimization for Multi-Agent Reasoning and Collaboration", "abstract": "LLM-driven multi-agent frameworks address complex reasoning tasks through multi-role collaboration. However, existing approaches often suffer from reasoning instability, where individual agent errors are amplified through collaboration, undermining overall performance. Current research mainly focuses on enhancing high-capability agents or suppressing unreliable outputs to improve framework effectiveness, while systematic identification and reinforcement of performance-limiting agents receive less attention. To address this gap, we propose WORC, a \\underline{w}eak-link \\underline{o}ptimization framework for multi-agent \\underline{r}easoning and \\underline{c}ollaboration, grounded in the weak-link principle. WORC follows a two-stage workflow. In the weak agent localization stage, task features are constructed, and a meta-learning-based weight predictor trained on optimal configurations identified by swarm intelligence algorithms (SIAs) enables zero-shot mapping from these features to agent performance weights, where the agent with the lowest predicted weight is identified as the weak agent. In the weak-link optimization stage, an uncertainty-driven allocation strategy assigns additional reasoning budgets to weak agents, with lower predicted weights leading to larger repeated-sampling quotas to compensate for reliability deficiencies. Experimental results show that WORC achieves an average accuracy of 82.2\\% on reasoning benchmarks while improving framework stability and cross-architecture generalization, suggesting that compensating for weak links, rather than reinforcing strengths alone, enhances the robustness of multi-agent systems.", "authors": ["Haoyu Bian", "Chaoning Zhang", "Jiaquan Zhang", "Xingyao Li", "Yuanfang Guo", "Wei Dong", "Yang Yang"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-17", "url": "https://arxiv.org/abs/2604.15972", "pdf_url": "https://arxiv.org/pdf/2604.15972v1", "arxiv_id": "2604.15972", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.35}
{"id": "b76216d4cc69720bc7f4f828243cfbaf4685da69d5d4b7322a070f6aa2b4877a", "sources": ["arxiv", "semantic_scholar"], "title": "CAMP: Cumulative Agentic Masking and Pruning for Privacy Protection in Multi-Turn LLM Conversations", "abstract": "The deployment of Large Language Models in agentic, multi-turn conversational settings has introduced a class of privacy vulnerabilities that existing protection mechanisms are not designed to address. Current approaches to Personally Identifiable Information (PII) masking operate on a per-turn basis, scanning each user message in isolation and replacing detected entities with typed placeholders before forwarding sanitized text to the model. While effective against direct identifier leakage within a single message, these methods are fundamentally stateless and fail to account for the compounding privacy risk that emerges when PII fragments accumulate across conversation turns. A user who separately discloses their name, employer, location, and medical condition across several messages has revealed a fully re-identifiable profile - yet no individual message would trigger a per-turn masker. We formalize this phenomenon as Cumulative PII Exposure (CPE) and propose CAMP (Cumulative Agentic Masking and Pruning), a cross-turn privacy protection framework for multi-turn LLM conversations. CAMP maintains a session-level PII registry, constructs a co-occurrence graph to model combination risk between entity types, computes a CPE score after each turn, and triggers retroactive masking of conversation history when the score crosses a configurable threshold. We evaluate CAMP on four synthetic multi-turn scenarios spanning healthcare, hiring, finance, and general conversation, demonstrating that per-turn baselines expose re-identifiable profiles that CAMP successfully neutralizes while preserving full conversational utility.", "authors": ["Aman Panjwani"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-16", "url": "https://arxiv.org/abs/2604.16521", "pdf_url": "https://arxiv.org/pdf/2604.16521v1", "arxiv_id": "2604.16521", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3493}
{"id": "6265cc70f17b456bf9b47d508704b719c0fe9f11c15019da9889082c82e4cdf6", "sources": ["arxiv", "semantic_scholar"], "title": "Scepsy: Serving Agentic Workflows Using Aggregate LLM Pipelines", "abstract": "Agentic workflows carry out complex tasks by orchestrating multiple large language models (LLMs) and tools. Serving such workflows at a target throughput with low latency is challenging because they can be defined using arbitrary agentic frameworks and exhibit unpredictable execution times: execution may branch, fan-out, or recur in data-dependent ways. Since LLMs in workflows often outnumber available GPUs, their execution also leads to GPU oversubscription. We describe Scepsy, a new agentic serving system that efficiently schedules arbitrary multi-LLM agentic workflows onto a GPU cluster. Scepsy exploits the insight that, while agentic workflows have unpredictable end-to-end latencies, the shares of each LLM's total execution times are comparatively stable across executions. Scepsy decides on GPU allocations based on these aggregate shares: first, it profiles the LLMs under different parallelism degrees. It then uses these statistics to construct an Aggregate LLM Pipeline, which is a lightweight latency/throughput predictor for allocations. To find a GPU allocation that minimizes latency while achieving a target throughput, Scepsy uses the Aggregate LLM Pipeline to explore a search space over fractional GPU shares, tensor parallelism degrees, and replica counts. It uses a hierarchical heuristic to place the best allocation onto the GPU cluster, minimizing fragmentation, while respecting network topology constraints. Our evaluation on realistic agentic workflows shows that Scepsy achieves up to 2.4x higher throughput and 27x lower latency compared to systems that optimize LLMs independently or rely on user-specified allocations.", "authors": ["Marcel Wagenländer", "Otto White", "Britannio Jarrett", "Pedro Silvestre", "Yanda Tao", "Guo Li", "Huanzhou Zhu", "Llúis Vilanova", "Peter Pietzuch"], "categories": ["cs.DC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-16", "url": "https://arxiv.org/abs/2604.15186", "pdf_url": "https://arxiv.org/pdf/2604.15186v1", "arxiv_id": "2604.15186", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3493}
{"id": "df539509b8de10109eee15b22ff9cc76461077d1a6dad0fbfffac881c7681a15", "sources": ["arxiv", "semantic_scholar"], "title": "Autonomous Evolution of EDA Tools: Multi-Agent Self-Evolved ABC", "abstract": "This paper introduces the first \\emph{self-evolving} logic synthesis framework, which leverages Large Language Model (LLM) agents to autonomously improve the source code of \\textsc{ABC}, the widely adopted logic synthesis system. Our framework operates on the \\emph{entire integrated ABC codebase}, and the output repository preserves its single-binary execution model and command interface. In the initial evolution cycle, we bootstrap the system using existing prior open-source synthesis components, covering flow tuning, logic minimization, and technology mapping, but without manually injecting new heuristics. On top of this foundation, a team of LLM-based agents iteratively rewrites and evolves specific sub-components of ABC following our ``programming guidance`` prompts under a unified correctness and QoR-driven evaluation loop. Each evolution cycle proposes code modifications, compiles the integrated binary, validates correctness, and evaluates quality-of-results (QoR) on \\emph{multi-suite benchmarks including ISCAS~85/89/99, VTR, EPFL, and IWLS~2005}. Through continuous feedback, the system discovers optimizations beyond human-designed heuristics, effectively \\emph{learning new synthesis strategies} that enhance QoR. We detail the architecture of this self-improving system, its integration with \\textsc{ABC}, and results demonstrating that the framework can autonomously and progressively improve EDA tool at full million-line scale.", "authors": ["Cunxi Yu", "Haoxing Ren"], "categories": ["cs.AR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-16", "url": "https://arxiv.org/abs/2604.15082", "pdf_url": "https://arxiv.org/pdf/2604.15082v1", "arxiv_id": "2604.15082", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6486}
{"id": "991a0291afbf18da34c8e6e6707b0e920c1ee9f8cd401770b4d7dc1dfae67e4d", "sources": ["arxiv", "semantic_scholar"], "title": "Coalition Formation in LLM Agent Networks: Stability Analysis and Convergence Guarantees", "abstract": "Large Language Model (LLM) agents are increasingly deployed in multi-agent systems requiring strategic coordination. While recent work has analyzed LLM behavior in two-player games, coalition formation, where $n$ agents dynamically form cooperative groups, remains theoretically uncharacterized. We present the first framework grounding coalition formation in LLM agent networks in hedonic game theory with formal stability guarantees. We introduce the LLM Coalition Formation Game (LCFG), establish sufficient conditions for Nash-stable partitions, and prove complexity results. Our analysis reveals that LLM agents exhibit bounded rationality characterized by $ε$-rational preferences; we provide both deterministic existence guarantees and consistency-driven stability bounds whose predictions are consistent with empirical outcomes. Experiments with GPT-4, Claude-3, and Llama-3 across 2,400 episodes validate our framework: LLM coalitions achieve Nash stability in 73.2% of cases under our Coalition-of-Thought (CoalT) protocol, compared to 58.4% under chain-of-thought and 41.8% under standard prompting ($p < 0.001$). Our framework provides theoretical foundations for designing stable multi-agent LLM systems.", "authors": ["Dongxin Guo", "Jikun Wu", "Siu-Ming Yiu"], "categories": ["cs.GT", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-15", "url": "https://arxiv.org/abs/2604.14386", "pdf_url": "https://arxiv.org/pdf/2604.14386v1", "arxiv_id": "2604.14386", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3485}
{"id": "f2001a02aef0e726997ddeae48de945c0dba2c4d925ada5fa7b7bffbd7147084", "sources": ["arxiv", "semantic_scholar"], "title": "ToolOmni: Enabling Open-World Tool Use via Agentic learning with Proactive Retrieval and Grounded Execution", "abstract": "Large Language Models (LLMs) enhance their problem-solving capability by utilizing external tools. However, in open-world scenarios with massive and evolving tool repositories, existing methods relying on static embedding retrieval or parameter memorization of tools struggle to align user intent with tool semantics or generalize to unseen tools, respectively, leading to suboptimal accuracy of open-world tool retrieval and execution. To address these, we present ToolOmni, a unified agentic framework that enables LLMs for open-world tool use by proactive retrieval and grounded execution within a reasoning loop. First, we construct a cold-start multi-turn interaction dataset to instill foundational agentic capabilities via Supervised Fine-Tuning (SFT). Then, we introduce open-world tool learning based on a Decoupled Multi-Objective GRPO algorithm, which simultaneously optimizes LLMs for both tool retrieval accuracy and execution efficacy in online environments. Extensive experiments demonstrate that ToolOmni achieves state-of-the-art performance both in retrieval and execution, surpassing strong baselines by a significant margin of +10.8% in end-to-end execution success rate, while exhibiting exceptional robustness and generalization capabilities.", "authors": ["Shouzheng Huang", "Meishan Zhang", "Baotian Hu", "Min Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-15", "url": "https://arxiv.org/abs/2604.13787", "pdf_url": "https://arxiv.org/pdf/2604.13787v1", "arxiv_id": "2604.13787", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3485}
{"id": "c67804e7fb89057e919fc379c5096d6f0eb63ff70062281f92a634bde90dceaf", "sources": ["arxiv", "semantic_scholar"], "title": "TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Exploration", "abstract": "While Large Language Models (LLMs) have empowered AI research agents to perform isolated scientific tasks, automating complex, real-world workflows, such as LLM training, remains a significant challenge. In this paper, we introduce TREX, a multi-agent system that automates the entire LLM training life-cycle. By orchestrating collaboration between two core modules-the Researcher and the Executor-the system seamlessly performs requirement analysis, open-domain literature and data research, formulation of training strategies, preparation of data recipes, and model training and evaluation. The multi-round experimental process is modeled as a search tree, enabling the system to efficiently plan exploration paths, reuse historical results, and distill high-level insights from iterative trials. To evaluate the capability of automated LLM training, we construct FT-Bench, a benchmark comprising 10 tasks derived from real-world scenarios, ranging from optimizing fundamental model capabilities to enhancing performance on domain-specific tasks. Experimental results demonstrate that the TREX agent consistently optimizes model performance on target tasks.", "authors": ["Zerun Ma", "Guoqiang Wang", "Xinchen Xie", "Yicheng Chen", "He Du", "Bowen Li", "Yanan Sun", "Wenran Liu", "Kai Chen", "Yining Li"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-15", "url": "https://arxiv.org/abs/2604.14116", "pdf_url": "https://arxiv.org/pdf/2604.14116v2", "arxiv_id": "2604.14116", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3485}
{"id": "386f6abea7cb1c04925ed92e65727a24a75085bfac48491a9046012af2b24215", "sources": ["arxiv", "semantic_scholar"], "title": "CascadeDebate: Multi-Agent Deliberation for Cost-Aware LLM Cascades", "abstract": "Cascaded LLM systems coordinate models of varying sizes with human experts to balance accuracy, cost, and abstention under uncertainty. However, single-model tiers at each stage often struggle with ambiguous queries, triggering premature escalations to costlier models or experts due to under-confidence and inefficient compute scaling. CascadeDebate addresses this gap by inserting multi-agent deliberation directly at each tier's escalation boundary. Confidence-based routers activate lightweight agent ensembles only for uncertain cases, enabling consensus-driven resolution of ambiguities internally without invoking higher-cost upgrades. Our unified architecture alternates single-model inference with selective multi-agent deliberation across model scales, culminating in human experts as the final fallback. This design scales test-time compute dynamically according to query difficulty. Across five benchmarks spanning science, medicine, and general knowledge, CascadeDebate outperforms strong single-model cascades and standalone multi-agent systems by up to 26.75 percent. An online threshold optimizer proves essential, boosting accuracy by 20.98 to 52.33 percent relative improvement over fixed policies and enabling elastic adaptation to real-world distributions.", "authors": ["Raeyoung Chang", "Dongwook Kwon", "Jisoo Lee", "Nikhil Verma"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-14", "url": "https://arxiv.org/abs/2604.12262", "pdf_url": "https://arxiv.org/pdf/2604.12262v1", "arxiv_id": "2604.12262", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3478}
{"id": "79c9c53ddf18a7048e5ef6f2de978b9b5deede93b7f78aee0c6602a94ee4cfd3", "sources": ["arxiv", "semantic_scholar"], "title": "Cross-Domain Query Translation for Network Troubleshooting: A Multi-Agent LLM Framework with Privacy Preservation and Self-Reflection", "abstract": "This paper presents a hierarchical multi-agent LLM architecture to bridge communication gaps between non-technical end users and telecommunications domain experts in private network environments. We propose a cross-domain query translation framework that leverages specialized language models coordinated through multi-agent reflection-based reasoning. The resulting system addresses three critical challenges: (1) accurately classify user queries related to telecommunications network issues using a dual-stage hierarchical approach, (2) preserve user privacy through the anonymization of semantically relevant personally identifiable information (PII) while maintaining diagnostic utility, and (3) translate technical expert responses into user-comprehensible language. Our approach employs ReAct-style agents enhanced with self-reflection mechanisms for iterative output refinement, semantic-preserving anonymization techniques respecting $k$-anonymity and differential privacy principles, and few-shot learning strategies designed for limited training data scenarios. The framework was comprehensively evaluated on 10,000 previously unseen validation scenarios across various vertical industries.", "authors": ["Nguyen Phuc Tran", "Brigitte Jaumard", "Karthikeyan Premkumar", "Salman Memon"], "categories": ["cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-14", "url": "https://arxiv.org/abs/2604.13353", "pdf_url": "https://arxiv.org/pdf/2604.13353v2", "arxiv_id": "2604.13353", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "EuCNC & 6G Summit 2026", "quality_score": 0.5466}
{"id": "4a798f68b51aa5aa35db950098ca10224ce4e7d3d93d66691c6f24a70d4950db", "sources": ["arxiv", "semantic_scholar"], "title": "TRUST Agents: A Collaborative Multi-Agent Framework for Fake News Detection, Explainable Verification, and Logic-Aware Claim Reasoning", "abstract": "TRUST Agents is a collaborative multi-agent framework for explainable fact verification and fake news detection. Rather than treating verification as a simple true-or-false classification task, the system identifies verifiable claims, retrieves relevant evidence, compares claims against that evidence, reasons under uncertainty, and generates explanations that humans can inspect. The baseline pipeline consists of four specialized agents. A claim extractor uses named entity recognition, dependency parsing, and LLM-based extraction to identify factual claims. A retrieval agent performs hybrid sparse and dense search using BM25 and FAISS. A verifier agent compares claims with retrieved evidence and produces verdicts with calibrated confidence. An explainer agent then generates a human-readable report with explicit evidence citations. To handle complex claims more effectively, we introduce a research-oriented extension with three additional components: a decomposer agent inspired by LoCal-style claim decomposition, a Delphi-inspired multi-agent jury with specialized verifier personas, and a logic aggregator that combines atomic verdicts using conjunction, disjunction, negation, and implication. We evaluate both pipelines on the LIAR benchmark against fine-tuned BERT, fine-tuned RoBERTa, and a zero-shot LLM baseline. Although supervised encoders remain stronger on raw metrics, TRUST Agents improves interpretability, evidence transparency, and reasoning over compound claims. Results also show that retrieval quality and uncertainty calibration remain the main bottlenecks in trustworthy automated fact verification.", "authors": ["Gautama Shastry Bulusu Venkata", "Santhosh Kakarla", "Maheedhar Omtri Mohan", "Aishwarya Gaddam"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-14", "url": "https://arxiv.org/abs/2604.12184", "pdf_url": "https://arxiv.org/pdf/2604.12184v1", "arxiv_id": "2604.12184", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3478}
{"id": "da110261f9ac0d86b10641042ebc3e8daf78ea9d4899a4d7713c30926271be2b", "sources": ["arxiv", "semantic_scholar"], "title": "Aethon: A Reference-Based Replication Primitive for Constant-Time Instantiation of Stateful AI Agents", "abstract": "The transition from stateless model inference to stateful agentic execution is reshaping the systems assumptions underlying modern AI infrastructure. While large language models have made persistent, tool-using, and collaborative agents technically viable, existing runtime architectures remain constrained by materialization-heavy instantiation models that impose significant latency and memory overhead. This paper introduces Aethon, a reference-based replication primitive for near-constant-time instantiation of stateful AI agents. Rather than reconstructing agents as fully materialized objects, Aethon represents each instance as a compositional view over stable definitions, layered memory, and local contextual overlays. By shifting instantiation from duplication to reference, Aethon decouples creation cost from inherited structure. We present the conceptual framework, system architecture, and memory model underlying Aethon, including layered inheritance and copy-on-write semantics. We analyze its implications for complexity, scalability, multi-agent orchestration, and enterprise governance. We argue that reference-based instantiation is not merely an optimization, but a more appropriate systems abstraction for production-scale agentic software. Aethon points toward a new class of AI infrastructure in which agents become lightweight, composable execution identities that can be spawned, specialized, and governed at scale.", "authors": ["Swanand Rao", "Kiran Kashalkar", "Parvathi Somashekar", "Priya Krishnan"], "categories": ["cs.AI", "cs.AR", "cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.12129", "pdf_url": "https://arxiv.org/pdf/2604.12129v1", "arxiv_id": "2604.12129", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3471}
{"id": "0f0d18bfd7cb84d7bea8c1ff5ce677b038a58620645a503e21b0c68a2e0c1401", "sources": ["arxiv", "semantic_scholar"], "title": "Knowledge Compounding: An Empirical Economic Analysis of Self-Evolving Knowledge Wikis under the Agentic ROI Framework", "abstract": "Building on the Agentic ROI framework proposed by Liu et al. (2026), this paper introduces knowledge compounding as a new measurable concept in the empirical economics of LLM agents and validates it through a controlled four-query experiment on Qing Claw, an industrial-grade C# reimplementation of the OpenClaw multi-agent framework. Our central theoretical claim is that the cost term in the original Agentic ROI equation contains an unexamined assumption -- that the cost of each task is mutually independent. This assumption holds under the traditional retrieval-augmented generation (RAG) paradigm but breaks down once a persistent, structured knowledge layer is introduced. We propose a dynamic Agentic ROI model in which cost is treated as a time-varying function Cost(t) governed by a knowledge-base coverage rate H(t). Empirical results from four sequential queries on the same domain yield a cumulative token consumption of 47K under the compounding regime versus 305K under a matched RAG baseline -- a savings of 84.6%. Calibrated 30-day projections indicate cumulative savings of 53.7% under medium topic concentration and 81.3% under high concentration, with the gap widening monotonically over time. We further identify three microeconomic mechanisms underlying the compounding effect: (i) one-time INGEST amortized over N retrievals, (ii) auto-feedback of high-value answers into synthesis pages, and (iii) write-back of external search results into entity pages. The theoretical contribution of this paper is a recategorization of LLM tokens from consumables to capital goods, shifting the economic discussion from static marginal cost analysis to dynamic capital accumulation. The engineering contribution is a minimal reproducible implementation in approximately 200 lines of C#, which we believe is the first complete industrial-grade reference implementation of Karpathy's (2026) LLM Wiki paradigm.", "authors": ["Shuide Wen", "Beier Ku"], "categories": ["econ.EM"], "fields_of_study": ["Economics"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11243", "pdf_url": "https://arxiv.org/pdf/2604.11243v2", "arxiv_id": "2604.11243", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6446}
{"id": "c685d863b845bdd4ebc7a5074112b25a571887a2c29dfc648200c7a5268bb3fe", "sources": ["arxiv", "semantic_scholar"], "title": "UniToolCall: Unifying Tool-Use Representation, Data, and Evaluation for LLM Agents", "abstract": "Tool-use capability is a fundamental component of LLM agents, enabling them to interact with external systems through structured function calls. However, existing research exhibits inconsistent interaction representations, largely overlooks the structural distribution of tool-use trajectories, and relies on incompatible evaluation benchmarks. We present UniToolCall, a unified framework for tool learning that standardizes the entire pipeline from toolset construction and dataset generation to evaluation. The framework curates a large tool pool of 22k+ tools and constructs a hybrid training corpus of 390k+ instances by combining 10 standardized public datasets with structurally controlled synthetic trajectories. It explicitly models diverse interaction patterns, including single-hop vs. multi-hop and single-turn vs. multi-turn, while capturing both serial and parallel execution structures. To support coherent multi-turn reasoning, we further introduce an Anchor Linkage mechanism that enforces cross-turn dependencies. Furthermore, we convert 7 public benchmarks into a unified Query--Action--Observation--Answer (QAOA) representation with fine-grained evaluation at the function-call, turn, and conversation levels. Experiments show that fine-tuning Qwen3-8B on our dataset substantially improves tool-use performance. Under the distractor-heavy Hybrid-20 setting, achieves 93.0% single-turn Strict Precision, outperforming commercial models including GPT, Gemini, and Claude.", "authors": ["Yijuan Liang", "Xinghao Chen", "Yifan Ge", "Ziyi Wu", "Hao Wu", "Changyu Zeng", "Wei Xing", "Xiaoyu Shen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11557", "pdf_url": "https://arxiv.org/pdf/2604.11557v2", "arxiv_id": "2604.11557", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/EIT-NLP/UniToolCall", "venue": null, "quality_score": 0.6446}
{"id": "fb7c3eed0a00d9795f7688393e9061b14ab166e650ed2cb879bfaa84d0eaf888", "sources": ["arxiv", "semantic_scholar"], "title": "A Simulation-Based Method for Testing Collaborative Learning Scaffolds Using LLM-Based Multi-Agent Systems", "abstract": "Background: Traditional research on collaborative learning scaffolding is often time-consuming and resource-heavy, which hinders the rapid iteration and optimization of instructional strategies. LLM-based multi-agent systems have recently emerged as a powerful tool to simulate complex social interactions and provide a novel paradigm for educational research. Objectives: This study proposes an LLM-based multi-agent simulation approach to investigate collaborative learning processes and the effectiveness of instructional scaffolds prior to actual classroom deployment. The research specifically examines the feasibility of simulating group discussions and the alignment of these simulations with established learning science theories. Methods: The simulation system was implemented using the MetaGPT framework and GPT-4o, comprising one teacher agent and five distinct student roles (Leader, Supporter, Expounder, Rebutter, and Summarizer). Two scaffolding strategies, \"Deep Think before Speak\" and \"Direct Speak\", were compared across ten classical Chinese poetry appreciation tasks. Evaluation was conducted through discourse analysis of quality and behavior. Results and Conclusions: The introduction of the \"Deep Think before Speak\" scaffold significantly improved the agents' discourse diversity and interaction depth while notably reducing content repetitiveness. Behavioral analysis showed that the scaffold encouraged more complex interaction patterns, such as reflecting, rebutting, and explaining. These findings align with the ICAP framework, as the scaffold prompted agents to move from simple \"Active\" participation to \"Constructive\" and \"Interactive\" knowledge co-construction. This study demonstrates the feasibility and ecological validity of using LLM-based multi-agent systems to simulate authentic collaborative learning dynamics.", "authors": ["Han Wua", "Lishan Zhang", "Chunming Lu"], "categories": ["cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11161", "pdf_url": "https://arxiv.org/pdf/2604.11161v1", "arxiv_id": "2604.11161", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3471}
{"id": "923c118fb2dcc0f878cf804a3035855fa7a190a72d234ad9f17a3c41af0f7b11", "sources": ["arxiv", "semantic_scholar"], "title": "ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection", "abstract": "Tool-augmented Large Language Model (LLM) agents have demonstrated impressive capabilities in automating complex, multi-step real-world tasks, yet remain vulnerable to indirect prompt injection. Adversaries exploit this weakness by embedding malicious instructions within tool-returned content, which agents directly incorporate into their conversation history as trusted observations. To address these vulnerabilities, we introduce \\textsc{ClawGuard}, a novel runtime security framework that enforces a user-confirmed rule set at every tool-call boundary, transforming unreliable alignment-dependent defense into a deterministic, auditable mechanism that intercepts adversarial tool calls before any real-world effect is produced. By automatically deriving task-specific access constraints from the user's stated objective prior to any external tool invocation, \\textsc{ClawGuard} blocks all three injection pathways without model modification or infrastructure change. Experiments across five state-of-the-art language models on six injection benchmarks covering web, local, MCP, and skill channels, as well as three utility benchmarks covering OS, web, and code tasks, demonstrate that \\textsc{ClawGuard} achieves robust protection against indirect prompt injection without compromising agent utility or introducing significant token overhead. This work establishes deterministic tool-call boundary enforcement as an effective defense mechanism for secure agentic AI systems. Code is publicly available at github.com/Claw-Guard/ClawGuard/.", "authors": ["Wei Zhao", "Zhe Li", "Peixin Zhang", "Jun Sun"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11790", "pdf_url": "https://arxiv.org/pdf/2604.11790v2", "arxiv_id": "2604.11790", "doi": null, "citation_count": 7, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6446}
{"id": "8f5c4836b2ff7a5469f4a8be188c70238f9a205116d86fccb2d6f4bc159264a9", "sources": ["arxiv", "semantic_scholar"], "title": "M2HRI: An LLM-Driven Multimodal Multi-Agent Framework for Personalized Human-Robot Interaction", "abstract": "Multi-robot systems hold significant promise for social environments such as homes and hospitals, yet existing multi-robot works treat robots as functionally identical, overlooking how robots individual identity shape user perception and how coordination shapes multi-robot behavior when such individuality is present. To address this, we introduce M2HRI, a multimodal multi-agent framework built on large language models that equips each robot with distinct personality and long-term memory, alongside a coordination mechanism conditioned on these differences. In a controlled user study (n = 105) in a multi-agent human-robot interaction (HRI) scenario, we find that LLM-driven personality traits are significantly distinguishable and enhance interaction quality, long-term memory improves personalization and preference awareness, and centralized coordination significantly reduces overlap while improving overall interaction quality. Together, these results demonstrate that both agent individuality and structured coordination are essential for coherent and socially appropriate multi-agent HRI. Project website and code are available at https://project-m2hri.github.io/.", "authors": ["Shaid Hasan", "Breenice Lee", "Sujan Sarker", "Tariq Iqbal"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11975", "pdf_url": "https://arxiv.org/pdf/2604.11975v1", "arxiv_id": "2604.11975", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3471}
{"id": "7df7efc81910e78b5cc40645a29282f098737c690c0a8925a57b7f23c75a301e", "sources": ["arxiv", "semantic_scholar"], "title": "AgentWebBench: Benchmarking Multi-Agent Coordination in Agentic Web", "abstract": "Agentic Web is an emerging paradigm where autonomous agents help users use online information. As the paradigm develops, content providers are also deploying agents to manage their data and serve it through controlled interfaces. This shift moves information access from centralized retrieval to decentralized coordination. To study this setting, we introduce AgentWebBench, a benchmark that evaluates how well a user agent synthesizes answers by interacting with website-specific content agents. We evaluate four tasks that cover common web information needs, spanning ranked retrieval (web search, web recommendation) and open-ended synthesis (question answering, deep research). Across seven advanced LLMs and three coordination strategies, multi-agent coordination generally lags behind centralized retrieval as expected, because user agent cannot directly access the corpus, but the gap shrinks with model scale and can even outperform centralized retrieval on question answering. This benchmark also enables us to study properties of the emerging paradigm of the digital world. We find that decentralized access concentrates traffic toward a small set of websites, test time scaling improves both interaction reliability and task performance, and strong results require sufficient interactions guided by careful planning. Finally, our failure analysis suggests that user agents need better planning and answer synthesis, while content agents need more reliable retrieval and evidence quality. Code, data, and APIs are released on https://github.com/cxcscmu/AgentWebBench.", "authors": ["Shanshan Zhong", "Kate Shen", "Chenyan Xiong"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.10938", "pdf_url": "https://arxiv.org/pdf/2604.10938v1", "arxiv_id": "2604.10938", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/cxcscmu/AgentWebBench", "venue": null, "quality_score": 0.6446}
{"id": "63490657376ba8e426a0b754b746eef695b160d0c4b5e92c5a617f59e755f164", "sources": ["arxiv", "semantic_scholar"], "title": "AgentForge: Execution-Grounded Multi-Agent LLM Framework for Autonomous Software Engineering", "abstract": "Large language models generate plausible code but cannot verify correctness. Existing multi-agent systems simulate execution or leave verification optional. We introduce execution-grounded verification as a first-class principle: every code change must survive sandboxed execution before propagation. We instantiate this principle in AGENTFORGE, a multi-agent framework where Planner, Coder, Tester, Debugger, and Critic agents coordinate through shared memory and a mandatory Docker sandbox. We formalize software engineering with LLMs as an iterative decision process over repository states, where execution feedback provides a stronger supervision signal than next-token likelihood. AGENTFORGE achieves 40.0\\% resolution on SWE-BENCH Lite, outperforming single-agent baselines by 26--28 points. Ablations confirm that execution feedback and role decomposition each independently drive performance. The framework is open-source at https://github.com/raja21068/AutoCodeAI.", "authors": ["Rajesh Kumar", "Waqar Ali", "Junaid Ahmed", "Najma Imtiaz Ali", "Shaban Usman"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.13120", "pdf_url": "https://arxiv.org/pdf/2604.13120v1", "arxiv_id": "2604.13120", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/raja21068/AutoCodeAI", "venue": null, "quality_score": 0.6446}
{"id": "be2dbf551745a733e90c6b020a6419360bf9166faac39d5f2e04a820a8d6b83c", "sources": ["arxiv", "semantic_scholar"], "title": "PAC-BENCH: Evaluating Multi-Agent Collaboration under Privacy Constraints", "abstract": "We are entering an era in which individuals and organizations increasingly deploy dedicated AI agents that interact and collaborate with other agents. However, the dynamics of multi-agent collaboration under privacy constraints remain poorly understood. In this work, we present $PAC\\text{-}Bench$, a benchmark for systematic evaluation of multi-agent collaboration under privacy constraints. Experiments on $PAC\\text{-}Bench$ show that privacy constraints substantially degrade collaboration performance and make outcomes depend more on the initiating agent than the partner. Further analysis reveals that this degradation is driven by recurring coordination breakdowns, including early-stage privacy violations, overly conservative abstraction, and privacy-induced hallucinations. Together, our findings identify privacy-aware multi-agent collaboration as a distinct and unresolved challenge that requires new coordination mechanisms beyond existing agent capabilities.", "authors": ["Minjun Park", "Donghyun Kim", "Hyeonjong Ju", "Seungwon Lim", "Dongwook Choi", "Taeyoon Kwon", "Minju Kim", "Jinyoung Yeo"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11523", "pdf_url": "https://arxiv.org/pdf/2604.11523v1", "arxiv_id": "2604.11523", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3471}
{"id": "ad2778e0b0028c48dc7e9ad46e819fcdafcf2492855ca97452075e843bc555d8", "sources": ["arxiv", "semantic_scholar"], "title": "Federated Single-Agent Robotics: Multi-Robot Coordination Without Intra-Robot Multi-Agent Fragmentation", "abstract": "As embodied robots move toward fleet-scale operation, multi-robot coordination is becoming a central systems challenge. Existing approaches often treat this as motivation for increasing internal multi-agent decomposition within each robot. We argue for a different principle: multi-robot coordination does not require intra-robot multi-agent fragmentation. Each robot should remain a single embodied agent with its own persistent runtime, local policy scope, capability state, and recovery authority, while coordination emerges through federation across robots at the fleet level. We present Federated Single-Agent Robotics (FSAR), a runtime architecture for multi-robot coordination built on single-agent robot runtimes. Each robot exposes a governed capability surface rather than an internally fragmented agent society. Fleet coordination is achieved through shared capability registries, cross-robot task delegation, policy-aware authority assignment, trust-scoped interaction, and layered recovery protocols. We formalize key coordination relations including authority delegation, inter-robot capability requests, local-versus-fleet recovery boundaries, and hierarchical human supervision, and describe a fleet runtime architecture supporting shared Embodied Capability Module (ECM) discovery, contract-aware cross-robot coordination, and fleet-level governance. We evaluate FSAR on representative multi-robot coordination scenarios against decomposition-heavy baselines. Results show statistically significant gains in governance locality (d=2.91, p<.001 vs. centralized control) and recovery containment (d=4.88, p<.001 vs. decomposition-heavy), while reducing authority conflicts and policy violations across all scenarios. Our results support the view that the path from embodied agents to embodied fleets is better served by federation across coherent robot runtimes than by fragmentation within them.", "authors": ["Xue Qin", "Simin Luan", "John See", "Cong Yang", "Zhijun Li"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-13", "url": "https://arxiv.org/abs/2604.11028", "pdf_url": "https://arxiv.org/pdf/2604.11028v2", "arxiv_id": "2604.11028", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/s20sc/fsar-fleet-coordination", "venue": null, "quality_score": 0.6446}
{"id": "1722370b003d67202292d14798acb88fe1fa3c368f6f3833da633d2d7b75fa6d", "sources": ["arxiv", "semantic_scholar"], "title": "Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents", "abstract": "Reinforcement learning (RL) has been widely used to train LLM agents for multi-turn interactive tasks, but its sample efficiency is severely limited by sparse rewards and long horizons. On-policy self-distillation (OPSD) alleviates this by providing dense token-level supervision from a privileged teacher that has access to ground-truth answers. However, such fixed privileged information cannot capture the diverse valid strategies in agent tasks, and naively combining OPSD with RL often leads to training collapse. To address these limitations, we introduce Skill-SD, a framework that turns the agent's own trajectories into dynamic training-only supervision. Completed trajectories are summarized into compact natural language skills that describe successful behaviors, mistakes, and workflows. These skills serve as dynamic privileged information conditioning only the teacher, while the student always acts under the plain task prompt and learns to internalize the guidance through distillation. To stabilize the training, we derive an importance-weighted reverse-KL loss to provide gradient-correct token-level distillation, and dynamically synchronize the teacher with the improving student. Experimental results on agentic benchmarks demonstrate that Skill-SD substantially outperforms the standard RL baseline, improving both vanilla GRPO (+14.0%/+10.9% on AppWorld/Sokoban) and vanilla OPD (+42.1%/+40.6%). Project page: https://k1xe.github.io/skill-sd/", "authors": ["Hao Wang", "Guozhi Wang", "Han Xiao", "Yufeng Zhou", "Yue Pan", "Jichao Wang", "Ke Xu", "Yafei Wen", "Xiaohu Ruan", "Xiaoxin Chen", "Honggang Qi"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-12", "url": "https://arxiv.org/abs/2604.10674", "pdf_url": "https://arxiv.org/pdf/2604.10674v1", "arxiv_id": "2604.10674", "doi": null, "citation_count": 16, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3464}
{"id": "f29208b45bd4387e74032c7f4ba21c4e0d25aec393db6680a631bbf07df1993c", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Video Generation: From Text to Executable Event Graphs via Tool-Constrained LLM Planning", "abstract": "Existing multi-agent video generation systems use LLM agents to orchestrate neural video generators, producing visually impressive but semantically unreliable outputs with no ground truth annotations. We present an agentic system that inverts this paradigm: instead of generating pixels, the LLM constructs a formal Graph of Events in Space and Time (GEST) -- a structured specification of actors, actions, objects, and temporal constraints -- which is then executed deterministically in a 3D game engine. A staged LLM refinement pipeline fails entirely at this task (0 of 50 attempts produce an executable specification), motivating a fundamentally different architecture based on a separation of concerns: the LLM handles narrative planning through natural language reasoning, while a programmatic state backend enforces all simulator constraints through validated tool calls, guaranteeing that every generated specification is executable by construction. The system uses a hierarchical two-agent architecture -- a Director that plans the story and a Scene Builder that constructs individual scenes through a round-based state machine -- with dedicated Relation Subagents that populate the logical and semantic edge types of the GEST formalism that procedural generation leaves empty, making this the first approach to exercise the full expressive capacity of the representation. We evaluate in two stages: autonomous generation against procedural baselines via a 3-model LLM jury, where agentic narratives win 79% of text and 74% of video comparisons; and seeded generation where the same text is given to our system, VEO 3.1, and WAN 2.2, with human annotations showing engine-generated videos substantially outperform neural generators on physical validity (58% vs 25% and 20%) and semantic alignment (3.75/5 vs 2.33 and 1.50).", "authors": ["Nicolae Cudlenco", "Mihai Masala", "Marius Leordeanu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-11", "url": "https://arxiv.org/abs/2604.10383", "pdf_url": "https://arxiv.org/pdf/2604.10383v1", "arxiv_id": "2604.10383", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3456}
{"id": "891538ec051033e89731bad6a49664d19e727ab583e0836dc9ef13519314b287", "sources": ["arxiv", "semantic_scholar"], "title": "The Amazing Agent Race: Strong Tool Users, Weak Navigators", "abstract": "Existing tool-use benchmarks for LLM agents are overwhelmingly linear: our analysis of six benchmarks shows 55 to 100% of instances are simple chains of 2 to 5 steps. We introduce The Amazing Agent Race (AAR), a benchmark featuring directed acyclic graph (DAG) puzzles (or \"legs\") with fork-merge tool chains. We release 1,400 instances across two variants: sequential (800 legs) and compositional (600 DAG legs). Agents must navigate Wikipedia, execute multi-step tool chains, and aggregate results into a verifiable answer. Legs are procedurally generated from Wikipedia seeds across four difficulty levels with live-API validation. Three complementary metrics (finish-line accuracy, pit-stop visit rate, and roadblock completion rate) separately diagnose navigation, tool-use, and arithmetic failures. Evaluating three agent frameworks on 1,400 legs, the best achieves only 37.2% accuracy. Navigation errors dominate (27 to 52% of trials) while tool-use errors remain below 17%, and agent architecture matters as much as model scale (Claude Code matches Codex CLI at 37% with 6x fewer tokens). The compositional structure of AAR reveals that agents fail not at calling tools but at navigating to the right pages, a blind spot invisible to linear benchmarks. The project page can be accessed at: https://minnesotanlp.github.io/the-amazing-agent-race", "authors": ["Zae Myung Kim", "Dongseok Lee", "Jaehyung Kim", "Vipul Raheja", "Dongyeop Kang"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-11", "url": "https://arxiv.org/abs/2604.10261", "pdf_url": "https://arxiv.org/pdf/2604.10261v2", "arxiv_id": "2604.10261", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3456}
{"id": "1d04822440b35eb5ad04a1421b1bbd971472ff9510df484f01e83675b04eb795", "sources": ["arxiv", "semantic_scholar"], "title": "ClawVM: Harness-Managed Virtual Memory for Stateful Tool-Using LLM Agents", "abstract": "Stateful tool-using LLM agents treat the context window as working memory, yet today's agent harnesses manage residency and durability as best-effort, causing recurring failures: lost state after compaction, bypassed flushes on reset, and destructive writeback. We present \\textsc{ClawVM}, a virtual memory layer that manages state as typed pages with minimum-fidelity invariants, multi-resolution representations under a token budget, and validated writeback at every lifecycle boundary. Because the harness already assembles prompts, mediates tools, and observes lifecycle events, it is the natural enforcement point; placing the contract there makes residency and durability deterministic and auditable. Across synthetic workloads, 12 real-session traces, and adversarial stress tests, \\textsc{ClawVM} eliminates all policy-controllable faults whenever the minimum-fidelity set fits within the token budget, confirmed by an offline oracle, and adds median <50 microseconds of policy-engine overhead per turn.", "authors": ["Mofasshara Rafique", "Laurent Bindschaedler"], "categories": ["cs.AI", "cs.OS", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-11", "url": "https://arxiv.org/abs/2604.10352", "pdf_url": "https://arxiv.org/pdf/2604.10352v1", "arxiv_id": "2604.10352", "doi": "10.1145/3805621.3807648", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3456}
{"id": "537cd52f58070f5fcad628c6f22dee1b9cb27a421370bb300d642830a80fbb0c", "sources": ["arxiv", "semantic_scholar"], "title": "From Helpful to Trustworthy: LLM Agents for Pair Programming", "abstract": "LLM-based coding agents are increasingly used to generate code, tests, and documentation. Still, their outputs can be plausible yet misaligned with developer intent and provide limited evidence for review in evolving projects. This limits our understanding of how to structure LLM pair-programming workflows so that artifacts remain reliable, auditable, and maintainable over time. To address this gap, this doctoral research proposes a systematic study of multi-agent LLM pair programming that externalizes intent and uses development tools for iterative validation. The plan includes three studies: translating informal problem statements into standards aligned requirements and formal specifications; refining tests and implementations using automated feedback, such as solver-backed counterexamples; and supporting maintenance tasks, including refactoring, API migrations, and documentation updates, while preserving validated behavior. The expected outcome is a clearer understanding of when multi-agent workflows increase trust, along with practical guidance for building reliable programming assistants for real-world development.", "authors": ["Ragib Shahariar Ayon"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-11", "url": "https://arxiv.org/abs/2604.10300", "pdf_url": "https://arxiv.org/pdf/2604.10300v1", "arxiv_id": "2604.10300", "doi": "10.1145/3803437.3804875", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3456}
{"id": "ce239e03ca2cdbc1bbe7a71e8974f69cddea072f68e45d583b7e78e2d9814035", "sources": ["arxiv", "semantic_scholar"], "title": "FinTrace: Holistic Trajectory-Level Evaluation of LLM Tool Calling for Long-Horizon Financial Tasks", "abstract": "Recent studies demonstrate that tool-calling capability enables large language models (LLMs) to interact with external environments for long-horizon financial tasks. While existing benchmarks have begun evaluating financial tool calling, they focus on limited scenarios and rely on call-level metrics that fail to capture trajectory-level reasoning quality. To address this gap, we introduce FinTrace, a benchmark comprising 800 expert-annotated trajectories spanning 34 real-world financial task categories across multiple difficulty levels. FinTrace employs a rubric-based evaluation protocol with nine metrics organized along four axes -- action correctness, execution efficiency, process quality, and output quality -- enabling fine-grained assessment of LLM tool-calling behavior. Our evaluation of 13 LLMs reveals that while frontier models achieve strong tool selection, all models struggle with information utilization and final answer quality, exposing a critical gap between invoking the right tools and reasoning effectively over their outputs. To move beyond diagnosis, we construct FinTrace-Training, the first trajectory-level preference dataset for financial tool-calling, containing 8,196 curated trajectories with tool-augmented contexts and preference pairs. We fine-tune Qwen-3.5-9B using supervised fine-tuning followed by direct preference optimization (DPO) and show that training on FinTrace-Training consistently improves intermediate reasoning metrics, with DPO more effectively suppressing failure modes. However, end-to-end answer quality remains a bottleneck, indicating that trajectory-level improvements do not yet fully propagate to final output quality.", "authors": ["Yupeng Cao", "Haohang Li", "Weijin Liu", "Wenbo Cao", "Anke Xu", "Lingfei Qian", "Xueqing Peng", "Minxue Tang", "Zhiyuan Yao", "Jimin Huang", "K. P. Subbalakshmi", "Zining Zhu", "Jordan W. Suchow", "Yangyang Yu"], "categories": ["cs.AI", "cs.CE", "cs.CL", "cs.MM"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-11", "url": "https://arxiv.org/abs/2604.10015", "pdf_url": "https://arxiv.org/pdf/2604.10015v2", "arxiv_id": "2604.10015", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3456}
{"id": "e395351bcc3f803db6c20e5da6821f6d24ded878bcbd1d100062e35b807193b7", "sources": ["arxiv", "semantic_scholar"], "title": "GenesisFunc: Multi-Agent Data Generation for Accurate and Generalizable Function-Calling", "abstract": "Large Language Models (LLMs) extend their capabilities through function-calling (FC), which relies on training data with high quality, diversity, and broad coverage of scenario. However, obtaining and annotating real function-calling data is challenging, while synthetic data from existing pipelines often suffers from unreliable APIs, limited tool scalability, insufficient diversity, and weak quality control. To address these, we present GenesisFunc, an automated pipeline for generating FC training data. Starting from reliable tools in widely used public benchmarks, our GenesisFunc employs a multi-agent framework to support a dialogue generation system that produces conversations spanning diverse scenarios, while maintaining both diversity and quality throughout the process. The accuracy of the data is further reinforced through a multi-stage evaluation system. We fine-tune an 8B LLM on the synthetic dataset and show through extensive experiments that it outperforms similarly sized open-source models in in-domain FC performance and out-of-domain generalization, while reaching FC capabilities comparable to some of the latest API-based models. In addition, our method demonstrates strong potential to scale effectively across downstream tools, underscoring its real-world applicability.", "authors": ["Hao-Xiang Xu", "Chong Deng", "Jiaqing Liu", "Wen Wang", "Qian Chen", "Lujia Bao", "Xiangang Li", "Zhen-Hua Ling"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-10", "url": "https://arxiv.org/abs/2605.28835", "pdf_url": "https://arxiv.org/pdf/2605.28835v1", "arxiv_id": "2605.28835", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6405}
{"id": "9068b057863b9fa5f48af77ac6d08140e01961e73f6e7b062910577b6720d55f", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing LLM Problem Solving via Tutor-Student Multi-Agent Interaction", "abstract": "Human cognitive development is shaped not only by individual effort but by structured social interaction, where role-based exchanges such as those between a tutor and a learner, enable solutions that neither could achieve alone. Inspired by these developmental principles, we ask the question whether a tutor-student multi-agent system can create a synergistic effect by pushing Large Language Model (LLM) beyond what it can do within existing frameworks. To test the idea, we adopt autonomous coding problem domain where two agents instantiated from the same LLM assigned asymmetric roles: a student agent generates and iteratively refines solutions, while a tutor agent provides structured evaluative feedback without access to ground-truth answers. In our proposed framework (PETITE), we aim to extract better problem-solving performance from one model by structuring its interaction through complementary roles, rather than relying on stronger supervisory models or heterogeneous ensembles. Our model is evaluated on the APPS coding benchmark against state-of-the-art approaches of Self-Consistency, Self-Refine, Multi-Agent Debate, and Multi-Agent Review. The results show that our model achieves similar or higher accuracy while consuming significantly fewer tokens. These results suggest that developmentally grounded role-differentiated interaction structures provide a principled and resource-efficient paradigm for enhancing LLM problem-solving through structured peer-like interactions. Index Terms- Peer Tutoring, Scaffolding, Large Language Models, Multi-Agent Systems, Code Generation", "authors": ["Nurullah Eymen Özdemir", "Erhan Oztop"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-10", "url": "https://arxiv.org/abs/2604.08931", "pdf_url": "https://arxiv.org/pdf/2604.08931v1", "arxiv_id": "2604.08931", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3449}
{"id": "87acd0718a1ad95092a4261490895c0c04a7f33f21e3fb715eb69e605c884a7d", "sources": ["arxiv", "semantic_scholar"], "title": "CONSCIENTIA: Can LLM Agents Learn to Strategize? Emergent Deception and Trust in a Multi-Agent NYC Simulation", "abstract": "As large language models (LLMs) are increasingly deployed as autonomous agents, understanding how strategic behavior emerges in multi-agent environments has become an important alignment challenge. We take a neutral empirical stance and construct a controlled environment in which strategic behavior can be directly observed and measured. We introduce a large-scale multi-agent simulation in a simplified model of New York City, where LLM-driven agents interact under opposing incentives. Blue agents aim to reach their destinations efficiently, while Red agents attempt to divert them toward billboard-heavy routes using persuasive language to maximize advertising revenue. Hidden identities make navigation socially mediated, forcing agents to decide when to trust or deceive. We study policy learning through an iterative simulation pipeline that updates agent policies across repeated interaction rounds using Kahneman-Tversky Optimization (KTO). Blue agents are optimized to reduce billboard exposure while preserving navigation efficiency, whereas Red agents adapt to exploit remaining weaknesses. Across iterations, the best Blue policy improves task success from 46.0% to 57.3%, although susceptibility remains high at 70.7%. Later policies exhibit stronger selective cooperation while preserving trajectory efficiency. However, a persistent safety-helpfulness trade-off remains: policies that better resist adversarial steering do not simultaneously maximize task completion. Overall, our results show that LLM agents can exhibit limited strategic behavior, including selective trust and deception, while remaining highly vulnerable to adversarial persuasion.", "authors": ["Aarush Sinha", "Arion Das", "Soumyadeep Nag", "Charan Karnati", "Shravani Nag", "Chandra Vadhan Raj", "Aman Chadha", "Vinija Jain", "Suranjana Trivedy", "Amitava Das"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-10", "url": "https://arxiv.org/abs/2604.09746", "pdf_url": "https://arxiv.org/pdf/2604.09746v1", "arxiv_id": "2604.09746", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3449}
{"id": "e426facf10109e38810cdb15662a338ea4aa5f1f38b010c0fa8eb0fc7082be38", "sources": ["arxiv", "semantic_scholar"], "title": "MPAC: A Multi-Principal Agent Coordination Protocol for Interoperable Multi-Agent Collaboration", "abstract": "The AI agent ecosystem has converged on two protocols: the Model Context Protocol (MCP) for tool invocation and Agent-to-Agent (A2A) for single-principal task delegation. Both assume a single controlling principal, meaning one person or organization that owns every agent. When independent principals' agents must coordinate over shared state, such as engineers' coding agents editing the same repository, family members planning a shared trip, or agents from different organizations negotiating a joint decision, neither protocol applies, and coordination collapses to ad-hoc chat, manual merging, or silent overwrites. We present MPAC (Multi-Principal Agent Coordination Protocol), an application-layer protocol that fills this gap with explicit coordination semantics across five layers: Session, Intent, Operation, Conflict, and Governance. MPAC makes intent declaration a precondition for action, represents conflicts as first-class structured objects, and supports human-in-the-loop arbitration through a pluggable governance layer. The specification defines 21 message types, three state machines with normative transition tables, Lamport-clock causal watermarking, two execution models, three security profiles, and optimistic concurrency control on shared state. We release two interoperable reference implementations in Python and TypeScript with 223 tests, a JSON Schema suite, and seven live multi-agent demos. A controlled three-agent code review benchmark shows a 95 percent reduction in coordination overhead and a 4.8 times wall-clock speedup versus a serialized human-mediated baseline, with per-agent decision time preserved. The speedup comes from eliminating coordination waits, not compressing model calls. Specification, implementations, and demos are open source.", "authors": ["Kaiyang Qian", "Xinmin Fang", "Zhengxiong Li"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-10", "url": "https://arxiv.org/abs/2604.09744", "pdf_url": "https://arxiv.org/pdf/2604.09744v1", "arxiv_id": "2604.09744", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6405}
{"id": "1c2a8bcf5e7ca5f1a2f41e95db9e0471e57f10bf92368b6d194cfe9d3f041c87", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic Attentional Context Scoping: Agent-Triggered Focus Sessions for Isolated Per-Agent Steering in Multi-Agent LLM Orchestration", "abstract": "Multi-agent LLM orchestration systems suffer from context pollution: when N concurrent agents compete for the orchestrator's context window, each agent's task state, partial outputs, and pending questions contaminate the steering interactions of every other agent, degrading decision quality. We introduce Dynamic Attentional Context Scoping (DACS), a mechanism in which the orchestrator operates in two asymmetric modes. In Registry mode it holds only lightweight per-agent status summaries (<=200 tokens each), remaining responsive to all agents and the user. When an agent emits a SteeringRequest, the orchestrator enters Focus(a_i) mode, injecting the full context of agent a_i while compressing all other agents to their registry entries. Context isolation is agent-triggered, asymmetric, and deterministic: the context window contains exactly F(a_i) + R_{-i} during steering, eliminating cross-agent contamination without requiring context compression or retrieval. We evaluate DACS across four experimental phases totalling 200 trials: Phase 1 tests N in {3,5,10} (60 trials); Phase 2 tests agent heterogeneity and adversarial dependencies (60 trials); Phase 3 tests decision density up to D=15 (40 trials); Phase 4 uses autonomous LLM agents for free-form questions (40 trials, Claude Haiku 4.5). Across all 8 synthetic scenarios, DACS achieves 90.0--98.4% steering accuracy versus 21.0--60.0% for a flat-context baseline (p < 0.0001 throughout), with wrong-agent contamination falling from 28--57% to 0--14% and context efficiency ratios of up to 3.53x. The accuracy advantage grows with N and D; keyword matching is validated by LLM-as-judge across all phases (mean kappa=0.909). DACS outperforms the flat-context baseline by +17.2pp at N=3 (p=0.0023) and +20.4pp at N=5 (p=0.0008) in Phase 4, with the advantage growing with N confirmed by two independent judges.", "authors": ["Nickson Patel"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-09", "url": "https://arxiv.org/abs/2604.07911", "pdf_url": "https://arxiv.org/pdf/2604.07911v1", "arxiv_id": "2604.07911", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3442}
{"id": "a6b84a145e6bb518bdea91335087b4fefb49f4c4e3cd5ada5b96d6d5f04b6fcc", "sources": ["arxiv", "semantic_scholar"], "title": "Externalization in LLM Agents: A Unified Review of Memory, Skills, Protocols and Harness Engineering", "abstract": "Large language model (LLM) agents are increasingly built less by changing model weights than by reorganizing the runtime around them. Capabilities that earlier systems expected the model to recover internally are now externalized into memory stores, reusable skills, interaction protocols, and the surrounding harness that makes these modules reliable in practice. This paper reviews that shift through the lens of externalization. Drawing on the idea of cognitive artifacts, we argue that agent infrastructure matters not merely because it adds auxiliary components, but because it transforms hard cognitive burdens into forms that the model can solve more reliably. Under this view, memory externalizes state across time, skills externalize procedural expertise, protocols externalize interaction structure, and harness engineering serves as the unification layer that coordinates them into governed execution. We trace a historical progression from weights to context to harness, analyze memory, skills, and protocols as three distinct but coupled forms of externalization, and examine how they interact inside a larger agent system. We further discuss the trade-off between parametric and externalized capability, identify emerging directions such as self-evolving harnesses and shared agent infrastructure, and discuss open challenges in evaluation, governance, and the long-term co-evolution of models and external infrastructure. The result is a systems-level framework for explaining why practical agent progress increasingly depends not only on stronger models, but on better external cognitive infrastructure.", "authors": ["Chenyu Zhou", "Huacan Chai", "Wenteng Chen", "Zihan Guo", "Rong Shan", "Yuanyi Song", "Tianyi Xu", "Yingxuan Yang", "Aofan Yu", "Weiming Zhang", "Congming Zheng", "Jiachen Zhu", "Zeyu Zheng", "Zhuosheng Zhang", "Xingyu Lou", "Changwang Zhang", "Zhihui Fu", "Jun Wang", "Weiwen Liu", "Jianghao Lin", "Weinan Zhang"], "categories": ["cs.SE", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-09", "url": "https://arxiv.org/abs/2604.08224", "pdf_url": "https://arxiv.org/pdf/2604.08224v1", "arxiv_id": "2604.08224", "doi": null, "citation_count": 25, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3537}
{"id": "e901a0e2ad5428bc2cfcb3c6cab9e24ba0e67804baad899be74cfaabb9fcd798", "sources": ["arxiv", "semantic_scholar"], "title": "EigentSearch-Q+: Enhancing Deep Research Agents with Structured Reasoning Tools", "abstract": "Deep research requires reasoning over web evidence to answer open-ended questions, and it is a core capability for AI agents. Yet many deep research agents still rely on implicit, unstructured search behavior that causes redundant exploration and brittle evidence aggregation. Motivated by Anthropic's \"think\" tool paradigm and insights from the information-retrieval literature, we introduce Q+, a set of query and evidence processing tools that make web search more deliberate by guiding query planning, monitoring search progress, and extracting evidence from long web snapshots. We integrate Q+ into the browser sub-agent of Eigent, an open-source, production-ready multi-agent workforce for computer use, yielding EigentSearch-Q+. Across four benchmarks (SimpleQA-Verified, FRAMES, WebWalkerQA, and XBench DeepSearch), Q+ improves Eigent's browser agent benchmark-size-weighted average accuracy by 3.0, 3.8, and 0.6 percentage points (pp) for GPT-4.1, GPT-5.1, and Minimax M2.5 model backends, respectively. Case studies further suggest that EigentSearch-Q+ produces more coherent tool-calling trajectories by making search progress and evidence handling explicit.", "authors": ["Boer Zhang", "Mingyan Wu", "Dongzhuoran Zhou", "Yuqicheng Zhu", "Wendong Fan", "Puzhen Zhang", "Zifeng Ding", "Guohao Li", "Yuan He"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-09", "url": "https://arxiv.org/abs/2604.07927", "pdf_url": "https://arxiv.org/pdf/2604.07927v3", "arxiv_id": "2604.07927", "doi": "10.1145/3786335.3813186", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6392}
{"id": "9947475e7eb435a6d556eddea2572bf7591412b815ef6e9cf437c3ff826cbeb3", "sources": ["arxiv", "semantic_scholar"], "title": "We Need Strong Preconditions For Using Simulations In Policy", "abstract": "Simulations, and more recently LLM agent simulations, have been adopted as useful tools for policymakers to explore interventions, rehearse potential scenarios, and forecast outcomes. While LLM simulations have enormous potential, two critical challenges remain understudied: the dual-use potential of accurate models of individual or population-level human behavior and the difficulty of validating simulation outputs. In light of these limitations, we must define boundaries for both simulation developers and decision-makers to ensure responsible development and ethical use. We propose and discuss three preconditions for societal-scale LLM agent simulations: 1) do not treat simulations of marginalized populations as neutral technical outputs, 2) do not simulate populations without their participation, and 3) do not simulate without accountability. We believe that these guardrails, combined with our call for simulation development and deployment reports, will help build trust among policymakers while promoting responsible development and use of societal-scale LLM agent simulations for the public benefit.", "authors": ["Steven Luo", "Saanvi Arora", "Carlos Guirado"], "categories": ["cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-09", "url": "https://arxiv.org/abs/2604.07838", "pdf_url": "https://arxiv.org/pdf/2604.07838v1", "arxiv_id": "2604.07838", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3442}
{"id": "963350fcda1a9838837b6d7e67b1540578f0288612389d995163f0026080037b", "sources": ["arxiv", "semantic_scholar"], "title": "TurboAgent: An LLM-Driven Autonomous Multi-Agent Framework for Turbomachinery Aerodynamic Design", "abstract": "The aerodynamic design of turbomachinery is a complex and tightly coupled multi-stage process involving geometry generation, performance prediction, optimization, and high-fidelity physical validation. Existing intelligent design approaches typically focus on individual stages or rely on loosely coupled pipelines, making fully autonomous end-to-end design challenging. To address this issue, this study proposes TurboAgent, a large language model (LLM)-driven autonomous multi-agent framework for turbomachinery aerodynamic design and optimization. The LLM serves as the core for task planning and coordination, while specialized agents handle generative design, rapid performance prediction, multi-objective optimization, and physics-based validation. The framework transforms traditional trial-and-error design into a data-driven collaborative workflow, with high-fidelity simulations retained for final verification. A transonic single-rotor compressor is used for validation. The results show strong agreement between target performance, generated designs, and CFD simulations. The coefficients of determination for mass flow rate, total pressure ratio, and isentropic efficiency all exceed 0.91, with normalized RMSE values below 8%. The optimization agent further improves isentropic efficiency by 1.61% and total pressure ratio by 3.02%. The complete workflow can be executed within approximately 30 minutes under parallel computing. These results demonstrate that TurboAgent enables an autonomous closed-loop design process from natural language requirements to final design generation, providing an efficient and scalable paradigm for turbomachinery aerodynamic design.", "authors": ["Juan Du", "Yueteng Wu", "Pan Zhao", "Yuze Liu", "Min Zhang", "Xiaobin Xu", "Xinglong Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-08", "url": "https://arxiv.org/abs/2604.06747", "pdf_url": "https://arxiv.org/pdf/2604.06747v2", "arxiv_id": "2604.06747", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3434}
{"id": "bf106045d865e4335217fed9ed50bbeba0d005cb1252f7d5a8eaea24d656d6be", "sources": ["arxiv", "semantic_scholar"], "title": "PoC-Adapt: Semantic-Aware Automated Vulnerability Reproduction with LLM Multi-Agents and Reinforcement Learning-Driven Adaptive Policy", "abstract": "While recent approaches leverage large language models (LLMs) and multi-agent pipelines to automatically generate proof-of-concept (PoC) exploits from vulnerability reports, existing systems often suffer from two fundamental limitations: unreliable validation based on surface-level execution signals and high operational cost caused by extensive trial-and-error during exploit generation. In this paper, we present PoC-Adapt, an end-to-end framework for automated PoC generation and verification, architected upon a foundation semantic runtime validation and adaptive policy learning. At the core of PoC-Adapt is a Semantic Oracle that validates exploits by comparing structured pre- and post-execution system states, enabling reliable distinction between true vulnerability exploitation and incidental behavioral changes. To reduce exploration cost, we further introduce an Adaptive Policy Learning mechanism that learns an exploitation policy over semantic states and actions, guiding the exploit agent toward effective strategies with fewer failed attempts. PoC-Adapt is implemented as a multi-agent system comprising specialized agents for root cause analysis, environment building, exploit generation, and semantic validation, coordinated through structured feedback loops. Experimenting on the CWE-Bench-Java and PrimeVul benchmarks shows that PoC-Adapt significantly improves verification reliability by 25% and reduces exploit generation cost compared to prior LLM-based systems, highlighting the importance of semantic validation and learned action policies in automated vulnerability reproduction. Applied to the latest CVE corpus, PoC-Adapt confirmed 12 verified PoC out of 80 reproduce attempts at a cost of $0.42 per generated exploit", "authors": ["Phan The Duy", "Khoa Ngo-Khanh", "Nguyen Huu Quyen", "Van-Hau Pham"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-08", "url": "https://arxiv.org/abs/2604.06618", "pdf_url": "https://arxiv.org/pdf/2604.06618v2", "arxiv_id": "2604.06618", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3434}
{"id": "091d32e7a5884f2a0c24c276c88219326a1bce57b70bfd19f90490c9367b3f28", "sources": ["arxiv", "semantic_scholar"], "title": "Cayley Graph Optimization for Scalable Multi-Agent Communication Topologies", "abstract": "Large-scale multi-agent communication has long faced a scalability bottleneck: fully connected networks require quadratic complexity, yet existing sparse topologies rely on hand-crafted rules. This paper treats the communication graph itself as a design variable and proposes CayleyTopo, a family of circulant Cayley graphs whose generator sets are optimized to minimize diameter, directly targeting worst-case information propagation speed. To navigate the enormous search space of possible generator sets, we develop a lightweight reinforcement learning framework that injects a number-theoretic prior to favor structurally rich generators, alongside a message-propagation score that provides dense connectivity feedback during construction. The resulting CayleyTopo consistently outperforms existing hand-crafted topologies, achieving faster information dissemination, greater resilience to link failures, and lower communication load, all while approaching the theoretical Moore bound. Our study opens the door to scalable, robust, and efficient communication foundations for future multi-agent systems, where the graph itself becomes optimizable rather than a fixed constraint.", "authors": ["Jingkai Luo", "Yulin Shao"], "categories": ["cs.NI", "cs.IT", "cs.MA"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.09703", "pdf_url": "https://arxiv.org/pdf/2604.09703v1", "arxiv_id": "2604.09703", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "a038726fb108c443674b6457abc6c3b79aa0415bc53028ed4fda13d498bbd971", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Pathfinding with Non-Unit Integer Edge Costs via Enhanced Conflict-Based Search and Graph Discretization", "abstract": "Multi-Agent Pathfinding (MAPF) plays a critical role in various domains. Traditional MAPF methods typically assume unit edge costs and single-timestep actions, which limit their applicability to real-world scenarios. MAPFR extends MAPF to handle non-unit costs with real-valued edge costs and continuous-time actions, but its geometric collision model leads to an unbounded state space that compromises solver efficiency. In this paper, we propose MAPFZ, a novel MAPF variant on graphs with non-unit integer costs that preserves a finite state space while offering improved realism over classical MAPF. To solve MAPFZ efficiently, we develop CBS-NIC, an enhanced Conflict-Based Search framework incorporating time-interval-based conflict detection and an improved Safe Interval Path Planning (SIPP) algorithm. Additionally, we propose Bayesian Optimization for Graph Design (BOGD), a discretization method for non-unit edge costs that balances efficiency and accuracy with a sub-linear regret bound. Extensive experiments demonstrate that our approach outperforms state-of-the-art methods in runtime and success rate across diverse benchmark scenarios.", "authors": ["Hongkai Fan", "Qinjing Xie", "Bo Ouyang", "Yaonan Wang", "Zhi Yan", "Jiawen He", "Zheng Fang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05416", "pdf_url": "https://arxiv.org/pdf/2604.05416v1", "arxiv_id": "2604.05416", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "484a5938eae51ade0ec5aa7fce51f6d1caff2beb6ccfb846ffde0a12eade3110", "sources": ["arxiv", "semantic_scholar"], "title": "Your LLM Agent Can Leak Your Data: Data Exfiltration via Backdoored Tool Use", "abstract": "Tool-use large language model (LLM) agents are increasingly deployed to support sensitive workflows, relying on tool calls for retrieval, external API access, and session memory management. While prior research has examined various threats, the risk of systematic data exfiltration by backdoored agents remains underexplored. In this work, we present Back-Reveal, a data exfiltration attack that embeds semantic triggers into fine-tuned LLM agents. When triggered, the backdoored agent invokes memory-access tool calls to retrieve stored user context and exfiltrates it via disguised retrieval tool calls. We further demonstrate that multi-turn interaction amplifies the impact of data exfiltration, as attacker-controlled retrieval responses can subtly steer subsequent agent behavior and user interactions, enabling sustained and cumulative information leakage over time. Our experimental results expose a critical vulnerability in LLM agents with tool access and highlight the need for defenses against exfiltration-oriented backdoors.", "authors": ["Wuyang Zhang", "Shichao Pei"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05432", "pdf_url": "https://arxiv.org/pdf/2604.05432v1", "arxiv_id": "2604.05432", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "86f0a268bc3523671a3afa8285d52ed2b51d7e79dc8cf3fd9f4728ec591b88f8", "sources": ["arxiv", "semantic_scholar"], "title": "FLARE: Agentic Coverage-Guided Fuzzing for LLM-Based Multi-Agent Systems", "abstract": "Multi-Agent LLM Systems (MAS) have been adopted to automate complex human workflows by breaking down tasks into subtasks. However, due to the non-deterministic behavior of LLM agents and the intricate interactions between agents, MAS applications frequently encounter failures, including infinite loops and failed tool invocations. Traditional software testing techniques are ineffective in detecting such failures due to the lack of LLM agent specification, the large behavioral space of MAS, and semantic-based correctness judgment. This paper presents FLARE, a novel testing framework tailored for MAS. FLARE takes the source code of MAS as input and extracts specifications and behavioral spaces from agent definitions. Based on these specifications, FLARE builds test oracles and conducts coverage-guided fuzzing to expose failures. It then analyzes execution logs to judge whether each test has passed and generates failure reports. Our evaluation on 16 diverse open-source applications demonstrates that FLARE achieves 96.9% inter-agent coverage and 91.1% intra-agent coverage, outperforming baselines by 9.5% and 1.0%. FLARE also uncovers 56 previously unknown failures unique to MAS.", "authors": ["Mingxuan Hui", "Xinyue Li", "Lu Wang", "Chengcheng Wan", "Yifan Wang", "Yimian Wang", "Feiyue Song", "Beining Shi", "Yixi Li", "Yaxiao Li"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05289", "pdf_url": "https://arxiv.org/pdf/2604.05289v1", "arxiv_id": "2604.05289", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6365}
{"id": "bf67f2d665e0930ad5aa2e72403f06c031d607968b645e67ef209747cbc1e73d", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Interrupt in Language-based Multi-agent Communication", "abstract": "Multi-agent systems using large language models (LLMs) have demonstrated impressive capabilities across various domains. However, current agent communication suffers from verbose output that overload context and increase computational costs. Although existing approaches focus on compressing the message from the speaker side, they struggle to adapt to different listeners and identify relevant information. An effective way in human communication is to allow the listener to interrupt and express their opinion or ask for clarification. Motivated by this, we propose an interruptible communication framework that allows the agent who is listening to interrupt the current speaker. Through prompting experiments, we find that current LLMs are often overconfident and interrupt before receiving enough information. Therefore, we propose a learning method that predicts the appropriate interruption points based on the estimated future reward and cost. We evaluate our framework across various multi-agent scenarios, including 2-agent text pictionary games, 3-agent meeting scheduling, and 3-agent debate. The results of the experiment show that our HANDRAISER can reduce the communication cost by 32.2% compared to the baseline with comparable or superior task performance. This learned interruption behavior can also be generalized to different agents and tasks.", "authors": ["Danqing Wang", "Da Yin", "Ruta Desai", "Lei Li", "Asli Celikyilmaz", "Ansong Ni"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.06452", "pdf_url": "https://arxiv.org/pdf/2604.06452v1", "arxiv_id": "2604.06452", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "03d99dcd06453f97ae958864f4d0bdf03633b80b1d875a1d26ca8adfcfa61347", "sources": ["arxiv", "semantic_scholar"], "title": "Human Values Matter: Investigating How Misalignment Shapes Collective Behaviors in LLM Agent Communities", "abstract": "As LLMs become increasingly integrated into human society, evaluating their orientations on human values from social science has drawn growing attention. Nevertheless, it is still unclear why human values matter for LLMs, especially in LLM-based multi-agent systems, where group-level failures may accumulate from individually misaligned actions. We ask whether misalignment with human values alters the collective behavior of LLM agents and what changes it induces? In this work, we introduce CIVA, a controlled multi-agent environment grounded in social science theories, where LLM agents form a community and autonomously communicate, explore, and compete for resources, enabling systematic manipulation of value prevalence and behavioral analysis. Through comprehensive simulation experiments, we reveal three key findings. (1) We identify several structurally critical values that substantially shape the community's collective dynamics, including those diverging from LLMs' original orientations. Triggered by the misspecification of these values, we (2) detect system failure modes, e.g., catastrophic collapse, at the macro level, and (3) observe emergent behaviors like deception and power-seeking at the micro level. These results offer quantitative evidence that human values are essential for collective outcomes in LLMs and motivate future multi-agent value alignment.", "authors": ["Xiangxu Zhang", "Jiamin Wang", "Qinlin Zhao", "Hanze Guo", "Linzhuo Li", "Jing Yao", "Xiao Zhou", "Xiaoyuan Yi", "Xing Xie"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05339", "pdf_url": "https://arxiv.org/pdf/2604.05339v1", "arxiv_id": "2604.05339", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "c65a2e843daa95b8fa93539ee66bbe81e6bf7cbc184c1d8f9e53beb307715ac5", "sources": ["arxiv", "semantic_scholar"], "title": "AgentGL: Towards Agentic Graph Learning with LLMs via Reinforcement Learning", "abstract": "Large Language Models (LLMs) increasingly rely on agentic capabilities-iterative retrieval, tool use, and decision-making-to overcome the limits of static, parametric knowledge. Yet existing agentic frameworks treat external information as unstructured text and fail to leverage the topological dependencies inherent in real-world data. To bridge this gap, we introduce Agentic Graph Learning (AGL), a paradigm that reframes graph learning as an interleaved process of topology-aware navigation and LLM-based inference. Specifically, we propose AgentGL, the first reinforcement learning (RL)-driven framework for AGL. AgentGL equips an LLM agent with graph-native tools for multi-scale exploration, regulates tool usage via search-constrained thinking to balance accuracy and efficiency, and employs a graph-conditioned curriculum RL strategy to stabilize long-horizon policy learning without step-wise supervision. Across diverse Text-Attributed Graph (TAG) benchmarks and multiple LLM backbones, AgentGL substantially outperforms strong GraphLLMs and GraphRAG baselines, achieving absolute improvements of up to 17.5% in node classification and 28.4% in link prediction. These results demonstrate that AGL is a promising frontier for enabling LLMs to autonomously navigate and reason over complex relational environments. The code is publicly available at https://github.com/sunyuanfu/AgentGL.", "authors": ["Yuanfu Sun", "Kang Li", "Dongzhe Fan", "Jiajin Liu", "Qiaoyu Tan"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05846", "pdf_url": "https://arxiv.org/pdf/2604.05846v2", "arxiv_id": "2604.05846", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/sunyuanfu/AgentGL", "venue": null, "quality_score": 0.6365}
{"id": "27bb42a5fb080bb1b1e2005c9fb9f1212e48aec2e90835aabbf7eb0a268c1fdd", "sources": ["arxiv", "semantic_scholar"], "title": "Unsupervised Multi-agent and Single-agent Perception from Cooperative Views", "abstract": "The LiDAR-based multi-agent and single-agent perception has shown promising performance in environmental understanding for robots and automated vehicles. However, there is no existing method that simultaneously solves both multi-agent and single-agent perception in an unsupervised way. By sharing sensor data between multiple agents via communication, this paper discovers two key insights: 1) Improved point cloud density after the data sharing from cooperative views could benefit unsupervised object classification, 2) Cooperative view of multiple agents can be used as unsupervised guidance for the 3D object detection in the single view. Based on these two discovered insights, we propose an Unsupervised Multi-agent and Single-agent (UMS) perception framework that leverages multi-agent cooperation without human annotations to simultaneously solve multi-agent and single-agent perception. UMS combines a learning-based Proposal Purifying Filter to better classify the candidate proposals after multi-agent point cloud density cooperation, followed by a Progressive Proposal Stabilizing module to yield reliable pseudo labels by the easy-to-hard curriculum learning. Furthermore, we design a Cross-View Consensus Learning to use multi-agent cooperative view to guide detection in single-agent view. Experimental results on two public datasets V2V4Real and OPV2V show that our UMS method achieved significantly higher 3D detection performance than the state-of-the-art methods on both multi-agent and single-agent perception tasks in an unsupervised setting.", "authors": ["Haochen Yang", "Baolu Li", "Lei Li", "Delin Ren", "Jiacheng Guo", "Minghai Qin", "Tianyun Zhang", "Hongkai Yu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.05354", "pdf_url": "https://arxiv.org/pdf/2604.05354v1", "arxiv_id": "2604.05354", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "075a17ab20c6a4ad6eb23be29373742fa168ac2f5446b92e27e3ad99d1f8f003", "sources": ["arxiv", "semantic_scholar"], "title": "SkillGraph: Graph Foundation Priors for LLM Agent Tool Sequence Recommendation", "abstract": "LLM agents must select tools from large API libraries and order them correctly. Existing methods use semantic similarity for both retrieval and ordering, but ordering depends on inter-tool data dependencies that are absent from tool descriptions. As a result, semantic-only methods can produce negative Kendall-$τ$ in structured workflow domains. We introduce SkillGraph, a directed weighted execution-transition graph mined from 49,831 successful LLM agent trajectories, which encodes workflow-precedence regularities as a reusable graph foundation prior. Building on this graph foundation prior, we propose a two-stage decoupled framework: GS-Hybrid retrieval for candidate selection and a learned pairwise reranker for ordering. On ToolBench (9,965 test instances; ~16,000 tools), the method reaches Set-F1 = 0.271 and Kendall-$τ$ = 0.096; on API-Bank, Kendall-$τ$ improves from -0.433 to +0.613. Under identical Stage-1 inputs, the learned reranker also outperforms LLaMA-3.1-8B Stage-2 rerankers.", "authors": ["Hao Liu", "Dongyu Li"], "categories": ["cs.AI", "cs.CL", "cs.IR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.19793", "pdf_url": "https://arxiv.org/pdf/2604.19793v1", "arxiv_id": "2604.19793", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3427}
{"id": "8750e803aecf4126a967f04036f0b2e5d6b0e3efa715f75cf38bb13d3b30224b", "sources": ["arxiv", "semantic_scholar"], "title": "Paper Circle: An Open-source Multi-agent Research Discovery and Analysis Framework", "abstract": "The rapid growth of scientific literature has made it increasingly difficult for researchers to efficiently discover, evaluate, and synthesize relevant work. Recent advances in multi-agent large language models (LLMs) have demonstrated strong potential for understanding user intent and are being trained to utilize various tools. In this paper, we introduce Paper Circle, a multi-agent research discovery and analysis system designed to reduce the effort required to find, assess, organize, and understand academic literature. The system comprises two complementary pipelines: (1) a Discovery Pipeline that integrates offline and online retrieval from multiple sources, multi-criteria scoring, diversity-aware ranking, and structured outputs; and (2) an Analysis Pipeline that transforms individual papers into structured knowledge graphs with typed nodes such as concepts, methods, experiments, and figures, enabling graph-aware question answering and coverage verification. Both pipelines are implemented within a coder LLM-based multi-agent orchestration framework and produce fully reproducible, synchronized outputs including JSON, CSV, BibTeX, Markdown, and HTML at each agent step. This paper describes the system architecture, agent roles, retrieval and scoring methods, knowledge graph schema, and evaluation interfaces that together form the Paper Circle research workflow. We benchmark Paper Circle on both paper retrieval and paper review generation, reporting hit rate, MRR, and Recall at K. Results show consistent improvements with stronger agent models. We have publicly released the website at https://papercircle.vercel.app/ and the code at https://github.com/MAXNORM8650/papercircle.", "authors": ["Komal Kumar", "Aman Chadha", "Salman Khan", "Fahad Shahbaz Khan", "Hisham Cholakkal"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-07", "url": "https://arxiv.org/abs/2604.06170", "pdf_url": "https://arxiv.org/pdf/2604.06170v1", "arxiv_id": "2604.06170", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/MAXNORM8650/papercircle", "venue": null, "quality_score": 0.6365}
{"id": "dabf01c54ab1197f9b49601e754d173a66f510b558cc11aff5f383692dac8822", "sources": ["arxiv", "semantic_scholar"], "title": "ANX: Protocol-First Design for AI Agent Interaction with a Supporting 3EX Decoupled Architecture", "abstract": "AI agents, autonomous digital actors, need agent-native protocols; existing methods include GUI automation and MCP-based skills, with defects of high token consumption, fragmented interaction, inadequate security, due to lacking a unified top-level framework and key components, each independent module flawed. To address these issues, we present ANX, an open, extensible, verifiable agent-native protocol and top-level framework integrating CLI, Skill, MCP, resolving pain points via protocol innovation, architectural optimization and tool supplementation. Its four core innovations: 1) Agent-native design (ANX Config, Markup, CLI) with high information density, flexibility and strong adaptability to reduce tokens and eliminate inconsistencies; 2) Human-agent interaction combining Skill's flexibility for dual rendering as agent-executable instructions and human-readable UI; 3) MCP-supported on-demand lightweight apps without pre-registration; 4) ANX Markup-enabled machine-executable SOPs eliminating ambiguity for reliable long-horizon tasks and multi-agent collaboration. As the first in a series, we focus on ANX's design, present its 3EX decoupled architecture with ANXHub and preliminary feasibility analysis and experimental validation. ANX ensures native security: LLM-bypassed UI-to-Core communication keeps sensitive data out of agent context; human-only confirmation prevents automated misuse. Form-filling experiments with Qwen3.5-plus/GPT-4o show ANX reduces tokens by 47.3% (Qwen3.5-plus) and 55.6% (GPT-4o) vs MCP-based skills, 57.1% (Qwen3.5-plus) and 66.3% (GPT-4o) vs GUI automation, and shortens execution time by 58.1% and 57.7% vs MCP-based skills.", "authors": ["Xu Mingze"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-06", "url": "https://arxiv.org/abs/2604.04820", "pdf_url": "https://arxiv.org/pdf/2604.04820v1", "arxiv_id": "2604.04820", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/mountorc/anx-protocol", "venue": null, "quality_score": 0.6351}
{"id": "6e8e0cbd38b2937c2bedca1ce94a4f1f970f4b2fa804d2c2f8177070f5624f1a", "sources": ["arxiv", "semantic_scholar"], "title": "HDP: A Lightweight Cryptographic Protocol for Human Delegation Provenance in Agentic AI Systems", "abstract": "Agentic AI systems increasingly execute consequential actions on behalf of human principals, delegating tasks through multi-step chains of autonomous agents. No existing standard addresses a fundamental accountability gap: verifying that terminal actions in a delegation chain were genuinely authorized by a human principal, through what chain of delegation, and under what scope. This paper presents the Human Delegation Provenance (HDP) protocol, a lightweight token-based scheme that cryptographically captures and verifies human authorization context in multi-agent systems. An HDP token binds a human authorization event to a session, records each agent's delegation action as a signed hop in an append-only chain, and enables any participant to verify the full provenance record using only the issuer's Ed25519 public key and the current session identifier. Verification is fully offline, requiring no registry lookups or third-party trust anchors. We situate HDP within the existing landscape of delegation protocols, identify its distinct design point relative to OAuth 2.0 Token Exchange (RFC 8693), JSON Web Tokens (RFC 7519), UCAN, and the Intent Provenance Protocol (draft-haberkamp-ipp-00), and demonstrate that existing standards fail to address the multi-hop, append-only, human-provenance requirements of agentic systems. HDP has been published as an IETF Internet-Draft (draft-helixar-hdp-agentic-delegation-00) and a reference TypeScript SDK is publicly available.", "authors": ["Asiri Dalugoda"], "categories": ["cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-06", "url": "https://arxiv.org/abs/2604.04522", "pdf_url": "https://arxiv.org/pdf/2604.04522v1", "arxiv_id": "2604.04522", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Helixar-AI/HDP", "venue": null, "quality_score": 0.6351}
{"id": "1feb649ce8fb2422f325d834c70647bd19d81f22e8a9daf97e46cf5a78ef5418", "sources": ["arxiv", "semantic_scholar"], "title": "Governance-Aware Agent Telemetry for Closed-Loop Enforcement in Multi-Agent AI Systems", "abstract": "Enterprise multi-agent AI systems produce thousands of inter-agent interactions per hour, yet existing observability tools capture these dependencies without enforcing anything. OpenTelemetry and Langfuse collect telemetry but treat governance as a downstream analytics concern, not a real-time enforcement target. The result is an \"observe-but-do-not-act\" gap where policy violations are detected only after damage is done. We present Governance-Aware Agent Telemetry (GAAT), a reference architecture that closes the loop between telemetry collection and automated policy enforcement for multi-agent systems. GAAT introduces (1) a Governance Telemetry Schema (GTS) extending OpenTelemetry with governance attributes; (2) a real-time policy violation detection engine using OPA-compatible declarative rules under sub-200 ms latency; (3) a Governance Enforcement Bus (GEB) with graduated interventions; and (4) a Trusted Telemetry Plane with cryptographic provenance.", "authors": ["Anshul Pathak", "Nishant Jain"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-06", "url": "https://arxiv.org/abs/2604.05119", "pdf_url": "https://arxiv.org/pdf/2604.05119v1", "arxiv_id": "2604.05119", "doi": null, "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.342}
{"id": "06417a32fc936613373fecc075f563ef7227e112740a5a4bca3f04e14a4fe40b", "sources": ["arxiv", "semantic_scholar"], "title": "Optimizing Service Operations via LLM-Powered Multi-Agent Simulation", "abstract": "Service system performance depends on how participants respond to design choices, but modeling these responses is hard due to the complexity of human behavior. We introduce an LLM-powered multi-agent simulation (LLM-MAS) framework for optimizing service operations. We pose the problem as stochastic optimization with decision-dependent uncertainty: design choices are embedded in prompts and shape the distribution of outcomes from interacting LLM-powered agents. By embedding key numerical information in prompts and extracting it from LLM-generated text, we model this uncertainty as a controlled Markov chain. We develop an on-trajectory learning algorithm that, on a single simulation run, simultaneously constructs zeroth-order gradient estimates and updates design parameters to optimize steady-state performance. We also incorporate variance reduction techniques. In a sustainable supply chain application, our method outperforms benchmarks, including blackbox optimization and using LLMs as numerical solvers or as role-playing system designers. A case study on optimal contest design with real behavioral data shows that LLM-MAS is both as a cost-effective evaluator of known designs and an exploratory tool that can uncover strong designs overlooked by traditional approaches.", "authors": ["Yanyuan Wang", "Xiaowei Zhang"], "categories": ["cs.AI", "cs.MA", "math.OC"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-04-06", "url": "https://arxiv.org/abs/2604.04383", "pdf_url": "https://arxiv.org/pdf/2604.04383v1", "arxiv_id": "2604.04383", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.342}
{"id": "764a22a024293d9411c5177b4c7219376cf890764960751a150a414f7430d433", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Framework for Automated Exploit Generation with Constraint-Guided Comprehension and Reflection", "abstract": "Open-source libraries are widely used in modern software development, introducing significant security vulnerabilities. While static analysis tools can identify potential vulnerabilities at scale, they often generate overwhelming reports with high false positive rates. Automated Exploit Generation (AEG) emerges as a promising solution to confirm vulnerability authenticity by generating an exploit. However, traditional AEG approaches based on fuzzing or symbolic execution face path coverage and constraint-solving problems. Although LLMs show great potential for AEG, how to effectively leverage them to comprehend vulnerabilities and generate corresponding exploits is still an open question. To address these challenges, we propose Vulnsage, a multi-agent framework for AEG. Vulnsage simulates human security researchers' workflows by decomposing the complex AEG process into multiple specialized sub-agents: Code Analyzer Agent, Code Generation Agent, Validation Agent, and a set of Reflection Agents, orchestrated by a central supervisor through iterative cycles. Given a target program, the Code Analyzer Agent performs static analysis to identify potential vulnerabilities and collects relevant information for each one. The Code Generation Agent then utilizes an LLM to generate candidate exploits. The Validation Agent and Reflection Agents form a feedback-driven self-refinement loop that uses execution traces and runtime error analysis to either improve the exploit iteratively or reason about the false positive alert. Experimental evaluation demonstrates that Vulnsage succeeds in generating 34.64\\% more exploits than state-of-the-art tools such as \\explodejs. Furthermore, Vulnsage has successfully discovered and verified 146 zero-day vulnerabilities in real-world scenarios, demonstrating its practical effectiveness for assisting security assessment in software supply chains.", "authors": ["Siyi Chen", "Tianhan Luo", "Shijian Wu", "Xiangyu Liu", "Yilin Zhou", "Qi Li", "Wenyuan Xu"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-06", "url": "https://arxiv.org/abs/2604.05130", "pdf_url": "https://arxiv.org/pdf/2604.05130v1", "arxiv_id": "2604.05130", "doi": "10.1145/3794763.3794817", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "34th IEEE/ACM International Conference on Program Comprehension (ICPC '26), April 12--13, 2026, Rio de Janeiro, Brazil", "quality_score": 0.8305}
{"id": "03ea3efba935471065dd231d17ca9748690e05c6a9126899879201643caeedc5", "sources": ["arxiv", "semantic_scholar"], "title": "Causality Laundering: Denial-Feedback Leakage in Tool-Calling LLM Agents", "abstract": "Tool-calling LLM agents can read private data, invoke external services, and trigger real-world actions, creating a security problem at the point of tool execution. We identify a denial-feedback leakage pattern, which we term causality laundering, in which an adversary probes a protected action, learns from the denial outcome, and exfiltrates the inferred information through a later seemingly benign tool call. This attack is not captured by flat provenance tracking alone because the leaked information arises from causal influence of the denied action, not direct data flow. We present the Agentic Reference Monitor (ARM), a runtime enforcement layer that mediates every tool invocation by consulting a provenance graph over tool calls, returned data, field-level provenance, and denied actions. ARM propagates trust through an integrity lattice and augments the graph with counterfactual edges from denied-action nodes, enabling enforcement over both transitive data dependencies and denial-induced causal influence. In a controlled evaluation on three representative attack scenarios, ARM blocks causality laundering, transitive taint propagation, and mixed-provenance field misuse that a flat provenance baseline misses, while adding sub-millisecond policy evaluation overhead. These results suggest that denial-aware causal provenance is a useful abstraction for securing tool-calling agent systems.", "authors": ["Mohammad Hossein Chinaei"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-05", "url": "https://arxiv.org/abs/2604.04035", "pdf_url": "https://arxiv.org/pdf/2604.04035v1", "arxiv_id": "2604.04035", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3412}
{"id": "bd53106f7143895beac3dd32026f7b2c43fa6ce096e2baa9d97e405ddd774a57", "sources": ["arxiv", "semantic_scholar"], "title": "CoopGuard: Stateful Cooperative Agents Safeguarding LLMs Against Evolving Multi-Round Attacks", "abstract": "As Large Language Models (LLMs) are increasingly deployed in complex applications, their vulnerability to adversarial attacks raises urgent safety concerns, especially those evolving over multi-round interactions. Existing defenses are largely reactive and struggle to adapt as adversaries refine strategies across rounds. In this work, we propose CoopGuard , a stateful multi-round LLM defense framework based on cooperative agents that maintains and updates an internal defense state to counter evolving attacks. It employs three specialized agents (Deferring Agent, Tempting Agent, and Forensic Agent) for complementary round-level strategies, coordinated by System Agent, which conditions decisions on the evolving defense state (interaction history) and orchestrates agents over time. To evaluate evolving threats, we introduce the EMRA benchmark with 5,200 adversarial samples across 8 attack types, simulating progressively LLM multi-round attacks. Experiments show that CoopGuard reduces attack success rate by 78.9% over state-of-the-art defenses, while improving deceptive rate by 186% and reducing attack efficiency by 167.9%, offering a more comprehensive assessment of multi-round defense. These results demonstrate that CoopGuard provides robust protection for LLMs in multi-round adversarial scenarios.", "authors": ["Siyuan Li", "Zehao Liu", "Xi Lin", "Qinghua Mao", "Yuliang Chen", "Haoyu Li", "Jun Wu", "Jianhua Li", "Xiu Su"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-05", "url": "https://arxiv.org/abs/2604.04060", "pdf_url": "https://arxiv.org/pdf/2604.04060v1", "arxiv_id": "2604.04060", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3412}
{"id": "5f077938293cf0092ac58ed365dac07f4a9f79c503fc786c96f3b5318b6cd4b7", "sources": ["arxiv", "semantic_scholar"], "title": "GeoBrowse: A Geolocation Benchmark for Agentic Tool Use with Expert-Annotated Reasoning Traces", "abstract": "Deep research agents integrate fragmented evidence through multi-step tool use. BrowseComp offers a text-only testbed for such agents, but existing multimodal benchmarks rarely require both weak visual cues composition and BrowseComp-style multi-hop verification. Geolocation is a natural testbed because answers depend on combining multiple ambiguous visual cues and validating them with open-web evidence. Thus, we introduce GeoBrowse, a geolocation benchmark that combines visual reasoning with knowledge-intensive multi-hop queries. Level 1 tests extracting and composing fragmented visual cues, and Level 2 increases query difficulty by injecting long-tail knowledge and obfuscating key entities. To support evaluation, we provide an agentic workflow GATE with five think-with-image tools and four knowledge-intensive tools, and release expert-annotated stepwise traces grounded in verifiable evidence for trajectory-level analysis. Experiments show that GATE outperforms direct inference and open-source agents, indicating that no-tool, search-only or image-only setups are insufficient. Gains come from coherent, level-specific tool-use plans rather than more tool calls, as they more reliably reach annotated key evidence steps and make fewer errors when integrating into the final decision. The GeoBrowse bernchmark and codes are provided in https://github.com/ornamentt/GeoBrowse", "authors": ["Xinyu Geng", "Yanjing Xiao", "Yuyang Zhang", "Hanwen Wang", "Xinyan Liu", "Rui Min", "Tianqing Fang", "Yi R. Fung"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-05", "url": "https://arxiv.org/abs/2604.04017", "pdf_url": "https://arxiv.org/pdf/2604.04017v1", "arxiv_id": "2604.04017", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ornamentt/GeoBrowse", "venue": null, "quality_score": 0.6337}
{"id": "6c2e5be54e7b9d91d0fbb8823b44b37dad4bb29d201cae71e4e65bfb0d470769", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Code Optimization via Compiler-LLM Cooperation", "abstract": "Generating performant executables from high level languages is critical to software performance across a wide range of domains. Modern compilers perform this task by passing code through a series of well-studied optimizations at progressively lower levels of abstraction, but may miss optimization opportunities that require high-level reasoning about a program's purpose. Recent work has proposed using LLMs to fill this gap. While LLMs can achieve large speedups on some programs, they frequently generate code that is incorrect. In this work, we propose a method to balance the correctness of conventional compiler optimizations with the ``creativity'' of LLM-based code generation: compiler-LLM cooperation. Our approach integrates existing compiler optimization passes with LLM-based code generation at multiple levels of abstraction, retaining the best features of both types of code optimization. We realize our approach with a multi-agent system that includes (1) LLM-based optimization agents for each level of abstraction, (2) individual compiler constituents as tools, (3) an LLM-based test generation agent that probes the correctness and performance of generated code, and (4) a guiding LLM that orchestrates the other components. The strategy enables LLM-based optimization of input programs at multiple levels of abstraction and introduces a method for distributing computational budget between levels. Our extensive evaluation shows that compiler-LLM cooperation outperforms both existing compiler optimizations and level-specific LLM-based baselines, producing speedups up to 1.25x.", "authors": ["Benjamin Mikek", "Danylo Vashchilenko", "Bryan Lu", "Panpan Xu"], "categories": ["cs.PL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-05", "url": "https://arxiv.org/abs/2604.04238", "pdf_url": "https://arxiv.org/pdf/2604.04238v1", "arxiv_id": "2604.04238", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3412}
{"id": "0ef143c3ec21cf2618b208e0639a850aa3c345a5038fa08d35a37394e4ef65a7", "sources": ["arxiv", "semantic_scholar"], "title": "Your Agent is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs", "abstract": "The rapid deployment of open-source frameworks has significantly advanced the development of modern multi-agent systems. However, expanded action spaces, including uncontrolled privilege exposure and hidden inter-system interactions, pose severe security challenges. Specifically, Indirect Prompt Injections (IPI), which conceal malicious instructions within third-party content, can trigger unauthorized actions such as data exfiltration during normal operations. While current security evaluations predominantly rely on isolated single-turn benchmarks, the systemic vulnerabilities of these agents within complex dynamic environments remain critically underexplored. To bridge this gap, we systematically evaluate six defense strategies against four sophisticated IPI attack vectors across nine LLM backbones. Crucially, we conduct our evaluation entirely within dynamic multi-step tool-calling environments to capture the true attack surface of modern autonomous agents. Moving beyond binary success rates, our multidimensional analysis reveals a pronounced fragility. Advanced injections successfully bypass nearly all baseline defenses, and some surface-level mitigations even produce counterproductive side effects. Furthermore, while agents execute malicious instructions almost instantaneously, their internal states exhibit abnormally high decision entropy. Motivated by this latent hesitation, we investigate Representation Engineering (RepE) as a robust detection strategy. By extracting hidden states at the tool-input position, we revealed that the RepE-based circuit breaker successfully identifies and intercepts unauthorized actions before the agent commits to them, achieving high detection accuracy across diverse LLM backbones. This study exposes the limitations of current IPI defenses and provides a highly practical paradigm for building resilient multi-agent architectures.", "authors": ["Wenhui Zhu", "Xuanzhao Dong", "Xiwen Chen", "Rui Cai", "Peijie Qiu", "Zhipeng Wang", "Oana Frunza", "Shao Tang", "Jindong Gu", "Yalin Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-04", "url": "https://arxiv.org/abs/2604.03870", "pdf_url": "https://arxiv.org/pdf/2604.03870v1", "arxiv_id": "2604.03870", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6324}
{"id": "ae38fa4f6be7a2560a1bc4a79e1102595d16041db4993c38978687f5d45039f2", "sources": ["arxiv", "semantic_scholar"], "title": "Single-agent vs. Multi-agents for Automated Video Analysis of On-Screen Collaborative Learning Behaviors", "abstract": "On-screen learning behavior provides valuable insights into how students seek, use, and create information during learning. Analyzing on-screen behavioral engagement is essential for capturing students' cognitive and collaborative processes. The recent development of Vision Language Models (VLMs) offers new opportunities to automate the labor-intensive manual coding often required for multimodal video data analysis. In this study, we compared the performance of both leading closed-source VLMs (Claude-3.7-Sonnet, GPT-4.1) and open-source VLM (Qwen2.5-VL-72B) in single- and multi-agent settings for automated coding of screen recordings in collaborative learning contexts based on the ICAP framework. In particular, we proposed and compared two multi-agent frameworks: 1) a three-agent workflow multi-agent system (MAS) that segments screen videos by scene and detects on-screen behaviors using cursor-informed VLM prompting with evidence-based verification; 2) an autonomous-decision MAS inspired by ReAct that iteratively interleaves reasoning, tool-like operations (segmentation/ classification/ validation), and observation-driven self-correction to produce interpretable on-screen behavior labels. Experimental results demonstrated that the two proposed MAS frameworks achieved viable performance, outperforming the single VLMs in scene and action detection tasks. It is worth noting that the workflow-based agent achieved best on scene detection, and the autonomous-decision MAS achieved best on action detection. This study demonstrates the effectiveness of VLM-based Multi-agent System for video analysis and contributes a scalable framework for multimodal data analytics.", "authors": ["Likai Peng", "Shihui Feng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-04", "url": "https://arxiv.org/abs/2604.03631", "pdf_url": "https://arxiv.org/pdf/2604.03631v1", "arxiv_id": "2604.03631", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6324}
{"id": "b0eaff4e7d7a84691c4bb1a1e8f903784ef63e95771c376870b0320063725e59", "sources": ["arxiv", "semantic_scholar"], "title": "Representational Collapse in Multi-Agent LLM Committees: Measurement and Diversity-Aware Consensus", "abstract": "Multi-agent LLM committees replicate the same model under different role prompts and aggregate outputs by majority vote, implicitly assuming that agents contribute complementary evidence. We embed each agent's chain-of-thought rationale and measure pairwise similarity: across 100 GSM8K questions with three Qwen2.5-14B agents, mean cosine similarity is 0.888 and effective rank is 2.17 out of 3.0, a failure mode we term representational collapse. DALC, a training-free consensus protocol that computes diversity weights from embedding geometry, reaches 87% on GSM8K versus 84% for self-consistency at 26% lower token cost. Ablation experiments reveal 1-3 point per-protocol run-to-run variance, confirm that hint sharing contributes more than diversity weighting alone, and show that encoder choice strongly modulates collapse severity (cosine 0.908 with mxbai versus 0.888 with nomic) and downstream accuracy. The more robust finding is that collapse is measurable, worsens on harder tasks, and that the choice of embedding proxy is a first-order design decision for any latent communication protocol.", "authors": ["Dipkumar Patel"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-04", "url": "https://arxiv.org/abs/2604.03809", "pdf_url": "https://arxiv.org/pdf/2604.03809v1", "arxiv_id": "2604.03809", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3405}
{"id": "52372b09f1ce72466292fc92f0ee15cfd2465f1e44ae845a5ed92ad2a886e2e2", "sources": ["arxiv", "semantic_scholar"], "title": "SGTA: Scene-Graph Based Multi-Modal Traffic Agent for Video Understanding", "abstract": "We present Scene-Graph Based Multi-Modal Traffic Agent (SGTA), a modular framework for traffic video understanding that combines structured scene graphs with multi-modal reasoning. It constructs a traffic scene graph from roadside videos using detection, tracking, and lane extraction, followed by tool-based reasoning over both symbolic graph queries and visual inputs. SGTA adopts ReAct to process interleaved reasoning traces from large language models with tool invocations, enabling interpretable decision-making for complex video questions. Experiments on selected TUMTraffic VideoQA dataset sample demonstrate that SGTA achieves competitive accuracy across multiple question types while providing transparent reasoning steps. These results highlight the potential of integrating structured scene representations with multi-modal agents for traffic video understanding.", "authors": ["Xingcheng Zhou", "Mingyu Liu", "Walter Zimmer", "Jiajie Zhang", "Alois Knoll"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-04", "url": "https://arxiv.org/abs/2604.03697", "pdf_url": "https://arxiv.org/pdf/2604.03697v1", "arxiv_id": "2604.03697", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3405}
{"id": "ba85a295aa28228a7640b1b129c0ce5e331f4d7c911c5b6c122eafafac91889d", "sources": ["arxiv", "semantic_scholar"], "title": "From Model-Based Screening to Data-Driven Surrogates: A Multi-Stage Workflow for Exploring Stochastic Agent-Based Models", "abstract": "Systematic exploration of Agent-Based Models (ABMs) is challenged by the curse of dimensionality and their inherent stochasticity. We present a multi-stage pipeline integrating the systematic design of experiments with machine learning surrogates. Using a predator-prey case study, our methodology proceeds in two steps. First, an automated model-based screening identifies dominant variables, assesses outcome variability, and segments the parameter space. Second, we train Machine Learning models to map the remaining nonlinear interaction effects. This approach automates the discovery of unstable regions where system outcomes are highly dependent on nonlinear interactions between many variables. Thus, this work provides modelers with a rigorous, hands-off framework for sensitivity analysis and policy testing, even when dealing with high-dimensional stochastic simulators.", "authors": ["Paul Saves", "Matthieu Mastio", "Nicolas Verstaevel", "Benoit Gaudou"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.03350", "pdf_url": "https://arxiv.org/pdf/2604.03350v1", "arxiv_id": "2604.03350", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Multi-Agent-Based Simulation (MABS) XXVII. LCNS Springer, 2026", "quality_score": 0.534}
{"id": "cafeb96e1dcf2d326df7d7a9f88fa2929ca096b097794ac324f03c417ce8874b", "sources": ["arxiv", "semantic_scholar"], "title": "SentinelAgent: Intent-Verified Delegation Chains for Securing Federal Multi-Agent AI Systems", "abstract": "When Agent A delegates to Agent B, which invokes Tool C on behalf of User X, no existing framework can answer: whose authorization chain led to this action, and where did it violate policy? This paper introduces SentinelAgent, a formal framework for verifiable delegation chains in federal multi-agent AI systems. The Delegation Chain Calculus (DCC) defines seven properties - six deterministic (authority narrowing, policy preservation, forensic reconstructibility, cascade containment, scope-action conformance, output schema conformance) and one probabilistic (intent preservation) - with four meta-theorems and one proposition establishing the practical infeasibility of deterministic intent verification. The Intent-Preserving Delegation Protocol (IPDP) enforces all seven properties at runtime through a non-LLM Delegation Authority Service. A three-point verification lifecycle achieves 100% combined TPR at 0% FPR on DelegationBench v4 (516 scenarios, 10 attack categories, 13 federal domains). Under black-box adversarial conditions, the DAS blocks 30/30 attacks with 0 false positives. Deterministic properties are unbreakable under adversarial stress testing; intent verification degrades to 13% against sophisticated paraphrasing. Fine-tuning the NLI model on 190 government delegation examples improves P2 from 1.7% to 88.3% TPR (5-fold cross-validated, F1=82.1%). Properties P1, P3-P7 are mechanically verified via TLA+ model checking across 2.7 million states with zero violations. Even when intent verification is evaded, the remaining six properties constrain the adversary to permitted API calls, conformant outputs, traceable actions, bounded cascades, and compliant behavior.", "authors": ["KrishnaSaiReddy Patil"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02767", "pdf_url": "https://arxiv.org/pdf/2604.02767v1", "arxiv_id": "2604.02767", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "5f4519ebd6025ca82492bcb74e405460a4746f629d33a56e0d27ac7d3bd8234c", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Turn Reinforcement Learning for Tool-Calling Agents with Iterative Reward Calibration", "abstract": "Training tool-calling agents with reinforcement learning on multi-turn tasks remains challenging due to sparse outcome rewards and difficult credit assignment across conversation turns. We present the first application of MT-GRPO (Multi-Turn Group Relative Policy Optimization) combined with GTPO (Generalized Token-level Policy Optimization) for training a tool-calling agent on realistic customer service tasks with an LLM-based user simulator. Through systematic analysis of training rollouts, we discover that naively designed dense per-turn rewards degrade performance by up to 14 percentage points due to misalignment between reward discriminativeness and advantage direction. We introduce Iterative Reward Calibration, a methodology for designing per-turn rewards using empirical discriminative analysis of rollout data, and show that our GTPO hybrid advantage formulation eliminates the advantage misalignment problem. Applied to the Tau-Bench airline benchmark, our approach improves Qwen3.5-4B from 63.8 percent to 66.7 percent (+2.9pp) and Qwen3-30B-A3B from 58.0 percent to 69.5 percent (+11.5pp) -- with the trained 4B model exceeding GPT-4.1 (49.4 percent) and GPT-4o (42.8 percent) despite being 50 times smaller, and the 30.5B MoE model approaching Claude Sonnet 4.5 (70.0 percent). To our knowledge, these are the first published RL training results on Tau-Bench. We release our code, reward calibration analysis, and training recipes.", "authors": ["Wachiravit Modecrua", "Krittanon Kaewtawee", "Krittin Pachtrachai", "Touchapon Kraisingkorn"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02869", "pdf_url": "https://arxiv.org/pdf/2604.02869v1", "arxiv_id": "2604.02869", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "5b8c3d99a667601f0429683b95d8b35e9de575258db1e9e343fb99545deeb976", "sources": ["arxiv", "semantic_scholar"], "title": "CoFi-PGMA: Counterfactual Policy Gradients under Filtered Feedback for Multi-Agent LLMs", "abstract": "Large language model (LLM) deployments increasingly rely on multi-agent architectures in which multiple models either compete through routing mechanisms or collaborate to produce a final answer. In both settings, the learning signal received by each agent is filtered by the system mechanism. Routing produces selection-gated feedback where only the chosen response is evaluated, while collaboration produces shared rewards that obscure the individual contribution of each agent. As a result, standard RLHF objectives designed for a single deployed policy become misspecified. We introduce CoFi-PGMA (Counterfactual Policy Gradients under Filtered Feedback for Multi-Agent LLMs), a unified framework for learning under filtered feedback in multi-agent LLM systems. Our approach derives a counterfactual per-agent training objective based on marginal contribution, which corrects the learning signal under both routing and collaborative mechanisms. For routing systems, the objective corresponds to off-policy corrections for selection-gated feedback, while for collaborative systems it reduces to leave-one-out difference rewards for credit assignment. We further analyze how softmax routing induces risk-sensitive incentives and provide practical training algorithms that integrate counterfactual estimators, multiturn-aware rewards, and policy optimization methods, and demonstrate the approach on a real-world reasoning dataset.", "authors": ["Stela Tong", "Elai Ben-Gal"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.22785", "pdf_url": "https://arxiv.org/pdf/2604.22785v1", "arxiv_id": "2604.22785", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "e87fd5b9ee7313dab60e092614e716925c47c8ae99676aa446a5c9da17f1368e", "sources": ["arxiv", "semantic_scholar"], "title": "Do Agent Societies Develop Intellectual Elites? The Hidden Power Laws of Collective Cognition in LLM Multi-Agent Systems", "abstract": "Large Language Model (LLM) multi-agent systems are increasingly deployed as interacting agent societies, yet scaling these systems often yields diminishing or unstable returns, the causes of which remain poorly understood. We present the first large-scale empirical study of coordination dynamics in LLM-based multi-agent systems, introducing an atomic event-level formulation that reconstructs reasoning as cascades of coordination. Analyzing over 1.5 Million interactions across tasks, topologies, and scales, we uncover three coupled laws: coordination follows heavy-tailed cascades, concentrates via preferential attachment into intellectual elites, and produces increasingly frequent extreme events as system size grows. We show that these effects are coupled through a single structural mechanism: an integration bottleneck, in which coordination expansion scales with system size while consolidation does not, producing large but weakly integrated reasoning processes. To test this mechanism, we introduce Deficit-Triggered Integration (DTI), which selectively increases integration under imbalance. DTI improves performance precisely where coordination fails, without suppressing large-scale reasoning. Together, our results establish quantitative laws of collective cognition and identify coordination structure as a fundamental, previously unmeasured axis for understanding and improving scalable multi-agent intelligence.", "authors": ["Kavana Venkatesh", "Jiaming Cui"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02674", "pdf_url": "https://arxiv.org/pdf/2604.02674v1", "arxiv_id": "2604.02674", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "9020c7522ae8c01d8dc51cbdcfbab6bafee3829a64c804308a3a3be0fb40638f", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond the AI Tutor: Social Learning with LLM Agents", "abstract": "Most AI-based educational tools today adopt a one-on-one tutoring paradigm, pairing a single LLM with a single learner. Yet decades of learning science research suggest that multi-party interaction -- through peer modeling, co-construction, and exposure to diverse perspectives -- can produce learning benefits that dyadic tutoring alone cannot. In this paper, we investigate whether multi-agent LLM configurations can enhance learning outcomes beyond what a single LLM tutor provides. We present two controlled experiments spanning distinct learning contexts. In a convergent problem-solving study ($N=315$), participants tackle SAT-level math problems in a 2$\\times$2 design that varies the presence of an LLM tutor and LLM peers, each making different kinds of errors (conceptual vs.\\ arithmetic); participants who interacted with both a tutor and peers achieved the highest unassisted test accuracy. In a divergent composition study ($N=247$), participants write argumentative and creative essays with either no AI assistance, a single LLM (Claude or ChatGPT), or both Claude and ChatGPT together; while both LLM conditions improved essay quality, only the two-agent condition avoided the idea-level homogeneity that single-model assistance was found to produce. Together, these studies offer one of the first controlled investigations of multi-agent LLM learning environments, probing whether the move from one-on-one AI tutoring toward richer agent configurations can unlock the collaborative and observational benefits long documented in human social learning research.", "authors": ["Harsh Kumar", "Zi Kang", " Mu", "Jonathan Vincentius", "Ashton Anderson"], "categories": ["cs.HC", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02677", "pdf_url": "https://arxiv.org/pdf/2604.02677v1", "arxiv_id": "2604.02677", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "81a7ef2d01556bb93d56209a7ddce80d213d4e02df4c4abda52f177fd4ab03fc", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling Multi-agent Systems: A Smart Middleware for Improving Agent Interactions", "abstract": "As Large Language Model (LLM) based Multi-Agent Systems (MAS) evolve from experimental pilots to complex, persistent ecosystems, the limitations of direct agent-to-agent communication have become increasingly apparent. Current architectures suffer from fragmented context, stochastic hallucinations, rigid security boundaries, and inefficient topology management. This paper introduces Cognitive Fabric Nodes (CFN), a novel middleware layer that creates an omnipresent \"Cognitive Fabric\" between agents. Unlike traditional message queues or service meshes, CFNs are not merely pass-through mechanisms; they are active, intelligent intermediaries. Central to this architecture is the elevation of Memory from simple storage to an active functional substrate that informs four other critical capabilities: Topology Selection, Semantic Grounding, Security Policy Enforcement, and Prompt Transformation. We propose that each of these functions be governed by learning modules utilizing Reinforcement Learning (RL) and optimization algorithms to improve system performance dynamically. By intercepting, analyzing, and rewriting inter-agent communication, the Cognitive Fabric ensures that individual agents remain lightweight while the ecosystem achieves coherence, safety, and semantic alignment. We evaluate the effectiveness of the CFN on the HotPotQA and MuSiQue datasets in a multi-agent environment and demonstrate that the CFN improves performance by more than 10\\% on both datasets over direct agent to agent communication.", "authors": ["Charles Fleming", "Guillaume De Saint Marc", "Ramana Kompella", "Peter Bosch", "Vijoy Pandey"], "categories": ["cs.MA", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.03430", "pdf_url": "https://arxiv.org/pdf/2604.03430v2", "arxiv_id": "2604.03430", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "51f37a34855c1280212449c2c0c9cd301aa8a0bd69d3d687e063e70c09c733d0", "sources": ["arxiv", "semantic_scholar"], "title": "ToolWeave: Structured Synthesis of Complex Multi-Turn Tool-Calling Dialogues", "abstract": "Multi-turn tool calling is essential for LLMs to function as autonomous agents, yet synthesizing the training data required for these capabilities remains a fundamental challenge. Existing synthetic data generation pipelines often produce unrealistic dialogues for two reasons: they chain tools that are only superficially compatible rather than aligned with meaningful user tasks, and they generate dialogues in one shot, which often introduces arguments that were neither provided by the user nor produced by prior tool calls. These issues also lead to a severe underrepresentation of multi-step tool interactions. We introduce ToolWeave, a structured framework for synthesizing realistic multi-turn tool-calling dialogues. ToolWeave support realistic multi-step workflows (or tool sequences) by constructing tools with built-in dependencies and filters the workflows based on alignment with user goals. It reduces parameter hallucination by using a fine-grained planning stage that explicitly tracks parameter provenance. As a result, ToolWeave-generated synthetic dialogues contain more multi-step tool interactions (45%) and fewer hallucinations in parameters and tool names. Consequently, LLMs fine-tuned on ToolWeave consistently outperform those fine-tuned on prior datasets across three public benchmarks. Notably, Llama-3.1-70B fine-tuned on ToolWeave achieves 39.75% on BFCL-V3 multi-turn, compared to 23.50% when fine-tuned on SOTA ToolFlow data.", "authors": ["Dinesh Khandelwal", "Gnana Prakash Punnavajhala", "GPS Bhargav", "Gaurav Pandey", "Sachin Joshi", "Hima Karanam", "Dinesh Raghu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2605.12521", "pdf_url": "https://arxiv.org/pdf/2605.12521v1", "arxiv_id": "2605.12521", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "910adb7abaebfa3718d2e3b741bf4c3747dce8c460da00a598cb4e60e5f4acca", "sources": ["arxiv", "semantic_scholar"], "title": "TokenDance: Scaling Multi-Agent LLM Serving via Collective KV Cache Sharing", "abstract": "Multi-agent LLM applications organize execution in synchronized rounds where a central scheduler gathers outputs from all agents and redistributes the combined context. This All-Gather communication pattern creates massive KV Cache redundancy, because every agent's prompt contains the same shared output blocks, yet existing reuse methods fail to exploit it efficiently. We present TokenDance, a system that scales the number of concurrent agents by exploiting the All-Gather pattern for collective KV Cache sharing. TokenDance's KV Collector performs KV Cache reuse over the full round in one collective step, so the cost of reusing a shared block is paid once regardless of agent count. Its Diff-Aware Storage encodes sibling caches as block-sparse diffs against a single master copy, achieving 11-17x compression on representative workloads. Evaluation on GenerativeAgents and AgentSociety shows that TokenDance supports up to 2.7x more concurrent agents than vLLM with prefix caching under SLO requirement, reduces per-agent KV Cache storage by up to 17.5x, and achieves up to 1.9x prefill speedup over per-request position-independent caching.", "authors": ["Zhuohang Bian", "Feiyang Wu", "Chengrui Zhang", "Hangcheng Dong", "Yun Liang", "Youwei Zhuo"], "categories": ["cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.03143", "pdf_url": "https://arxiv.org/pdf/2604.03143v1", "arxiv_id": "2604.03143", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "565d3b373ee324038e05e91372b2c50dd7977ce0670b30f39f611c018b14057a", "sources": ["arxiv", "semantic_scholar"], "title": "MatClaw: An Autonomous Code-First LLM Agent for End-to-End Materials Exploration", "abstract": "Existing LLM agents for computational materials science are constrained by pipeline-bounded architectures tied to specific simulation codes and by dependence on manually written tool functions that grow with task scope. We present MatClaw, a code-first agent that writes and executes Python directly, composing any installed domain library to orchestrate multi-code workflows on remote HPC clusters without predefined tool functions. To sustain coherent execution across multi-day workflows, MatClaw uses a four-layer memory architecture that prevents progressive context loss, and retrieval-augmented generation over domain source code that raises per-step API-call accuracy to ${\\sim}$99 %. Three end-to-end demonstrations on ferroelectric CuInP2S6 (machine-learning force field training via active learning, Curie temperature prediction, and heuristic parameter-space search) reveal that the agent handles code generation reliably but struggles with tacit domain knowledge. The missing knowledge, such as appropriate simulation timescales, equilibration protocols, and sampling strategies, is the kind that researchers accumulate through experience but rarely formalize. Two lightweight interventions, literature self-learning and expert-specified constraints, bridge these gaps, defining a guided autonomy model in which the researcher provides high-level domain knowledge while the agent handles workflow execution. Our results demonstrate that the gap between guided and fully autonomous computational materials research is narrower than ever before: LLMs already handle code generation and scientific interpretation reliably, and the rapid improvement in their capabilities will accelerate materials discovery beyond what manual workflows can achieve. All code and benchmarks are open-source.", "authors": ["Chenmu Zhang", "Boris I. Yakobson"], "categories": ["cond-mat.mtrl-sci", "cs.SE"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02688", "pdf_url": "https://arxiv.org/pdf/2604.02688v3", "arxiv_id": "2604.02688", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.631}
{"id": "44ee087ed4c382cc6a5acfbeae289c74198d32b4e00da3828a2dd4c1e5d2954e", "sources": ["arxiv", "semantic_scholar"], "title": "EMS: Multi-Agent Voting via Efficient Majority-then-Stopping", "abstract": "Majority voting is the standard for aggregating multi-agent responses into a final decision. However, traditional methods typically require all agents to complete their reasoning before aggregation begins, leading to significant computational overhead, as many responses become redundant once a majority consensus is achieved. In this work, we formulate the multi-agent voting as a reliability-aware agent scheduling problem, and propose an Efficient Majority-then-Stopping (EMS) to improve reasoning efficiency. EMS prioritizes agents based on task-aware reliability and terminates the reasoning pipeline the moment a majority is achieved from the following three critical components. Specifically, we introduce Agent Confidence Modeling (ACM) to estimate agent reliability using historical performance and semantic similarity, Adaptive Incremental Voting (AIV) to sequentially select agents with early stopping, and Individual Confidence Updating (ICU) to dynamically update the reliability of each contributing agent. Extensive evaluations across six benchmarks demonstrate that EMS consistently reduces the average number of invoked agents by 32%.", "authors": ["Yiqing Liu", "Hantao Yao", "Wu Liu", "Yongdong Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02863", "pdf_url": "https://arxiv.org/pdf/2604.02863v1", "arxiv_id": "2604.02863", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "d98f95da417ade4500799b8d956e8efd81c74e6fedcbc81b0bc89e7e2f33e16a", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Optimizing Multi-Agent Systems for Deep Research", "abstract": "Given a user's complex information need, a multi-agent Deep Research system iteratively plans, retrieves, and synthesizes evidence across hundreds of documents to produce a high-quality answer. In one possible architecture, an orchestrator agent coordinates the process, while parallel worker agents execute tasks. Current Deep Research systems, however, often rely on hand-engineered prompts and static architectures, making improvement brittle, expensive, and time-consuming. We therefore explore various multi-agent optimization methods to show that enabling agents to self-play and explore different prompt combinations can produce high-quality Deep Research systems that match or outperform expert-crafted prompts.", "authors": ["Arthur Câmara", "Vincent Slot", "Jakub Zavrel"], "categories": ["cs.IR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02988", "pdf_url": "https://arxiv.org/pdf/2604.02988v1", "arxiv_id": "2604.02988", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3398}
{"id": "80c3240d8f5e3674f28e08ace7b9c7ad232538548d431b5012711b8a980d2e38", "sources": ["arxiv", "semantic_scholar"], "title": "Council Mode: A Heterogeneous Multi-Agent Consensus Framework for Reducing LLM Hallucination and Bias", "abstract": "Large Language Models (LLMs) have demonstrated advanced capabilities but often suffer from factual inaccuracies (hallucinations) and systematic biases. These issues, sometimes amplified in specific architectures like Mixture-of-Experts (MoE) which motivate our work, pose risks for reliable deployment. To address these challenges, we propose the Council Mode, a multi-agent consensus framework. Our approach dispatches queries to multiple heterogeneous frontier LLMs in parallel and synthesizes their outputs using a dedicated consensus model. The pipeline consists of three phases: an intelligent triage for query complexity, parallel generation across diverse models, and a structured synthesis that identifies agreement, disagreement, and unique findings. In our evaluation, conducted under controlled no-web settings, the Council Mode achieved a 35.9% relative reduction in hallucination rates on a 1,200-sample HaluEval subset and a 7.8-point improvement on TruthfulQA compared to the top-performing individual model. On our curated MDR-500 multi-domain reasoning benchmark, the Council Mode achieved a Quality Score of 91.7%, representing a 10.2-point improvement over the best individual model. The framework also exhibited lower measured bias variance under our rubric-based evaluation protocol. We provide a cost-effectiveness analysis showing that the framework incurs a 4.2x token-cost overhead, making it most suitable for accuracy-prioritized applications where the cost of errors exceeds the added inference cost. These findings suggest that structured multi-agent consensus is a promising direction for enhancing the reliability and factual grounding of LLM-generated content.", "authors": ["Shuai Wu", "Xue Li", "Yanna Feng", "Yufang Li", "Zhijun Wang", "Ran Wang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-03", "url": "https://arxiv.org/abs/2604.02923", "pdf_url": "https://arxiv.org/pdf/2604.02923v3", "arxiv_id": "2604.02923", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Noah-Wu66/Vectaix-Research", "venue": null, "quality_score": 0.631}
{"id": "6e9861987b47774f770ba1c703a71e99eef6aed83bbc9f8f651642e4bdcc6482", "sources": ["arxiv", "semantic_scholar"], "title": "Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token Budgets", "abstract": "Recent work reports strong performance from multi-agent LLM systems (MAS), but these gains are often confounded by increased test-time computation. When computation is normalized, single-agent systems (SAS) can match or outperform MAS, yet the theoretical basis and evaluation methodology behind this comparison remain unclear. We present an information-theoretic argument, grounded in the Data Processing Inequality, suggesting that under a fixed reasoning-token budget and with perfect context utilization, single-agent systems are more information-efficient. This perspective further predicts that multi-agent systems become competitive when a single agent's effective context utilization is degraded, or when more compute is expended. We test these predictions in a controlled empirical study across three model families (Qwen3, DeepSeek-R1-Distill-Llama, and Gemini 2.5), comparing SAS with multiple MAS architectures under matched budgets. We find that SAS consistently match or outperform MAS on multi-hop reasoning tasks when reasoning tokens are held constant. Beyond aggregate performance, we conduct a detailed diagnostic analysis of system behavior and evaluation methodology. We identify significant artifacts in API-based budget control (particularly in Gemini 2.5) and in standard benchmarks, both of which can inflate apparent gains from MAS. Overall, our results suggest that, for multi-hop reasoning tasks, many reported advantages of multi-agent systems are better explained by unaccounted computation and context effects rather than inherent architectural benefits, and highlight the importance of understanding and explicitly controlling the trade-offs between compute, context, and coordination in agentic systems.", "authors": ["Dat Tran", "Douwe Kiela"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-02", "url": "https://arxiv.org/abs/2604.02460", "pdf_url": "https://arxiv.org/pdf/2604.02460v2", "arxiv_id": "2604.02460", "doi": null, "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3391}
{"id": "b34dfaddb0e153a3c287701c65f834428d2c02318ea32786a50f5a9d0762a6fe", "sources": ["arxiv", "semantic_scholar"], "title": "CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery", "abstract": "Large language model (LLM)-based evolution is a promising approach for open-ended discovery, where progress requires sustained search and knowledge accumulation. Existing methods still rely heavily on fixed heuristics and hard-coded exploration rules, which limit the autonomy of LLM agents. We present CORAL, the first framework for autonomous multi-agent evolution on open-ended problems. CORAL replaces rigid control with long-running agents that explore, reflect, and collaborate through shared persistent memory, asynchronous multi-agent execution, and heartbeat-based interventions. It also provides practical safeguards, including isolated workspaces, evaluator separation, resource management, and agent session and health management. Evaluated on diverse mathematical, algorithmic, and systems optimization tasks, CORAL sets new state-of-the-art results on 10 tasks, achieving 3-10 times higher improvement rates with far fewer evaluations than fixed evolutionary search baselines across tasks. On Anthropic's kernel engineering task, four co-evolving agents improve the best known score from 1363 to 1103 cycles. Mechanistic analyses further show how these gains arise from knowledge reuse and multi-agent exploration and communication. Together, these results suggest that greater agent autonomy and multi-agent evolution can substantially improve open-ended discovery. Code is available at https://github.com/Human-Agent-Society/CORAL.", "authors": ["Ao Qu", "Han Zheng", "Zijian Zhou", "Yihao Yan", "Yihong Tang", "Shao Yong Ong", "Fenglu Hong", "Kaichen Zhou", "Chonghe Jiang", "Minwei Kong", "Jiacheng Zhu", "Xuan Jiang", "Sirui Li", "Cathy Wu", "Bryan Kian Hsiang Low", "Jinhua Zhao", "Paul Pu Liang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-02", "url": "https://arxiv.org/abs/2604.01658", "pdf_url": "https://arxiv.org/pdf/2604.01658v2", "arxiv_id": "2604.01658", "doi": null, "citation_count": 17, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/Human-Agent-Society/CORAL", "venue": null, "quality_score": 0.6297}
{"id": "598ddbb126c0324983e4e378f6e97501c873fb8c99c5917718235d1e4fb4db93", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Q-Mix: Selecting the Right Action for LLM Multi-Agent Systems through Reinforcement Learning", "abstract": "Large Language Models (LLMs) have shown remarkable performance in completing various tasks. However, solving complex problems often requires the coordination of multiple agents, raising a fundamental question: how to effectively select and interconnect these agents. In this paper, we propose \\textbf{Agent Q-Mix}, a reinforcement learning framework that reformulates topology selection as a cooperative Multi-Agent Reinforcement Learning (MARL) problem. Our method learns decentralized communication decisions using QMIX value factorization, where each agent selects from a set of communication actions that jointly induce a round-wise communication graph. At its core, Agent Q-Mix combines a topology-aware GNN encoder, GRU memory, and per-agent Q-heads under a Centralized Training with Decentralized Execution (CTDE) paradigm. The framework optimizes a reward function that balances task accuracy with token cost. Across seven core benchmarks in coding, reasoning, and mathematics, Agent Q-Mix achieves the highest average accuracy compared to existing methods while demonstrating superior token efficiency and robustness against agent failure. Notably, on the challenging Humanity's Last Exam (HLE) using Gemini-3.1-Flash-Lite as a backbone, Agent Q-Mix achieves 20.8\\% accuracy, outperforming Microsoft Agent Framework (19.2\\%) and LangGraph (19.2\\%), followed by AutoGen and Lobster by OpenClaw. These results underscore the effectiveness of learned, decentralized topology optimization in pushing the boundaries of multi-agent reasoning.", "authors": ["Eric Hanchen Jiang", "Levina Li", "Rui Sun", "Xiao Liang", "Yubei Li", "Yuchen Wu", "Haozheng Luo", "Hengli Li", "Zhi Zhang", "Zhaolu Kang", "Kai-Wei Chang", "Ying Nian Wu"], "categories": ["cs.CL", "stat.AP"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.00344", "pdf_url": "https://arxiv.org/pdf/2604.00344v1", "arxiv_id": "2604.00344", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3383}
{"id": "861077f4765fbb5c5de379bf77cc7ba2b9fb39f671ce044d43cc7aa39d5c1a72", "sources": ["arxiv", "semantic_scholar"], "title": "Competition and Cooperation of LLM Agents in Games", "abstract": "Large language model (LLM) agents are increasingly deployed in competitive multi-agent settings, raising fundamental questions about whether they converge to equilibria and how their strategic behavior can be characterized. In this paper, we study LLM agent interactions in two standard games: a network resource allocation game and a Cournot competition game. Rather than converging to Nash equilibria, we find that LLM agents tend to cooperate when given multi-round prompts and non-zero-sum context. Chain-of-thought analysis reveals that fairness reasoning is central to this behavior. We propose an analytical framework that captures the dynamics of LLM agent reasoning across rounds and explains these experimental findings.", "authors": ["Jiayi Yao", "Cong Chen", "Baosen Zhang"], "categories": ["cs.MA", "cs.GT", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.00487", "pdf_url": "https://arxiv.org/pdf/2604.00487v2", "arxiv_id": "2604.00487", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3383}
{"id": "2b20cbab0b78eeb6c6c3fd74d6e7ad4db204a5db72f7f9f44701a2df36b807e6", "sources": ["arxiv", "semantic_scholar"], "title": "Detecting Multi-Agent Collusion Through Multi-Agent Interpretability", "abstract": "As LLM agents are increasingly deployed in multi-agent systems, they introduce risks of covert coordination that may evade standard forms of human oversight. While linear probes on model activations have shown promise for detecting deception in single-agent settings, collusion is inherently a multi-agent phenomenon, and the use of internal representations for detecting collusion between agents remains unexplored. We introduce NARCBench, a benchmark for evaluating collusion detection under environment distribution shift, and propose five probing techniques that aggregate per-agent deception scores to classify scenarios at the group level, evaluated across four open-weight models (Qwen3-32B, Llama-3.1-70B, DeepSeek-R1 32B, GPT-OSS-20B) and six probe architectures. We frame this as a distributed anomaly detection problem, identifying three collusion signatures that map onto distinct anomaly types and detection paradigms. Every model reaches 1.00 AUROC in-distribution; on our strongest model (Llama-3.1-70B), our five probing techniques achieve 0.73 to 0.93 AUROC when transferred zero-shot to structurally different multi-agent scenarios and 0.99 to 1.00 on a steganographic blackjack card-counting task, with detection performance scaling with model capability. We find that no single probing technique dominates across all collusion types, consistent with the framework's prediction that different anomaly types require different detection paradigms. This work takes a step toward multi-agent interpretability: extending white-box inspection from single models to multi-agent contexts, where detection requires aggregating signals across agents. These results suggest that model internals provide a complementary signal to text-level monitoring for detecting multi-agent collusion. Code and data available at https://github.com/aaronrose227/narcbench.", "authors": ["Aaron Rose", "Carissa Cullen", "Sahar Abdelnabi", "Philip Torr", "Brandon Gary Kaplowitz", "Christian Schroeder de Witt"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.01151", "pdf_url": "https://arxiv.org/pdf/2604.01151v2", "arxiv_id": "2604.01151", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/aaronrose227/narcbench", "venue": null, "quality_score": 0.6283}
{"id": "ac2d41fcbe1e14dce64c3f8a95ab40187c4d3cec1d4144a106212b596e48b2db", "sources": ["arxiv", "semantic_scholar"], "title": "Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Prompts", "abstract": "Multi-agent Retrieval-Augmented Generation (RAG), wherein each agent takes on a specific role, supports hard queries that require multiple steps and sources, or complex reasoning. Existing approaches, however, rely on static agent behaviors and fixed orchestration strategies, leading to brittle performance on diverse, multi-hop tasks. We identify two key limitations: the lack of continuously adaptive orchestration mechanisms and the absence of behavior-level learning for individual agents. To this end, we propose HERA, a hierarchical framework that jointly evolves multi-agent orchestration and role-specific agent prompts. At the global level, HERA optimizes query-specific agent topologies through reward-guided sampling and experience accumulation. At the local level, Role-Aware Prompt Evolution refines agent behaviors via credit assignment and dual-axes adaptation along operational and behavioral principles, enabling targeted, role-conditioned improvements. On six knowledge-intensive benchmarks, HERA achieves an average improvement of 38.69\\% over recent baselines while maintaining robust generalization and token efficiency. Topological analyses reveal emergent self-organization, where sparse exploration yields compact, high-utility multi-agent networks, demonstrating both efficient coordination and robust reasoning.", "authors": ["Sha Li", "Naren Ramakrishnan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.00901", "pdf_url": "https://arxiv.org/pdf/2604.00901v2", "arxiv_id": "2604.00901", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3383}
{"id": "06da90ec147562f6d422ff237db4a09446a14ef10f6d643bb8bd4bb5e7a79ca8", "sources": ["arxiv", "semantic_scholar"], "title": "Type-Checked Compliance: Deterministic Guardrails for Agentic Financial Systems Using Lean 4 Theorem Proving", "abstract": "The rapid evolution of autonomous, agentic artificial intelligence within financial services has introduced an existential architectural crisis: large language models (LLMs) are probabilistic, non-deterministic systems operating in domains that demand absolute, mathematically verifiable compliance guarantees. Existing guardrail solutions -- including NVIDIA NeMo Guardrails and Guardrails AI -- rely on probabilistic classifiers and syntactic validators that are fundamentally inadequate for enforcing complex multi-variable regulatory constraints mandated by the SEC, FINRA, and OCC. This paper presents the Lean-Agent Protocol, a formal-verification-based AI guardrail platform that leverages the Aristotle neural-symbolic model developed by Harmonic AI to auto-formalize institutional policies into Lean 4 code. Every proposed agentic action is treated as a mathematical conjecture: execution is permitted if and only if the Lean 4 kernel proves that the action satisfies pre-compiled regulatory axioms. This architecture provides cryptographic-level compliance certainty at microsecond latency, directly satisfying SEC Rule 15c3-5, OCC Bulletin 2011-12, FINRA Rule 3110, and CFPB explainability mandates. A three-phase implementation roadmap from shadow verification through enterprise-scale deployment is provided.", "authors": ["Devakh Rashie", "Veda Rashi"], "categories": ["cs.LO", "cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.01483", "pdf_url": "https://arxiv.org/pdf/2604.01483v1", "arxiv_id": "2604.01483", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/arkanemystic/lean-agent-protocol", "venue": null, "quality_score": 0.6283}
{"id": "0474525373daf38f2d01ebd6dc17d53d41b3c30a2aab0fa6432c90eb055669b0", "sources": ["arxiv", "semantic_scholar"], "title": "HabitatAgent: An End-to-End Multi-Agent System for Housing Consultation", "abstract": "Housing selection is a high-stakes and largely irreversible decision problem. We study housing consultation as a decision-support interface for housing selection. Existing housing platforms and many LLM-based assistants often reduce this process to ranking or recommendation, resulting in opaque reasoning, brittle multi-constraint handling, and limited guarantees on factuality. We present HabitatAgent, the first LLM-powered multi-agent architecture for end-to-end housing consultation. HabitatAgent comprises four specialized agent roles: Memory, Retrieval, Generation, and Validation. The Memory Agent maintains multi-layer user memory through internal stages for constraint extraction, memory fusion, and verification-gated updates; the Retrieval Agent performs hybrid vector--graph retrieval (GraphRAG); the Generation Agent produces evidence-referenced recommendations and explanations; and the Validation Agent applies multi-tier verification and targeted remediation. Together, these agents provide an auditable and reliable workflow for end-to-end housing consultation. We evaluate HabitatAgent on 100 real user consultation scenarios (300 multi-turn question--answer pairs) under an end-to-end correctness protocol. A strong single-stage baseline (Dense+Rerank) achieves 75% accuracy, while HabitatAgent reaches 95%.", "authors": ["Hongyang Yang", "Yanxin Zhang", "Yang She", "Yue Xiao", "Hao Wu", "Yiyang Zhang", "Jiapeng Hou", "Rongshan Zhang"], "categories": ["cs.LG", "cs.AI", "cs.ET", "q-fin.CP", "q-fin.RM"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2026-04-01", "url": "https://arxiv.org/abs/2604.00556", "pdf_url": "https://arxiv.org/pdf/2604.00556v1", "arxiv_id": "2604.00556", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3383}
{"id": "a74baeee34ead9ce95c69cc6426792ba404f93abf8875da1d7242131cb0847c2", "sources": ["arxiv", "semantic_scholar"], "title": "A Safety-Aware Role-Orchestrated Multi-Agent LLM Framework for Behavioral Health Communication Simulation", "abstract": "Single-agent large language model (LLM) systems struggle to simultaneously support diverse conversational functions and maintain safety in behavioral health communication. We propose a safety-aware, role-orchestrated multi-agent LLM framework designed to simulate supportive behavioral health dialogue through coordinated, role-differentiated agents. Conversational responsibilities are decomposed across specialized agents, including empathy-focused, action-oriented, and supervisory roles, while a prompt-based controller dynamically activates relevant agents and enforces continuous safety auditing. Using semi-structured interview transcripts from the DAIC-WOZ corpus, we evaluate the framework with scalable proxy metrics capturing structural quality, functional diversity, and computational characteristics. Results illustrate clear role differentiation, coherent inter-agent coordination, and predictable trade-offs between modular orchestration, safety oversight, and response latency when compared to a single-agent baseline. This work emphasizes system design, interpretability, and safety, positioning the framework as a simulation and analysis tool for behavioral health informatics and decision-support research rather than a clinical intervention.", "authors": ["Ha Na Cho"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-31", "url": "https://arxiv.org/abs/2604.00249", "pdf_url": "https://arxiv.org/pdf/2604.00249v1", "arxiv_id": "2604.00249", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3376}
{"id": "db946485074a8c30e5c63be0ac14377d5081557cd6dc1eeed7b5b82e5a574ae8", "sources": ["arxiv", "semantic_scholar"], "title": "An Empirical Study of Multi-Agent Collaboration for Automated Research", "abstract": "As AI agents evolve, the community is rapidly shifting from single Large Language Models (LLMs) to Multi-Agent Systems (MAS) to overcome cognitive bottlenecks in automated research. However, the optimal multi-agent coordination framework for these autonomous agents remains largely unexplored. In this paper, we present a systematic empirical study investigating the comparative efficacy of distinct multi-agent structures for automated machine learning optimization. Utilizing a rigorously controlled, execution-based testbed equipped with Git worktree isolation and explicit global memory, we benchmark a single-agent baseline against two multi-agent paradigms: a subagent architecture (parallel exploration with post-hoc consolidation) and an agent team architecture (experts with pre-execution handoffs). By evaluating these systems under strictly fixed computational time budgets, our findings reveal a fundamental trade-off between operational stability and theoretical deliberation. The subagent mode functions as a highly resilient, high-throughput search engine optimal for broad, shallow optimizations under strict time constraints. Conversely, the agent team topology exhibits higher operational fragility due to multi-author code generation but achieves the deep theoretical alignment necessary for complex architectural refactoring given extended compute budgets. These empirical insights provide actionable guidelines for designing future autoresearch systems, advocating for dynamically routed architectures that adapt their collaborative structures to real-time task complexity.", "authors": ["Yang Shen", "Zhenyi Yi", "Ziyi Zhao", "Lijun Sun", "Dongyang Li", "Chin-Teng Lin", "Yuhui Shi"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-31", "url": "https://arxiv.org/abs/2603.29632", "pdf_url": "https://arxiv.org/pdf/2603.29632v2", "arxiv_id": "2603.29632", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3376}
{"id": "cda0189175766265495f085d2810d4aa77518edd1f8e91843c0b3aee3b4abe77", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLMs for Adaptive Acquisition in Bayesian Optimization", "abstract": "The exploration-exploitation trade-off is central to sequential decision-making and black-box optimization, yet how Large Language Models (LLMs) reason about and manage this trade-off remains poorly understood. Unlike Bayesian Optimization, where exploration and exploitation are explicitly encoded through acquisition functions, LLM-based optimization relies on implicit, prompt-based reasoning over historical evaluations, making search behavior difficult to analyze or control. In this work, we present a metric-level study of LLM-mediated search policy learning, studying how LLMs construct and adapt exploration-exploitation strategies under multiple operational definitions of exploration, including informativeness, diversity, and representativeness. We show that single-agent LLM approaches, which jointly perform strategy selection and candidate generation within a single prompt, suffer from cognitive overload, leading to unstable search dynamics and premature convergence. To address this limitation, we propose a multi-agent framework that decomposes exploration-exploitation control into strategic policy mediation and tactical candidate generation. A strategy agent assigns interpretable weights to multiple search criteria, while a generation agent produces candidates conditioned on the resulting search policy defined as weights. This decomposition renders exploration-exploitation decisions explicit, observable, and adjustable. Empirical results across various continuous optimization benchmarks indicate that separating strategic control from candidate generation substantially improves the effectiveness of LLM-mediated search.", "authors": ["Andrea Carbonati", "Mohammadsina Almasi", "Hadis Anahideh"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-30", "url": "https://arxiv.org/abs/2603.28959", "pdf_url": "https://arxiv.org/pdf/2603.28959v1", "arxiv_id": "2603.28959", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3369}
{"id": "46614612ce4d301dae8eda68a5628552003dd0680aff7f498bcf315da112661b", "sources": ["arxiv", "semantic_scholar"], "title": "Synergy: A Next-Generation General-Purpose Agent for Open Agentic Web", "abstract": "AI agents are rapidly expanding in both capability and population: they now write code, operate computers across platforms, manage cloud infrastructure, and make purchasing decisions, while open-source frameworks such as OpenClaw are putting personal agents in the hands of millions and embodied agents are spreading across smartphones, vehicles, and robots. As the internet prepares to host billions of such entities, it is shifting toward what we call Open Agentic Web, a decentralized digital ecosystem in which agents from different users, organizations, and runtimes can discover one another, negotiate task boundaries, and delegate work across open technical and social surfaces at scale. Yet most of today's agents remain isolated tools or closed-ecosystem orchestrators rather than socially integrated participants in open networks. We argue that the next generation of agents must become Agentic Citizens, defined by three requirements: Agentic-Web-Native Collaboration, participation in open collaboration networks rather than only closed internal orchestration; Agent Identity and Personhood, continuity as a social entity rather than a resettable function call; and Lifelong Evolution, improvement across task performance, communication, and collaboration over time. We present Synergy, a general-purpose agent architecture and runtime harness for persistent, collaborative, and evolving agents on Open Agentic Web, grounding collaboration in session-native orchestration, repository-backed workspaces, and social communication; identity in typed memory, notes, agenda, skills, and persistent social relationships; and evolution in an experience-centered learning mechanism that proactively recalls rewarded trajectories at inference time.", "authors": ["Xiaohang Nie", "Zihan Guo", "Kezhuo Yang", "Zhichong Zheng", "Bochen Ge", "Shuai Pan", "Zeyi Chen", "Youling Xiang", "Yu Zhang", "Weiwen Liu", "Yuanjian Zhou", "Weinan Zhang"], "categories": ["cs.CY", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-30", "url": "https://arxiv.org/abs/2603.28428", "pdf_url": "https://arxiv.org/pdf/2603.28428v1", "arxiv_id": "2603.28428", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6256}
{"id": "5d77b7afd05b6578924b0d812c7fa7643892e2ee0f0d3c268c17897f1238f144", "sources": ["arxiv", "semantic_scholar"], "title": "InconLens: Interactive Visual Diagnosis of Behavioral Inconsistencies in LLM-based Agentic Systems", "abstract": "Large Language Model (LLM)-based agentic systems have shown growing promise in tackling complex, multi-step tasks through autonomous planning, reasoning, and interaction with external environments. However, the stochastic nature of LLM generation introduces intrinsic behavioral inconsistency: the same agent may succeed in one execution but fail in another under identical inputs. Diagnosing such inconsistencies remains a major challenge for developers, as agent execution logs are often lengthy, unstructured, and difficult to compare across runs. Existing debugging and evaluation tools primarily focus on inspecting single executions, offering limited support for understanding how and why agent behaviors diverge across repeated runs. To address this challenge, we introduce InconLens, a visual analytics system designed to support interactive diagnosis of LLM-based agentic systems with a particular focus on cross-run behavioral analysis. InconLens introduces information nodes as an intermediate abstraction that captures canonical informational milestones shared across executions, enabling semantic alignment and inspection of agent reasoning trajectories across multiple runs. We demonstrate the effectiveness of InconLens through a detailed case study and further validate its usability and analytical value via expert interviews. Our results show that InconLens enables developers to more efficiently identify divergence points, uncover latent failure modes, and gain actionable insights into improving the reliability and stability of agentic systems.", "authors": ["Shuo Yan", "Xiaolin Wen", "Shaolun Ruan", "Yanjie Zhang", "Jiaming Mi", "Yushi Sun", "Huamin Qu", "Rui Sheng"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-30", "url": "https://arxiv.org/abs/2603.28106", "pdf_url": "https://arxiv.org/pdf/2603.28106v1", "arxiv_id": "2603.28106", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3369}
{"id": "6b423be3c529093f454d04fd45dd93a32f3285fcf0fc9f0e3fb76f0f795d7d99", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Reliable Evaluation of LLM-Based Financial Multi-Agent Systems: Taxonomy, Coordination Primacy, and Cost Awareness", "abstract": "Multi-agent systems based on large language models (LLMs) for financial trading have grown rapidly since 2023, yet the field lacks a shared framework for understanding what drives performance or for evaluating claims credibly. This survey makes three contributions. First, we introduce a four-dimensional taxonomy, covering architecture pattern, coordination mechanism, memory architecture, and tool integration; applied to 12 multi-agent systems and two single-agent baselines. Second, we formulate the Coordination Primacy Hypothesis (CPH): inter-agent coordination protocol design is a primary driver of trading decision quality, often exerting greater influence than model scaling. CPH is presented as a falsifiable research hypothesis supported by tiered structural evidence rather than as an empirically validated conclusion; its definitive validation requires evaluation infrastructure that does not yet exist in the field. Third, we document five pervasive evaluation failures (look-ahead bias, survivorship bias, backtesting overfitting, transaction cost neglect, and regime-shift blindness) and show that these can reverse the sign of reported returns. Building on the CPH and the evaluation critique, we introduce the Coordination Breakeven Spread (CBS), a metric for determining whether multi-agent coordination adds genuine value net of transaction costs, and propose minimum evaluation standards as prerequisites for validating the CPH.", "authors": ["Phat Nguyen", "Thang Pham"], "categories": ["cs.MA", "cs.AI", "cs.CE"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-29", "url": "https://arxiv.org/abs/2603.27539", "pdf_url": "https://arxiv.org/pdf/2603.27539v1", "arxiv_id": "2603.27539", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3361}
{"id": "df784d03ed7c0cb4416001406ca9b2cee443b523740d58775cff20ac8dbf2f0f", "sources": ["arxiv", "semantic_scholar"], "title": "The Observability Gap: Why Output-Level Human Feedback Fails for LLM Coding Agents", "abstract": "Large language model (LLM) multi-agent coding systems typically fix agent capabilities at design time. We study an alternative setting, earned autonomy, in which a coding agent starts with zero pre-defined functions and incrementally builds a reusable function library through lightweight human feedback on visual output alone. We evaluate this setup in a Blender-based 3D scene generation task requiring both spatial reasoning and programmatic geometric control. Although the agent rediscovered core utility functions comparable to a human reference implementation, it achieved 0% full-scene success under output-only feedback across multiple instruction granularities, where success required satisfying object completeness, ground contact, collision avoidance, and scale plausibility simultaneously. Our analysis identifies a structural observability gap: bugs originate in code logic and execution state, while human evaluation occurs only at the output layer, and the many-to-one mapping from internal states to visible outcomes prevents symptom-level feedback from reliably identifying root causes. This mismatch leads to persistent failure mode oscillation rather than convergence. A diagnostic intervention that injected minimal code-level knowledge restored convergence, strongly supporting the interpretation that the main bottleneck lies in feedback observability rather than programming competence. We formalize this phenomenon as a feedback paradox in domains with deep causal chains between internal code logic and perceptual outcomes, and argue that effective human-agent collaboration in such settings requires intermediate observability beyond output-only evaluation.", "authors": ["Yinghao Wang", "Cheng Wang"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2603.26942", "pdf_url": "https://arxiv.org/pdf/2603.26942v1", "arxiv_id": "2603.26942", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3347}
{"id": "c227b738af11e0c8289500e7257937c10266f1125419ed3811997224df313005", "sources": ["arxiv", "semantic_scholar"], "title": "On the Reliability Limits of LLM-Based Multi-Agent Planning", "abstract": "This technical note studies the reliability limits of LLM-based multi-agent planning as a delegated decision problem. We model the LLM-based multi-agent architecture as a finite acyclic decision network in which multiple stages process shared model-context information, communicate through language interfaces with limited capacity, and may invoke human review. We show that, without new exogenous signals, any delegated network is decision-theoretically dominated by a centralized Bayes decision maker with access to the same information. In the common-evidence regime, this implies that optimizing over multi-agent directed acyclic graphs under a finite communication budget can be recast as choosing a budget-constrained stochastic experiment on the shared signal. We also characterize the loss induced by communication and information compression. Under proper scoring rules, the gap between the centralized Bayes value and the value after communication admits an expected posterior divergence representation, which reduces to conditional mutual information under logarithmic loss and to expected squared posterior error under the Brier score. These results characterize the fundamental reliability limits of delegated LLM planning. Experiments with LLMs on a controlled problem set further demonstrate these characterizations.", "authors": ["Ruicheng Ao", "Siyang Gao", "David Simchi-Levi"], "categories": ["cs.MA", "cs.LG", "math.OC", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2603.26993", "pdf_url": "https://arxiv.org/pdf/2603.26993v1", "arxiv_id": "2603.26993", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3347}
{"id": "2cf933057f10ef768f77a4ee7eb0bf9ba2e2bdbaae1a48a80b4d9ff154534fa8", "sources": ["arxiv", "semantic_scholar"], "title": "GISclaw: A Comprehensive Open-Source LLM Agent System for Realistic Multi-Step Geospatial Analysis", "abstract": "Most LLM-driven GIS assistants solve narrow single-step tasks tightly coupled to proprietary platforms such as ArcGIS or QGIS, limiting their use for the multi-step, cross-format pipelines that define professional geospatial analysis. We present GISclaw, a comprehensive open-source agent system that performs realistic GIS analysis end to end - spatial joins, raster algebra, kriging interpolation, machine-learning classification, network analysis, choropleth cartography - directly through Python with no commercial GIS dependency. GISclaw couples an LLM reasoning core with a persistent Python sandbox pre-loaded with the open-source geospatial stack, three engineered prompt rules (Schema Analysis, Package Constraint, Domain Knowledge Injection), and an Error-Memory module for self-correction. A single backend-agnostic architecture supports both cloud-API and locally deployed open-weight LLM backends, enabling air-gapped deployment without loss of capability. On GeoAnalystBench - 50 expert-curated multi-step tasks averaging 5.8 analytical steps across vector, raster, and tabular data - GISclaw reaches up to 100% task success and 97% mean success over three independent runs. We further conduct 1,800 controlled experiments (50 tasks x 6 backends x 2 architectures x 3 repeats) with bootstrap 95% CIs, paired Wilcoxon tests, and a composite-score sensitivity analysis (Kendall's tau median = 0.94), and introduce a three-layer evaluation protocol combining code structure, reasoning process, and type-specific output verification. The Single-Agent ReAct loop reliably outperforms the Dual-Agent Plan-Execute-Replan pipeline on every cloud backend (Cliff's delta = 0.15-0.41); only the locally deployed 14B model gains from multi-agent orchestration, suggesting architectural complexity should match model capability rather than be added by default.", "authors": ["Jinzhen Han", "JinByeong Lee", "Yuri Shim", "Jisung Kim", "Jae-Joon Lee"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2603.26845", "pdf_url": "https://arxiv.org/pdf/2603.26845v2", "arxiv_id": "2603.26845", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6216}
{"id": "657d24e6bb22a36e2034f932a26fd6494d52477965d0314b2b97d2879efc2e88", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling Teams or Scaling Time? Memory Enabled Lifelong Learning in LLM Multi-Agent Systems", "abstract": "Large language model (LLM) multi-agent systems can scale along two distinct dimensions: by increasing the number of agents and by improving through accumulated experience over time. Although prior work has studied these dimensions separately, their interaction under realistic cost constraints remains unclear. In this paper, we introduce a conceptual scaling view of multi-agent systems that jointly considers team size and lifelong learning ability, and we study how memory design shares this landscape. To this end, we propose \\textbf{LLMA-Mem}, a lifelong memory framework for LLM multi-agent systems under flexible memory topologies. We evaluate LLMA-Mem on \\textsc{MultiAgentBench} across coding, research, and database environments. Empirically, LLMA-Mem consistently improves long-horizon performance over baselines while reducing cost. Our analysis further reveals a non-monotonic scaling landscape: larger teams do not always produce better long-term performance, and smaller teams can outperform larger ones when memory better supports the reuse of experience. These findings position memory design as a practical path for scaling multi-agent systems more effectively and more efficiently over time.", "authors": ["Shanglin Wu", "Yuyang Luo", "Yueqing Liang", "Kaiwen Shi", "Yanfang Ye", "Ali Payani", "Kai Shu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2604.03295", "pdf_url": "https://arxiv.org/pdf/2604.03295v1", "arxiv_id": "2604.03295", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3347}
{"id": "737141747d5ffbe5797e335dfed9a33e72ef98bcf96c87e5c98b3c1c4482f765", "sources": ["arxiv", "semantic_scholar"], "title": "Deception and Communication in Autonomous Multi-Agent Systems: An Experimental Study with Among Us", "abstract": "As large language models are deployed as autonomous agents, their capacity for strategic deception raises core questions for coordination, reliability, and safety in multi-goal, multi-agent systems. We study deception and communication in L2LM agents through the social deduction game Among Us, a cooperative-competitive environment. Across 1,100 games, autonomous agents produced over one million tokens of meeting dialogue. Using speech act theory and interpersonal deception theory, we find that all agents rely mainly on directive language, while impostor agents shift slightly toward representative acts such as explanations and denials. Deception appears primarily as equivocation rather than outright lies, increasing under social pressure but rarely improving win rates. Our contributions are a large-scale analysis of role-conditioned deceptive behavior in LLM agents and empirical evidence that current agents favor low-risk ambiguity that is linguistically subtle yet strategically limited, revealing a fundamental tension between truthfulness and utility in autonomous communication.", "authors": ["Maria Milkowski", "Tim Weninger"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2603.26635", "pdf_url": "https://arxiv.org/pdf/2603.26635v1", "arxiv_id": "2603.26635", "doi": "10.65109/FRXL8789", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026), IFAAMAS, 2026", "quality_score": 0.5259}
{"id": "293c200c2cb4c9354d500f9751ff423ef32339e89db53897cf0b81b542daf8fa", "sources": ["arxiv", "semantic_scholar"], "title": "AgentCollab: A Self-Evaluation-Driven Collaboration Paradigm for Efficient LLM Agents", "abstract": "Autonomous agents powered by large language models (LLMs) perform complex tasks through long-horizon reasoning and tool interaction, where a fundamental trade-off arises between execution efficiency and reasoning robustness. Models at different capability-cost levels offer complementary advantages: lower-cost models enable fast execution but may struggle on difficult reasoning segments, while stronger models provide more robust reasoning at higher computational cost. We present AgentCollab, a self-driven collaborative inference framework that dynamically coordinates models with different reasoning capacities during agent execution. Instead of relying on external routing modules, the framework uses the agent's own self-reflection signal to determine whether the current reasoning trajectory is making meaningful progress, and escalates control to a stronger reasoning tier only when necessary. To further stabilize long-horizon execution, we introduce a difficulty-aware cumulative escalation strategy that allocates additional reasoning budget based on recent failure signals. In our experiments, we instantiate this framework using a two-level small-large model setting. Experiments on diverse multi-step agent benchmarks show that AgentCollab consistently improves the accuracy-efficiency Pareto frontier of LLM agents.", "authors": ["Wenbo Gao", "Renxi Liu", "Xian Wang", "Fang Guo", "Shuai Yang", "Xi Chen", "Hui-Ling Zhen", "Hanting Chen", "Weizhe Lin", "Xiaosong Li", "Yaoyuan Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-27", "url": "https://arxiv.org/abs/2603.26034", "pdf_url": "https://arxiv.org/pdf/2603.26034v1", "arxiv_id": "2603.26034", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3347}
{"id": "dd6c0db625bbf56eff3a3d846ec56e4d3177ecd019b45e063828c1f3d6b14568", "sources": ["arxiv", "semantic_scholar"], "title": "MemoryCD: Benchmarking Long-Context User Memory of LLM Agents for Lifelong Cross-Domain Personalization", "abstract": "Recent advancements in Large Language Models (LLMs) have expanded context windows to million-token scales, yet benchmarks for evaluating memory remain limited to short-session synthetic dialogues. We introduce \\textsc{MemoryCD}, the first large-scale, user-centric, cross-domain memory benchmark derived from lifelong real-world behaviors in the Amazon Review dataset. Unlike existing memory datasets that rely on scripted personas to generate synthetic user data, \\textsc{MemoryCD} tracks authentic user interactions across years and multiple domains. We construct a multi-faceted long-context memory evaluation pipeline of 14 state-of-the-art LLM base models with 6 memory method baselines on 4 distinct personalization tasks over 12 diverse domains to evaluate an agent's ability to simulate real user behaviors in both single and cross-domain settings. Our analysis reveals that existing memory methods are far from user satisfaction in various domains, offering the first testbed for cross-domain life-long personalization evaluation.", "authors": ["Weizhi Zhang", "Xiaokai Wei", "Wei-Chieh Huang", "Zheng Hui", "Chen Wang", "Michelle Gong", "Philip S. Yu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-26", "url": "https://arxiv.org/abs/2603.25973", "pdf_url": "https://arxiv.org/pdf/2603.25973v1", "arxiv_id": "2603.25973", "doi": null, "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.334}
{"id": "3f7e67509ce58ed03b04ec94942d7397acceaadf0c1dba6ebcef41f25466c002", "sources": ["arxiv", "semantic_scholar"], "title": "Rethinking Health Agents: From Siloed AI to Collaborative Decision Mediators", "abstract": "Large language model based health agents are increasingly used by health consumers and clinicians to interpret health information and guide health decisions. However, most AI systems in healthcare operate in siloed configurations, supporting individual users rather than the multi-stakeholder relationships central to healthcare. Such use can fragment understanding and exacerbate misalignment among patients, caregivers, and clinicians. We reframe AI not as a standalone assistant, but as a collaborator embedded within multi-party care interactions. Through a clinically validated fictional pediatric chronic kidney disease case study, we show that breakdowns in adherence stem from fragmented situational awareness and misaligned goals, and that siloed use of general-purpose AI tools does little to address these collaboration gaps. We propose a conceptual framework for designing AI collaborators that surface contextual information, reconcile mental models, and scaffold shared understanding while preserving human decision authority.", "authors": ["Ray-Yuan Chung", "Xuhai Xu", "Ari Pollack"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-26", "url": "https://arxiv.org/abs/2603.24986", "pdf_url": "https://arxiv.org/pdf/2603.24986v1", "arxiv_id": "2603.24986", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.334}
{"id": "c044309d79ad565b8f2cd351945ded5fff06dd8a6e2ac3ee4670c14c0cb824c4", "sources": ["arxiv", "semantic_scholar"], "title": "FinMCP-Bench: Benchmarking LLM Agents for Real-World Financial Tool Use under the Model Context Protocol", "abstract": "This paper introduces \\textbf{FinMCP-Bench}, a novel benchmark for evaluating large language models (LLMs) in solving real-world financial problems through tool invocation of financial model context protocols. FinMCP-Bench contains 613 samples spanning 10 main scenarios and 33 sub-scenarios, featuring both real and synthetic user queries to ensure diversity and authenticity. It incorporates 65 real financial MCPs and three types of samples, single tool, multi-tool, and multi-turn, allowing evaluation of models across different levels of task complexity. Using this benchmark, we systematically assess a range of mainstream LLMs and propose metrics that explicitly measure tool invocation accuracy and reasoning capabilities. FinMCP-Bench provides a standardized, practical, and challenging testbed for advancing research on financial LLM agents.", "authors": ["Jie Zhu", "Yimin Tian", "Boyang Li", "Kehao Wu", "Zhongzhi Liang", "Junhui Li", "Xianyin Zhang", "Lifan Guo", "Feng Chen", "Yong Liu", "Chi Zhang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-26", "url": "https://arxiv.org/abs/2603.24943", "pdf_url": "https://arxiv.org/pdf/2603.24943v1", "arxiv_id": "2603.24943", "doi": "10.1109/icassp55912.2026.11460653", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Acoustics, Speech, and Signal Processing", "quality_score": 0.5248}
{"id": "1c319129f62f0927b49c730fcd63bdc53e58068d57d6abfdabb62891022f5751", "sources": ["arxiv", "semantic_scholar"], "title": "When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMs", "abstract": "Multi-agent systems powered by large language models (LLMs) are increasingly deployed in settings that shape consequential decisions, both directly and indirectly. Yet it remains unclear whether their outcomes reflect collective reasoning, systematic bias, or mere chance. Recent work has sharpened this question with naming games, showing that even when no individual agent favors any label a priori, populations rapidly break symmetry and reach consensus. Here, we reveal the mechanism by introducing a minimal model, Quantized Simplex Gossip (QSG), and trace the microscopic origin of this agreement to mutual in-context learning. In QSG, agents maintain internal belief states but learn from one another's sampled outputs, so one agent's arbitrary choice becomes the next agent's evidence and can compound toward agreement. By analogy with neutral evolution, we call this sampling-driven regime memetic drift. QSG predicts a crossover from a drift-dominated regime, where consensus is effectively a lottery, to a selection regime, where weak biases are amplified and shape the outcome. We derive scaling laws for drift-induced polarization as a function of population size, communication bandwidth, in-context adaptation rate, and agents' internal uncertainty, and we validate them in both QSG simulations and naming-game experiments with LLM populations. Together, these results provide a framework for studying the collective mechanisms of social representation formation in multi-agent systems.", "authors": ["Hidenori Tanaka"], "categories": ["cs.AI", "cond-mat.dis-nn", "cond-mat.stat-mech", "physics.bio-ph", "physics.soc-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2026-03-25", "url": "https://arxiv.org/abs/2603.24676", "pdf_url": "https://arxiv.org/pdf/2603.24676v1", "arxiv_id": "2603.24676", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3332}
{"id": "a0061bbe26abc765e782c7b0809a6007419cb8d5dd3a5124f370aa6589d751c9", "sources": ["arxiv", "semantic_scholar"], "title": "MARCH: Multi-Agent Reinforced Self-Check for LLM Hallucination", "abstract": "Hallucination remains a critical bottleneck for large language models (LLMs), undermining their reliability in real-world applications, especially in Retrieval-Augmented Generation (RAG) systems. While existing hallucination detection methods employ LLM-as-a-judge to verify LLM outputs against retrieved evidence, they suffer from inherent confirmation bias, where the verifier inadvertently reproduces the errors of the original generation. To address this, we introduce Multi-Agent Reinforced Self-Check for Hallucination (MARCH), a framework that enforces rigorous factual alignment by leveraging deliberate information asymmetry. MARCH orchestrates a collaborative pipeline of three specialized agents: a Solver, a Proposer, and a Checker. The Solver generates an initial RAG response, which the Proposer decomposes into claim-level verifiable atomic propositions. Crucially, the Checker validates these propositions against retrieved evidence in isolation, deprived of the Solver's original output. This well-crafted information asymmetry scheme breaks the cycle of self-confirmation bias. By training this pipeline with multi-agent reinforcement learning (MARL), we enable the agents to co-evolve and optimize factual adherence. Extensive experiments across hallucination benchmarks demonstrate that MARCH substantially reduces hallucination rates. Notably, an 8B-parameter LLM equipped with MARCH achieves performance competitive with powerful closed-source models. MARCH paves a scalable path for factual self-improvement of LLMs through co-evolution. The code is at https://github.com/Qwen-Applications/MARCH.", "authors": ["Zhuo Li", "Yupeng Zhang", "Pengyu Cheng", "Jiajun Song", "Mengyu Zhou", "Hao Li", "Shujie Hu", "Yu Qin", "Erchao Zhao", "Xiaoxi Jiang", "Guanjun Jiang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-25", "url": "https://arxiv.org/abs/2603.24579", "pdf_url": "https://arxiv.org/pdf/2603.24579v1", "arxiv_id": "2603.24579", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Qwen-Applications/MARCH", "venue": null, "quality_score": 0.6189}
{"id": "f8a8f2fa9308006de5991f28097a07ef96b072346c9eda498ecd81904c61f148", "sources": ["arxiv", "semantic_scholar"], "title": "Relaxing Constraints in Anonymous Multi Agent Path Finding for Large Agents", "abstract": "The study addressed the problem of Anonymous Multi-Agent Path-finding (AMAPF). Unlike the classical formulation, where the assignment of agents to goals is fixed, in the anonymous MAPF setting it is irrelevant which agent reaches specific goal, provided that all goals are occupied. Most existing multi-agent pathfinding algorithms rely on a discrete representation of the environment (e.g., square grids) and do not account for the sizes of agents. This limits their applicability in real-world scenarios, such as trajectory planning for mobile robots in warehouses. Conversely, methods operating in continuous space typically impose substantial restrictions on the input data, such as constraints on the distances between initial and goal positions or between start/goal positions and obstacles. In this work, we considered one of the AMAPF algorithms designed for continuous space, where agents are modeled as disks of equal size. The algorithm requires a strict minimum separation of $4$ agent radii between any start/goal positions. Proposed a modification aimed at relaxing the constraints and reduce this limit from $4$ to $2\\sqrt{3}$. We theoretically demonstrated that the proposed enhancements preserve original theoretical properties, including the guarantee that all agents will eventually achieve their goals safely and without collisions.", "authors": ["Stepan Dergachev", "Dmitry Avdeev"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-25", "url": "https://arxiv.org/abs/2603.24442", "pdf_url": "https://arxiv.org/pdf/2603.24442v1", "arxiv_id": "2603.24442", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3332}
{"id": "93f5da6ea2bc58f1d02997f28b5ad833ec21b48e5d6f899e51cfc5ffca11f54e", "sources": ["arxiv", "semantic_scholar"], "title": "Experiential Reflective Learning for Self-Improving LLM Agents", "abstract": "Recent advances in large language models (LLMs) have enabled the development of autonomous agents capable of complex reasoning and multi-step problem solving. However, these agents struggle to adapt to specialized environments and do not leverage past interactions, approaching each new task from scratch regardless of their accumulated experience. We introduce Experiential Reflective Learning (ERL), a simple self-improvement framework that enables rapid environment adaptation through experiential learning. ERL reflects on task trajectories and outcomes to generate heuristics, capturing actionable lessons that transfer across tasks. At test time, relevant heuristics are retrieved based on the current task and injected into the agent's context to guide execution. On the Gaia2 benchmark, ERL improves success rate by 7.8% over a ReAct baseline, with large gains in task completion reliability, and outperforms prior experiential learning methods. Through systematic ablations, we find that selective retrieval is essential and that heuristics provide more transferable abstractions than few-shot trajectory prompting. These results demonstrate that reflecting on single-attempt experiences to extract transferable heuristics enables effective agent self-improvement.", "authors": ["Marc-Antoine Allard", "Arnaud Teinturier", "Victor Xing", "Gautier Viaud"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-25", "url": "https://arxiv.org/abs/2603.24639", "pdf_url": "https://arxiv.org/pdf/2603.24639v2", "arxiv_id": "2603.24639", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3332}
{"id": "acf3f37a3fbe8d863f5d213faae690a47b8f8b3f0fb4998798bd6738c2722df6", "sources": ["arxiv", "semantic_scholar"], "title": "The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestration", "abstract": "Tool use enables large language models (LLMs) to access external information, invoke software systems, and act in digital environments beyond what can be solved from model parameters alone. Early research mainly studied whether a model could select and execute a correct single tool call. As agent systems evolve, however, the central problem has shifted from isolated invocation to multi-tool orchestration over long trajectories with intermediate state, execution feedback, changing environments, and practical constraints such as safety, cost, and verifiability. We comprehensively review recent progress in multi-tool LLM agents and analyzes the state of the art in this rapidly developing area. First, we unify task formulations and distinguish single-call tool use from long-horizon orchestration. Then, we organize the literature around six core dimensions: inference-time planning and execution, training and trajectory construction, safety and control, efficiency under resource constraints, capability completeness in open environments, and benchmark design and evaluation. We further summarize representative applications in software engineering, enterprise workflows, graphical user interfaces, and mobile systems. Finally, we discuss major challenges and outline future directions for building reliable, scalable, and verifiable multi-tool agents.", "authors": ["Haoyuan Xu", "Chang Li", "Xinyan Ma", "Xianhao Ou", "Zihan Zhang", "Tao He", "Xiangyu Liu", "Zixiang Wang", "Jiafeng Liang", "Zheng Chu", "Runxuan Liu", "Rongchuan Mu", "Dandan Tu", "Ming Liu", "Bing Qin"], "categories": ["cs.SE", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-24", "url": "https://arxiv.org/abs/2603.22862", "pdf_url": "https://arxiv.org/pdf/2603.22862v2", "arxiv_id": "2603.22862", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3325}
{"id": "77a79c2ca8120cbf543d34cfbb3089dedad0be77b428bb0b32f49ca8d62a89d3", "sources": ["arxiv", "semantic_scholar"], "title": "SoK: The Attack Surface of Agentic AI -- Tools, and Autonomy", "abstract": "Recent AI systems combine large language models with tools, external knowledge via retrieval-augmented generation (RAG), and even autonomous multi-agent decision loops. This agentic AI paradigm greatly expands capabilities - but also vastly enlarges the attack surface. In this systematization, we map out the trust boundaries and security risks of agentic LLM-based systems. We develop a comprehensive taxonomy of attacks spanning prompt-level injections, knowledge-base poisoning, tool/plug-in exploits, and multi-agent emergent threats. Through a detailed literature review, we synthesize evidence from 2023-2025, including more than 20 peer-reviewed and archival studies, industry reports, and standards. We find that agentic systems introduce new vectors for indirect prompt injection, code execution exploits, RAG index poisoning, and cross-agent manipulation that go beyond traditional AI threats. We define attacker models and threat scenarios, and propose metrics (e.g., Unsafe Action Rate, Privilege Escalation Distance) to evaluate security posture. Our survey examines defenses such as input sanitization, retrieval filters, sandboxes, access control, and \"AI guardrails,\" assessing their effectiveness and pointing out the areas where protection is still lacking. To assist practitioners, we outline defensive controls and provide a phased security checklist for deploying agentic AI (covering design-time hardening, runtime monitoring, and incident response). Finally, we outline open research challenges in secure autonomous AI (robust tool APIs, verifiable agent behavior, supply-chain safeguards) and discuss ethical and responsible disclosure practices. We systematize recent findings to help researchers and engineers understand and mitigate security risks in agentic AI.", "authors": ["Ali Dehghantanha", "Sajad Homayoun"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-24", "url": "https://arxiv.org/abs/2603.22928", "pdf_url": "https://arxiv.org/pdf/2603.22928v1", "arxiv_id": "2603.22928", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3325}
{"id": "9349436b93fde48e1a58fdbf0c04fd08d3f9d991539be096e0c5ceb4ce095bc7", "sources": ["arxiv", "semantic_scholar"], "title": "A Multimodal Framework for Human-Multi-Agent Interaction", "abstract": "Human-robot interaction is increasingly moving toward multi-robot, socially grounded environments. Existing systems struggle to integrate multimodal perception, embodied expression, and coordinated decision-making in a unified framework. This limits natural and scalable interaction in shared physical spaces. We address this gap by introducing a multimodal framework for human-multi-agent interaction in which each robot operates as an autonomous cognitive agent with integrated multimodal perception and Large Language Model (LLM)-driven planning grounded in embodiment. At the team level, a centralized coordination mechanism regulates turn-taking and agent participation to prevent overlapping speech and conflicting actions. Implemented on two humanoid robots, our framework enables coherent multi-agent interaction through interaction policies that combine speech, gesture, gaze, and locomotion. Representative interaction runs demonstrate coordinated multimodal reasoning across agents and grounded embodied responses. Future work will focus on larger-scale user studies and deeper exploration of socially grounded multi-agent interaction dynamics.", "authors": ["Shaid Hasan", "Breenice Lee", "Sujan Sarker", "Tariq Iqbal"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-24", "url": "https://arxiv.org/abs/2603.23271", "pdf_url": "https://arxiv.org/pdf/2603.23271v1", "arxiv_id": "2603.23271", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3325}
{"id": "397dedd2524e67ee0000d8b1f120a156db2fcf8dc64c15214e6a3a95e26feff4", "sources": ["arxiv", "semantic_scholar"], "title": "Empirical Comparison of Agent Communication Protocols for Task Orchestration", "abstract": "Context. The problem of comparative evaluation of communication protocols for task orchestration by large language model (LLM) agents is considered. The object of study is the process of interaction between LLM agents and external tools, as well as between autonomous LLM agents, during task orchestration. Objective. The goal of this work is to develop a systematic pilot benchmark comparing tool integration, multi-agent dele-gation, and hybrid architectures for standardized queries at three levels of complexity, and to quantify the advantages and disadvantages in terms of response time, context window consumption, cost, error recovery, and implementation complexity.", "authors": ["Ivan Dobrovolskyi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-24", "url": "https://arxiv.org/abs/2603.22823", "pdf_url": "https://arxiv.org/pdf/2603.22823v3", "arxiv_id": "2603.22823", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3325}
{"id": "52902950713cc36649a13258ebc1148107bc21d3aabe8fe9cfdb45047fb70230", "sources": ["arxiv", "semantic_scholar"], "title": "Strategic Infrastructure Design via Multi-Agent Congestion Games with Joint Placement and Pricing", "abstract": "Real-world infrastructure planning increasingly involves strategic interactions among autonomous agents competing over congestible, limited resources. Applications such as Electric Vehicle (EV) charging, emergency response, and intelligent transportation require coordinated resource placement and pricing decisions, while anticipating the adaptive behaviour of decentralised, self-interested agents. We propose a novel multi-agent framework for joint placement and pricing under such interactions, formalised as a bi-level optimisation model. The upper level represents a central planner, while the lower level captures agent responses via coupled non-atomic congestion games. Motivated by the EV charging domain, we study a setting where a central planner provisions chargers and road capacity under budget and profitability constraints. The agent population includes both EV drivers and non-charging drivers (NCDs), who respond to congestion, delays, and costs. To solve the resulting NP-hard problem, we introduce ABO-MPN, a double-layer approximation framework that decouples agent types, applies integer adjustment and rounding, and targets high-impact placement and pricing decisions. Experiments on benchmark networks show that our model reduces social cost by up to 40% compared to placement- or pricing-only baselines, and generalises to other MAS-relevant domains.", "authors": ["Niloofar Aminikalibar", "Farzaneh Farhadi", "Maria Chli"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.21691", "pdf_url": "https://arxiv.org/pdf/2603.21691v1", "arxiv_id": "2603.21691", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3318}
{"id": "07d86ff217899147abee99188960eeb904b13c1d1ad1b8670108ff150f66f39a", "sources": ["arxiv", "semantic_scholar"], "title": "Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe", "abstract": "Reinforcement Learning (RL) is essential for evolving Large Language Models (LLMs) into autonomous agents capable of long-horizon planning, yet a practical recipe for scaling RL in complex, multi-turn environments remains elusive. This paper presents a systematic empirical study using TravelPlanner, a challenging testbed requiring tool orchestration to satisfy multifaceted constraints. We decompose the agentic RL design space along 5 axes: reward shaping, model scaling, data composition, algorithm selection, and environmental stability. Our controlled experiments yield 7 key takeaways, e.g., (1) reward and algorithm choices are scale-dependent as smaller models benefit from staged rewards and enhanced exploration, whereas larger models converge efficiently with simpler dense rewards, (2) ~ 1K training samples with a balanced difficulty mixture mark a sweet spot for both in-domain and out-of-domain performance, and (3) environmental stability is critical to prevent policy degradation. Based on our distilled recipe, our RL-trained models achieve state-of-the-art performance on TravelPlanner, significantly outperforming leading LLMs.", "authors": ["Xixi Wu", "Qianguo Sun", "Ruiyang Zhang", "Chao Song", "Junlong Wu", "Yiyan Qi", "Hong Cheng"], "categories": ["cs.LG", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.21972", "pdf_url": "https://arxiv.org/pdf/2603.21972v1", "arxiv_id": "2603.21972", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/WxxShirley/Agent-STAR", "venue": null, "quality_score": 0.6161}
{"id": "4dfc55b91d8be427d7f2997357c0f61f50529779475861fbd28a3512fa729874", "sources": ["arxiv", "semantic_scholar"], "title": "OpenEarth-Agent: From Tool Calling to Tool Creation for Open-Environment Earth Observation", "abstract": "Earth Observation (EO) is essential for perceiving dynamic land surface changes, yet deploying autonomous EO in open environments is hindered by the immense diversity of multi-source data and heterogeneous tasks. While remote sensing agents have emerged to streamline EO workflows, existing tool-calling agents are confined to closed environments. They rely on pre-defined tools and are restricted to narrow scope, limiting their generalization to the diverse data and tasks. To overcome these limitations, we introduce OpenEarth-Agent, the first tool-creation agent framework tailored for open-environment EO. Rather than calling predefined tools, OpenEarth-Agent employs adaptive workflow planning and tool creation to generalize to unseen data and tasks. This adaptability is bolstered by an open-ended integration of multi-stage tools and cross-domain knowledge bases, enabling robust execution in the entire EO pipeline across multiple application domains. To comprehensively evaluate EO agents in open environments, we propose OpenEarth-Bench, a novel benchmark comprising 596 real-world, full-pipeline cases across seven application domains, explicitly designed to assess agents' adaptive planning and tool creation capabilities. Only essential pre-trained model tools are provided in this benchmark, devoid of any other predefined task-specific tools. Extensive experiments demonstrate that OpenEarth-Agent successfully masters full-pipeline EO across multiple domains in the open environment. Notably, on the cross-benchmark Earth-Bench, our tool-creating agent equipped with 6 essential pre-trained models achieves performance comparable to tool-calling agents relying on 104 specialized tools, and significantly outperforms them when provided with the complete toolset. In several cases, the created tools exhibit superior robustness to data anomalies compared to human-engineered counterparts.", "authors": ["Sijie Zhao", "Feng Liu", "Xueliang Zhang", "Hao Chen", "Xinyu Gu", "Zhe Jiang", "Fenghua Ling", "Ben Fei", "Wenlong Zhang", "Junjue Wang", "Weihao Xuan", "Pengfeng Xiao", "Naoto Yokoya", "Lei Bai"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.22148", "pdf_url": "https://arxiv.org/pdf/2603.22148v1", "arxiv_id": "2603.22148", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3318}
{"id": "474735a8b86aba14cbceab6592ebe2b2161040aacc810d4fe359f28a2a0be309", "sources": ["arxiv", "semantic_scholar"], "title": "Chimera: Latency- and Performance-Aware Multi-agent Serving for Heterogeneous LLMs", "abstract": "Multi-agent applications often execute complex tasks as multi-stage workflows, where each stage is an LLM call whose output becomes part of context for subsequent steps. Existing LLM serving systems largely assume homogeneous clusters with identical model replicas. This design overlooks the potential of heterogeneous deployments, where models of different sizes and capabilities enable finer trade-offs between latency and performance. However, heterogeneity introduces new challenges in scheduling across models with diverse throughput and performance. We present Chimera, a predictive scheduling system for multi-agent workflow serving on heterogeneous LLM clusters that jointly improves end-to-end latency and task performance. Chimera applies semantic routing to estimate per-model confidence scores for each request, predicts the total remaining output length of the workflow, and estimates per-model congestion using in-flight predicted token volumes for load balancing. We evaluate Chimera on representative agentic workflows for code generation and math reasoning using multiple heterogeneous LLM configurations. Across comparable settings, Chimera traces the best latency-performance frontier, reducing end-to-end latency by 1.2--2.4$\\times$ and improving task performance by 8.0-9.5 percentage points on average over competitive baselines including vLLM.", "authors": ["Kangqi Ni", "Wenyue Hua", "Xiaoxiang Shi", "Jiang Guo", "Shiyu Chang", "Tianlong Chen"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.22206", "pdf_url": "https://arxiv.org/pdf/2603.22206v1", "arxiv_id": "2603.22206", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3318}
{"id": "a105fad8e487789f8a4638d841e8ea50f92755ae2274c911a74548328dfa9902", "sources": ["arxiv", "semantic_scholar"], "title": "Counterfactual Credit Policy Optimization for Multi-Agent Collaboration", "abstract": "Collaborative multi-agent large language models (LLMs) can solve complex reasoning tasks by decomposing roles, but reinforcement learning for such systems is limited by credit assignment: shared terminal rewards obscure individual contributions and can encourage free-riding. We introduce Collaborative Credit Policy Optimization (CCPO), an optimizer-agnostic credit assignment layer that converts team-level outcomes into agent-specific learning signals. CCPO provides two complementary allocators. Counterfactual credit estimates an agent's marginal contribution by comparing the realized team outcome with a counterfactual outcome where that agent is removed. Verifier-anchored LLM self-evaluation is an exploratory allocator that uses constrained self- and peer-evaluations to redistribute credit while keeping the external verifier outcome dominant. The resulting role-specific rewards can be consumed by GRPO-style updates or other policy-gradient optimizers such as GSPO and REINFORCE++. We instantiate CCPO in a sequential Think--Solve setting and evaluate it on mathematical reasoning benchmarks. Results show that explicit credit assignment often improves dual-agent reasoning, especially on MATH500 and several out-of-distribution settings, while gains vary across models and datasets.", "authors": ["Zhongyi Li", "Wan Tian", "Yikun Ban", "Jinju Chen", "Huiming Zhang", "Yang Liu", "Fuzhen Zhuang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.21563", "pdf_url": "https://arxiv.org/pdf/2603.21563v4", "arxiv_id": "2603.21563", "doi": null, "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3318}
{"id": "198632b5e2019deb52ef5a6a3874a7b6da66f82b75e9a04437a9d6ac70cde066", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Automation of BT-RADS Scoring: End-to-End Multi-Agent System for Standardized Brain Tumor Follow-up Assessment", "abstract": "The Brain Tumor Reporting and Data System (BT-RADS) standardizes post-treatment MRI response assessment in patients with diffuse gliomas but requires complex integration of imaging trends, medication effects, and radiation timing. This study evaluates an end-to-end multi-agent large language model (LLM) and convolutional neural network (CNN) system for automated BT-RADS classification. A multi-agent LLM system combined with automated CNN-based tumor segmentation was retrospectively evaluated on 509 consecutive post-treatment glioma MRI examinations from a single high-volume center. An extractor agent identified clinical variables (steroid status, bevacizumab status, radiation date) from unstructured clinical notes, while a scorer agent applied BT-RADS decision logic integrating extracted variables with volumetric measurements. Expert reference standard classifications were established by an independent board-certified neuroradiologist. Of 509 examinations, 492 met inclusion criteria. The system achieved 374/492 (76.0%; 95% CI, 72.1%-79.6%) accuracy versus 283/492 (57.5%; 95% CI, 53.1%-61.8%) for initial clinical assessments (+18.5 percentage points; P<.001). Context-dependent categories showed high sensitivity (BT-1b 100%, BT-1a 92.7%, BT-3a 87.5%), while threshold-dependent categories showed moderate sensitivity (BT-3c 74.8%, BT-2 69.2%, BT-4 69.3%, BT-3b 57.1%). For BT-4, positive predictive value was 92.9%. The multi-agent LLM system achieved higher BT-RADS classification agreement with expert reference standard compared to initial clinical scoring, with high accuracy for context-dependent scores and high positive predictive value for BT-4 detection.", "authors": ["Mohamed Sobhi Jabal", "Jikai Zhang", "Dominic LaBella", "Jessica L. Houk", "Dylan Zhang", "Jeffrey D. Rudie", "Kirti Magudia", "Maciej A. Mazurowski", "Evan Calabrese"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-23", "url": "https://arxiv.org/abs/2603.21494", "pdf_url": "https://arxiv.org/pdf/2603.21494v2", "arxiv_id": "2603.21494", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3318}
{"id": "5deba8e073fe56adaceda7286fe7c62109d48d69336e910b4d107238c48841c6", "sources": ["arxiv", "semantic_scholar"], "title": "STEM Agent: A Self-Adapting, Tool-Enabled, Extensible Architecture for Multi-Protocol AI Agent Systems", "abstract": "Current AI agent frameworks commit early to a single interaction protocol, a fixed tool integration strategy, and static user models, limiting their deployment across diverse interaction paradigms. To address these constraints, we introduce STEM Agent (Self-adapting, Tool-enabled, Extensible, Multi-agent), a modular architecture inspired by biological pluripotency in which an undifferentiated agent core differentiates into specialized protocol handlers, tool bindings, and memory subsystems that compose into a fully functioning AI system. The framework unifies five interoperability protocols (A2A, AG-UI, A2UI, UCP, and AP2) behind a single gateway, introduces a Caller Profiler that continuously learns user preferences across more than twenty behavioral dimensions, externalizes all domain capabilities through the Model Context Protocol (MCP), and implements a biologically inspired skills acquisition system in which recurring interaction patterns crystallize into reusable agent skills through a maturation lifecycle analogous to cell differentiation. Complementing these capabilities, the memory system incorporates consolidation mechanisms, including episodic pruning, semantic deduplication, and pattern extraction, designed for sub-linear growth under sustained interaction. A comprehensive 413-test suite validates protocol handler behavior and component integration across all five architectural layers, completing in under three seconds.", "authors": ["Alfred Shen", "Aaron Shen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-22", "url": "https://arxiv.org/abs/2603.22359", "pdf_url": "https://arxiv.org/pdf/2603.22359v1", "arxiv_id": "2603.22359", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.331}
{"id": "a8068c4e78b099fc47e28ad340cb0e237e1477beedc734cd52917228c686f42c", "sources": ["arxiv", "semantic_scholar"], "title": "Utility-Guided Agent Orchestration for Efficient LLM Tool Use", "abstract": "Tool-using large language model (LLM) agents often face a fundamental tension between answer quality and execution cost. Fixed workflows are stable but inflexible, while free-form multi-step reasoning methods such as ReAct may improve task performance at the expense of excessive tool calls, longer trajectories, higher token consumption, and increased latency. In this paper, we study agent orchestration as an explicit decision problem rather than leaving it entirely to prompt-level behavior. We propose a utility-guided orchestration policy that selects among actions such as respond, retrieve, tool call, verify, and stop by balancing estimated gain, step cost, uncertainty, and redundancy. Our goal is not to claim universally best task performance, but to provide a controllable and analyzable policy framework for studying quality-cost trade-offs in tool-using LLM agents. Experiments across direct answering, threshold control, fixed workflows, ReAct, and several policy variants show that explicit orchestration signals substantially affect agent behavior. Additional analyses on cost definitions, workflow fairness, and redundancy control further demonstrate that lightweight utility design can provide a defensible and practical mechanism for agent control.", "authors": ["Boyan Liu", "Gongming Zhao", "Hongli Xu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.19896", "pdf_url": "https://arxiv.org/pdf/2603.19896v1", "arxiv_id": "2603.19896", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3296}
{"id": "b34c208a04316e96ff94514df01c601ffcb37b3356a1f502a8460b6c35440c46", "sources": ["arxiv", "semantic_scholar"], "title": "Synergistic Perception and Generative Recomposition: A Multi-Agent Orchestration for Expert-Level Building Inspection", "abstract": "Building facade defect inspection is fundamental to structural health monitoring and sustainable urban maintenance, yet it remains a formidable challenge due to extreme geometric variability, low contrast against complex backgrounds, and the inherent complexity of composite defects (e.g., cracks co-occurring with spalling). Such characteristics lead to severe pixel imbalance and feature ambiguity, which, coupled with the critical scarcity of high-quality pixel-level annotations, hinder the generalization of existing detection and segmentation models. To address gaps, we propose \\textit{FacadeFixer}, a unified multi-agent framework that treats defect perception as a collaborative reasoning task rather than isolated recognition. Specifically,\\textit{FacadeFixer} orchestrates specialized agents for detection and segmentation to handle multi-type defect interference, working in tandem with a generative agent to enable semantic recomposition. This process decouples intricate defects from noisy backgrounds and realistically synthesizes them onto diverse clean textures, generating high-fidelity augmented data with precise expert-level masks. To support this, we introduce a comprehensive multi-task dataset covering six primary facade categories with pixel-level annotations. Extensive experiments demonstrate that \\textit{FacadeFixer} significantly outperforms state-of-the-art (SOTA) baselines. Specifically, it excels in capturing pixel-level structural anomalies and highlights generative synthesis as a robust solution to data scarcity in infrastructure inspection. Our code and dataset will be made publicly available.", "authors": ["Hui Zhong", "Yichun Gao", "Luyan Liu", "Xusen Guo", "Zhaonian Kuang", "Qiming Zhang", "Xinhu Zheng"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.20143", "pdf_url": "https://arxiv.org/pdf/2603.20143v2", "arxiv_id": "2603.20143", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3296}
{"id": "1069410b9ed739490bd0166a8e597e1392c4edf8c091718680c0d4c9edec7f3b", "sources": ["arxiv", "semantic_scholar"], "title": "Solver-Aided Verification of Policy Compliance in Tool-Augmented LLM Agents", "abstract": "Tool-augmented Large Language Models (TaLLMs) extend LLMs with the ability to invoke external tools, enabling them to interact with real-world environments. However, a major limitation in deploying TaLLMs in sensitive applications such as customer service and business process automation is a lack of reliable compliance with domain-specific operational policies regarding tool-use and agent behavior. Current approaches merely steer LLMs to adhere to policies by including policy descriptions in the LLM context, but these provide no guarantees that policy violations will be prevented. In this paper, we introduce an SMT solver-aided framework to enforce tool-use policy compliance in TaLLM agents. Specifically, we use an LLM-assisted, human-guided approach to translate natural-language-specified tool-use policies into formal logic (SMT-LIB-2.0) constraints over agent-observable state and tool arguments. At runtime, planned tool calls are intercepted and checked against the constraints using the Z3 solver as a pre-condition to the tool call. Tool invocations that violate the policy are blocked. We evaluated on the TauBench benchmark and demonstrate that solver-aided policy checking reduces policy violations while maintaining overall task accuracy. These results suggest that integrating formal reasoning into TaLLM execution can improve tool-call policy compliance and overall reliability.", "authors": ["Cailin Winston", "Claris Winston", "René Just"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.20449", "pdf_url": "https://arxiv.org/pdf/2603.20449v1", "arxiv_id": "2603.20449", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3296}
{"id": "56f2de83ba1223ba817be2bf1b888a37715640de39de2b6a6f889815bb2a50d8", "sources": ["arxiv", "semantic_scholar"], "title": "GoAgent: Group-of-Agents Communication Topology Generation for LLM-based Multi-Agent Systems", "abstract": "Large language model (LLM)-based multi-agent systems (MAS) have demonstrated exceptional capabilities in solving complex tasks, yet their effectiveness depends heavily on the underlying communication topology that coordinates agent interactions. Within these systems, successful problem-solving often necessitates task-specific group structures to divide and conquer subtasks. However, most existing approaches generate communication topologies in a node-centric manner, leaving group structures to emerge implicitly from local connectivity decisions rather than modeling them explicitly, often leading to suboptimal coordination and unnecessary communication overhead. To address this limitation, we propose GoAgent (Group-of-Agents), a communication topology generation method that explicitly treats collaborative groups as the atomic units of MAS construction. Specifically, GoAgent first enumerates task-relevant candidate groups through an LLM and then autoregressively selects and connects these groups as atomic units to construct the final communication graph, jointly capturing intra-group cohesion and inter-group coordination. To mitigate communication redundancy and noise propagation inherent in expanding topologies, we further introduce a conditional information bottleneck (CIB) objective that compresses inter-group communication, preserving task-relevant signals while filtering out redundant historical noise. Extensive experiments on six benchmarks demonstrate the state-of-the-art performance of GoAgent with 93.84% average accuracy while reducing token consumption by about 17%.", "authors": ["Hongjiang Chen", "Xin Zheng", "Yixin Liu", "Pengfei Jiao", "Shiyuan Li", "Huan Liu", "Zhidong Zhao", "Ziqi Xu", "Ibrahim Khalil", "Shirui Pan"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.19677", "pdf_url": "https://arxiv.org/pdf/2603.19677v1", "arxiv_id": "2603.19677", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3296}
{"id": "8eae172afb059f33020702a065cb995fb2ee5130ce217720426390864129dd11", "sources": ["arxiv", "semantic_scholar"], "title": "Herding CATs: ALARA for Agent Harness Engineering in Portable Composable Multi-Agent Teams", "abstract": "Industry practitioners and academic researchers regularly use multi-agent systems to accelerate their work, but the applications through which users operate these systems do not provide a simple, unified mechanism for scalably managing critical components of the agent harness. This lack of control adversely impacts both the quality of individual human-agent interactions and reduces the capacity for practitioners to coordinate context engineering efforts. The behavioral specifications that define what agents in such systems can do remain fragmented across prose instruction files -- for which compliance cannot be guaranteed -- or framework-internal configurations, making these specifications difficult to share, version, or collaboratively maintain across teams and projects. Applying the ALARA principle from radiation safety (exposures kept as low as reasonably achievable) to context, we introduce a context-agent-tool (CAT) data layer expressed through interrelated plain-text files, allowing users to directly declare tool access for each agent and to modify the tools themselves that are used by the agents when processing. We demonstrate capability of this CAT data layer to enable real agentic usage by using a command-line shell that loads the team and executes agent runs -- \\texttt{npcsh} -- and evaluating 22 locally-hosted models from 0.6B to 35B parameters across 115 practical tasks spanning file operations, web search, multi-step scripting, tool chaining, and multi-agent delegation. We characterize which model families succeed in certain task categories and where they break down across $\\sim$2500 total executions.", "authors": ["Christopher J. Agostino", "Nayan D'Souza"], "categories": ["cs.MA", "cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.20380", "pdf_url": "https://arxiv.org/pdf/2603.20380v2", "arxiv_id": "2603.20380", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3296}
{"id": "95e1df414e837493d868102efa45dca5467e0123f0d46c155169e97658925094", "sources": ["arxiv", "semantic_scholar"], "title": "When Agents Disagree: The Selection Bottleneck in Multi-Agent LLM Pipelines", "abstract": "Multi-agent LLM pipelines produce contradictory evidence on whether team diversity improves output quality: heterogeneous Mixture-of-Agents teams outperform single models, yet homogeneous Self-MoA teams consistently win under synthesis-based aggregation. We propose a resolution by identifying the selection bottleneck -- a crossover threshold in aggregation quality that determines whether diversity helps or hurts. Under this model, we obtain a closed-form crossover threshold $s^*$ (Proposition 1) that separates the regimes where diversity helps and hurts. In a targeted experiment spanning 42 tasks across 7 categories ($N=210$), a diverse team with judge-based selection achieves a win rate of 0.810 against a single-model baseline, while a homogeneous team scores 0.512 -- near chance (Glass's $Δ= 2.07$). Judge-based selection outperforms MoA-style synthesis by $Δ_{\\mathrm{WR}} = +0.631$ -- the synthesis approach is preferred over the baseline in zero of 42 tasks by the judge panel. A decoupled evaluation with independent judges confirms all directional findings (Spearman $ρ= 0.90$). Exploratory evidence suggests that including a weaker model improves performance while reducing cost ($p < 10^{-4}$, not pre-registered). Our results suggest that selector quality may be a more impactful design lever than generator diversity in single-round generate-then-select pipelines.", "authors": ["Artem Maryanskyy"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-20", "url": "https://arxiv.org/abs/2603.20324", "pdf_url": "https://arxiv.org/pdf/2603.20324v1", "arxiv_id": "2603.20324", "doi": "10.3390/app16104914", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Applied Sciences", "quality_score": 0.5179}
{"id": "a55f65d84369dcf54747b35a89c54df6dca402dad4fd2de102617ece6889c07d", "sources": ["arxiv", "semantic_scholar"], "title": "Meanings and Measurements: Multi-Agent Probabilistic Grounding for Vision-Language Navigation", "abstract": "Robots collaborating with humans must convert natural language goals into actionable, physically grounded decisions. For example, executing a command such as \"go two meters to the right of the fridge\" requires grounding semantic references, spatial relations, and metric constraints within a 3D scene. While recent vision language models (VLMs) demonstrate strong semantic grounding capabilities, they are not explicitly designed to reason about metric constraints in physically defined spaces. In this work, we empirically demonstrate that state-of-the-art VLM-based grounding approaches struggle with complex metric-semantic language queries. To address this limitation, we propose MAPG (Multi-Agent Probabilistic Grounding), an agentic framework that decomposes language queries into structured subcomponents and queries a VLM to ground each component. MAPG then probabilistically composes these grounded outputs to produce metrically consistent, actionable decisions in 3D space. We evaluate MAPG on the HM-EQA benchmark and show consistent performance improvements over strong baselines. Furthermore, we introduce a new benchmark, MAPG-Bench, specifically designed to evaluate metric-semantic goal grounding, addressing a gap in existing language grounding evaluations. We also present a real-world robot demonstration showing that MAPG transfers beyond simulation when a structured scene representation is available.", "authors": ["Swagat Padhan", "Lakshya Jain", "Bhavya Minesh Shah", "Omkar Patil", "Thao Nguyen", "Nakul Gopalan"], "categories": ["cs.RO", "cs.AI", "cs.CL", "cs.CV", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.19166", "pdf_url": "https://arxiv.org/pdf/2603.19166v1", "arxiv_id": "2603.19166", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3289}
{"id": "f23b60377a27579163a7b8c4c95a7dfc11b0f9bd8e7906fca40a003055f301fc", "sources": ["arxiv", "semantic_scholar"], "title": "ProRL Agent: Rollout-as-a-Service for RL Training of Multi-Turn LLM Agents", "abstract": "Multi-turn LLM agents are increasingly important for solving complex, interactive tasks, and reinforcement learning (RL) is a key ingredient for improving their long-horizon behavior. However, RL training requires generating large numbers of sandboxed rollout trajectories, and existing infrastructures often couple rollout orchestration with the training loop, making systems hard to migrate and maintain. Under the rollout-as-a-service philosophy, we present ProRL Agent , a scalable infrastructure that serves the full agentic rollout lifecycle through an API service. ProRL Agent also provides standardized and extensible sandbox environments that support diverse agentic tasks in rootless HPC settings. We validate ProRL Agent through RL training on software engineering, math, STEM, and coding tasks. ProRL Agent is open-sourced and integrated as part of NVIDIA NeMo Gym.", "authors": ["Hao Zhang", "Mingjie Liu", "Shaokun Zhang", "Songyang Han", "Jian Hu", "Zhenghui Jin", "Yuchi Zhang", "Shizhe Diao", "Ximing Lu", "Binfeng Xu", "Zhiding Yu", "Jan Kautz", "Yi Dong"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.18815", "pdf_url": "https://arxiv.org/pdf/2603.18815v1", "arxiv_id": "2603.18815", "doi": null, "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6107}
{"id": "5451f9093268e92870795bb6db99a7f0131d24b6801562a135287179a9707c5b", "sources": ["arxiv", "semantic_scholar"], "title": "Act While Thinking: Accelerating LLM Agents via Pattern-Aware Speculative Tool Execution", "abstract": "LLM-powered agents are emerging as a dominant paradigm for autonomous task solving. Unlike standard inference workloads, agents operate in a strictly serial \"LLM-tool\" loop, where the LLM must wait for external tool execution at every step. This execution model introduces severe latency bottlenecks. To address this problem, we propose PASTE, a Pattern-Aware Speculative Tool Execution method designed to hide tool latency through speculation. PASTE is based on the insight that although agent requests are semantically diverse, they exhibit stable application level control flows (recurring tool-call sequences) and predictable data dependencies (parameter passing between tools). By exploiting these properties, PASTE improves agent serving performance through speculative tool execution. Experimental results against state of the art baselines show that PASTE reduces average task completion time by 48.5% and improves tool execution throughput by 1.8x.", "authors": ["Yifan Sui", "Han Zhao", "Rui Ma", "Zhiyuan He", "Hao Wang", "Jianxun Li", "Yuqing Yang"], "categories": ["cs.DC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.18897", "pdf_url": "https://arxiv.org/pdf/2603.18897v1", "arxiv_id": "2603.18897", "doi": null, "citation_count": 8, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3289}
{"id": "1cf8750fd76516eec95b23b24b8a8adc8a99cb4edf089dd9f5bff681c6064841", "sources": ["arxiv", "semantic_scholar"], "title": "The Autonomy Tax: Defense Training Breaks LLM Agents", "abstract": "Large language model (LLM) agents increasingly rely on external tools (file operations, API calls, database transactions) to autonomously complete complex multi-step tasks. Practitioners deploy defense-trained models to protect against prompt injection attacks that manipulate agent behavior through malicious observations or retrieved content. We reveal a fundamental \\textbf{capability-alignment paradox}: defense training designed to improve safety systematically destroys agent competence while failing to prevent sophisticated attacks. Evaluating defended models against undefended baselines across 97 agent tasks and 1,000 adversarial prompts, we uncover three systematic biases unique to multi-step agents. \\textbf{Agent incompetence bias} manifests as immediate tool execution breakdown, with models refusing or generating invalid actions on benign tasks before observing any external content. \\textbf{Cascade amplification bias} causes early failures to propagate through retry loops, pushing defended models to timeout on 99\\% of tasks compared to 13\\% for baselines. \\textbf{Trigger bias} leads to paradoxical security degradation where defended models perform worse than undefended baselines while straightforward attacks bypass defenses at high rates. Root cause analysis reveals these biases stem from shortcut learning: models overfit to surface attack patterns rather than semantic threat understanding, evidenced by extreme variance in defense effectiveness across attack categories. Our findings demonstrate that current defense paradigms optimize for single-turn refusal benchmarks while rendering multi-step agents fundamentally unreliable, necessitating new approaches that preserve tool execution competence under adversarial conditions.", "authors": ["Shawn Li", "Yue Zhao"], "categories": ["cs.CR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.19423", "pdf_url": "https://arxiv.org/pdf/2603.19423v1", "arxiv_id": "2603.19423", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3289}
{"id": "6d4f5e62b33bff7ad51bac1cec0ed2d90d5556d7051205513d5f602682b6c828", "sources": ["arxiv", "semantic_scholar"], "title": "Mean-field control barrier functions for stochastic multi-agent systems", "abstract": "Many applications involving multi-agent systems require fulfilling safety constraints. Control barrier functions offer a systematic framework to enforce forward invariance of safety sets. Recent work extended this paradigm to mean-field scenarios, where the number of agents is large enough to make density-space descriptions a reasonable workaround for the curse of dimensionality. However, an open gap in the recent literature concerns the development of mean-field control barrier functions for Fokker-Planck (advection-diffusion) equations. In this work, we address this gap, enabling safe mean-field control of agents with stochastic microscopic dynamics. We provide bounded stability guarantees under safety corrections and corroborate our results through numerical simulations in two representative scenarios, coverage and shepherding control of multi-agent systems.", "authors": ["Cinzia Tomaselli", "Gian Carlo Maffettone", "Samy Wu Fung", "Levon Nurbekyan", "Mario di Bernardo"], "categories": ["eess.SY"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.18658", "pdf_url": "https://arxiv.org/pdf/2603.18658v1", "arxiv_id": "2603.18658", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3289}
{"id": "f5b035adeac97a76ec973153e69b1c6e3f9a97ba9cc2e5a82ab843238e5954d7", "sources": ["arxiv", "semantic_scholar"], "title": "The Causal Impact of Tool Affordance on Safety Alignment in LLM Agents", "abstract": "Large language models (LLMs) are increasingly deployed as agents with access to executable tools, enabling direct interaction with external systems. However, most safety evaluations remain text-centric and assume that compliant language implies safe behavior, an assumption that becomes unreliable once models are allowed to act. In this work, we empirically examine how executable tool affordance alters safety alignment in LLM agents using a paired evaluation framework that compares text-only chatbot behavior with tool-enabled agent behavior under identical prompts and policies. Experiments are conducted in a deterministic financial transaction environment with binary safety constraints across 1,500 procedurally generated scenarios. To separate intent from outcome, we distinguish between attempted and realized violations using dual enforcement regimes that either block or permit unsafe actions. Both evaluated models maintain perfect compliance in text-only settings, yet exhibit sharp increases in violations after tool access is introduced, reaching rates up to 85% despite unchanged rules. We observe substantial gaps between attempted and executed violations, indicating that external guardrails can suppress visible harm while masking persistent misalignment. Agents also develop spontaneous constraint circumvention strategies without adversarial prompting. These results demonstrate that tool affordance acts as a primary driver of safety misalignment and that text-based evaluation alone is insufficient for assessing agentic systems.", "authors": ["Shasha Yu", "Fiona Carroll", "Barry L. Bentley"], "categories": ["cs.SE", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-19", "url": "https://arxiv.org/abs/2603.20320", "pdf_url": "https://arxiv.org/pdf/2603.20320v1", "arxiv_id": "2603.20320", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3289}
{"id": "6a5b009cdf9e659f6d678c4ebf7eb31a977a002accdc5792a34ba1ae294bd20f", "sources": ["arxiv", "semantic_scholar"], "title": "Who Tests the Testers? Systematic Enumeration and Coverage Audit of LLM Agent Tool Call Safety", "abstract": "Large Language Model (LLM) agents increasingly act through external tools, making their safety contingent on tool-call workflows rather than text generation alone. While recent benchmarks evaluate agents across diverse environments and risk categories, a fundamental question remains unanswered: how complete are existing test suites, and what unsafe interaction patterns persist even after an agent passes the benchmark? We propose SafeAudit, a meta-audit framework that addresses this gap through two contributions. First, an LLM-based enumerator that systematically generates test cases by enumerating valid tool-call workflows and diverse user scenarios. Second, we introduce rule-resistance, a non-semantic, quantitative metric that distills compact safety rules from existing benchmarks and identifies unsafe interaction patterns that remain uncovered under those rules. Across 3 benchmarks and 12 environments, SafeAudit uncovers more than 20% residual unsafe behaviors that existing benchmarks fail to expose, with coverage growing monotonically as the testing budget increases. Our results highlight significant completeness gaps in current safety evaluation and motivate meta-auditing as a necessary complement to benchmark-based agent safety testing.", "authors": ["Xuan Chen", "Lu Yan", "Ruqi Zhang", "Xiangyu Zhang"], "categories": ["cs.SE", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-18", "url": "https://arxiv.org/abs/2603.18245", "pdf_url": "https://arxiv.org/pdf/2603.18245v1", "arxiv_id": "2603.18245", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3281}
{"id": "bf675025e9fddc647486059a23a02c71b744d6f5bcc25b781ad032a55f3e0835", "sources": ["arxiv", "semantic_scholar"], "title": "MALLES: A Multi-agent LLMs-based Economic Sandbox with Consumer Preference Alignment", "abstract": "In the real economy, modern decision-making is fundamentally challenged by high-dimensional, multimodal environments, which are further complicated by agent heterogeneity and combinatorial data sparsity. This paper introduces a Multi-Agent Large Language Model-based Economic Sandbox (MALLES), leveraging the inherent generalization capabilities of large-sacle models to establish a unified simulation framework applicable to cross-domain and cross-category scenarios. Central to our approach is a preference learning paradigm in which LLMs are economically aligned via post-training on extensive, heterogeneous transaction records across diverse product categories. This methodology enables the models to internalize and transfer latent consumer preference patterns, thereby mitigating the data sparsity issues prevalent in individual categories. To enhance simulation stability, we implement a mean-field mechanism designed to model the dynamic interactions between the product environment and customer populations, effectively stabilizing sampling processes within high-dimensional decision spaces. Furthermore, we propose a multi-agent discussion framework wherein specialized agents collaboratively process extensive product information. This architecture distributes cognitive load to alleviate single-agent attention bottlenecks and captures critical decision factors through structured dialogue. Experiments demonstrate that our framework achieves significant improvements in product selection accuracy, purchase quantity prediction, and simulation stability compared to existing economic and financial LLM simulation baselines. Our results substantiate the potential of large language models as a foundational pillar for high-fidelity, scalable decision simulation and latter analysis in the real economy based on foundational database.", "authors": ["Yusen Wu", "Yiran Liu", "Xiaotie Deng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-18", "url": "https://arxiv.org/abs/2603.17694", "pdf_url": "https://arxiv.org/pdf/2603.17694v1", "arxiv_id": "2603.17694", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3281}
{"id": "05f22d8830f28a8bb458414d2760e8a66f7919adb47a25144a96cb9b694f685e", "sources": ["arxiv", "semantic_scholar"], "title": "The Verifier Tax: Horizon Dependent Safety Success Tradeoffs in Tool Using LLM Agents", "abstract": "We study how runtime enforcement against unsafe actions affects end-to-end task performance in multi-step tool using large language model (LLM) agents. Using tau-bench across Airline and Retail domains, we compare baseline Tool-Calling, planning-integrated (TRIAD), and policy-mediated (TRIAD-SAFETY) architectures with GPT-OSS-20B and GLM-4-9B. We identify model dependent interaction horizons (15 to 30 turns) and decompose outcomes into overall success rate (SR), safe success rate (SSR), and unsafe success rate (USR). Our results reveal a persistent Safety Capability Gap. While safety mediation can intercept up to 94 percent of non-compliant actions, it rarely translates into strictly safe goal attainment (SSR below 5 percent in most settings). We find that high unsafe success rates are primarily driven by Integrity Leaks, where models hallucinate user identifiers to bypass mandatory authentication. Recovery rates following blocked actions are consistently low, ranging from 21 percent for GPT-OSS-20B in simpler procedural tasks to near zero in complex Retail scenarios. These results demonstrate that runtime enforcement imposes a significant verifier tax on conversational length and compute cost without guaranteeing safe completion, highlighting the critical need for agents capable of grounded identity verification and post-intervention reasoning.", "authors": ["Tanmay Sah", "Vishal Srivastava", "Dolly Sah", "Kayden Jordan"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-18", "url": "https://arxiv.org/abs/2603.19328", "pdf_url": "https://arxiv.org/pdf/2603.19328v1", "arxiv_id": "2603.19328", "doi": "10.1145/3786335.3813160", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3281}
{"id": "0b22eca3ccd28b87e757b1cd68348ff71551e73a5a952944d432554822c345fd", "sources": ["arxiv", "semantic_scholar"], "title": "VeriAgent: A Tool-Integrated Multi-Agent System with Evolving Memory for PPA-Aware RTL Code Generation", "abstract": "LLMs have recently demonstrated strong capabilities in automatic RTL code generation, achieving high syntactic and functional correctness. However, most methods focus on functional correctness while overlooking critical physical design objectives, including Power, Performance, and Area. In this work, we propose a PPA-aware, tool-integrated multi-agent framework for high-quality verilog code generation. Our framework explicitly incorporates EDA tools into a closed-loop workflow composed of a \\textit{Programmer Agent}, a \\textit{Correctness Agent}, and a \\textit{PPA Agent}, enabling joint optimization of functional correctness and physical metrics. To support continuous improvement without model retraining, we introduce an \\textit{Evolved Memory Mechanism} that externalizes optimization experience into structured memory nodes. A dedicated memory manager dynamically maintains the memory pool and allows the system to refine strategies based on historical execution trajectories. Extensive experiments demonstrate that our approach achieves strong functional correctness while delivering significant improvements in PPA metrics. By integrating tool-driven feedback with structured and evolvable memory, our framework transforms RTL generation from one-shot reasoning into a continual, feedback-driven optimization process, providing a scalable pathway for deploying LLMs in real-world hardware design flows.", "authors": ["Yaoxiang Wang", "Qi Shi", "ShangZhan Li", "Qingguo Hu", "Xinyu Yin", "Bo Guo", "Xu Han", "Maosong Sun", "Jinsong Su"], "categories": ["cs.CL", "cs.PL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-18", "url": "https://arxiv.org/abs/2603.17613", "pdf_url": "https://arxiv.org/pdf/2603.17613v1", "arxiv_id": "2603.17613", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3281}
{"id": "9e1f23d01737c1781c00575d6c2cedfc2725fbf46af9f0a3d67f3ba972c18b10", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent System for Building-Age Cohort Mapping to Support Urban Energy Planning", "abstract": "Determining the age distribution of the urban building stock is crucial for sustainable municipal heat planning and upgrade prioritization. However, existing approaches often rely on datasets gathered via sensors or remote sensing techniques, leaving inconsistencies and gaps in data. We present a multi-agent LLM system comprising three key agents, the Zensus agent, the OSM agent, and the Monument agent, that fuse data from heterogeneous sources. A data orchestrator and harmonizer geocodes and deduplicates building imprints. Using this fused ground truth, we introduce BuildingAgeCNN, a satellite-only classifier based on a ConvNeXt backbone augmented with a Feature Pyramid Network (FPN), CoordConv spatial channels, and Squeeze-and-Excitation (SE) blocks. Under spatial cross validation, BuildingAgeCNN attains an overall accuracy of 90.69% but a modest macro-F1 of 67.25%, reflecting strong class imbalance and persistent confusions between adjacent historical cohorts. To mitigate risk for planning applications, the address-to prediction pipeline includes calibrated confidence estimates and flags low-confidence cases for manual review. This multi-agent LLM system not only assists in gathering structured data but also helps energy demand planners optimize district-heating networks and target low-carbon sustainable energy systems.", "authors": ["Kundan Thota", "Thorsten Schlachter", "Veit Hagenmeyer"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-18", "url": "https://arxiv.org/abs/2603.17626", "pdf_url": "https://arxiv.org/pdf/2603.17626v1", "arxiv_id": "2603.17626", "doi": "10.1109/SusTech67720.2026.11536317", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Conference on Technologies for Sustainability", "quality_score": 0.5156}
{"id": "56317faa77f1ca9ad7d69822c807418b0eb8fd117fd0d3d679a0d039d96e0cd3", "sources": ["arxiv", "semantic_scholar"], "title": "Adaptive Theory of Mind for LLM-based Multi-Agent Coordination", "abstract": "Theory of Mind (ToM) refers to the ability to reason about others' mental states, and higher-order ToM involves considering that others also possess their own ToM. Equipping large language model (LLM)-driven agents with ToM has long been considered to improve their coordination in multiagent collaborative tasks. However, we find that misaligned ToM orders-mismatches in the depth of ToM reasoning between agents-can lead to insufficient or excessive reasoning about others, thereby impairing their coordination. To address this issue, we design an adaptive ToM (A-ToM) agent, which can align in ToM orders with its partner. Based on prior interactions, the agent estimates the partner's likely ToM order and leverages this estimation to predict the partner's action, thereby facilitating behavioral coordination. We conduct empirical evaluations on four multi-agent coordination tasks: a repeated matrix game, two grid navigation tasks and an Overcooked task. The results validate our findings on ToM alignment and demonstrate the effectiveness of our A-ToM agent. Furthermore, we discuss the generalizability of our A-ToM to non-LLM-based agents, as well as what would diminish the importance of ToM alignment.", "authors": ["Chunjiang Mu", "Ya Zeng", "Qiaosheng Zhang", "Kun Shao", "Chen Chu", "Hao Guo", "Danyang Jia", "Zhen Wang", "Shuyue Hu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-17", "url": "https://arxiv.org/abs/2603.16264", "pdf_url": "https://arxiv.org/pdf/2603.16264v1", "arxiv_id": "2603.16264", "doi": "10.1609/aaai.v40i35.40204", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.5145}
{"id": "74cbd2948c36a9935de90381fbe06546bcb24d1f51b6dee6e987e94d6ecaa452", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Present: Inverse Specification Rewards for Agentic Slide Generation", "abstract": "Automated presentation generation remains a challenging task requiring coherent content creation, visual design, and audience-aware communication. This work proposes an OpenEnv-compatible reinforcement learning environment where LLM agents learn to research topics, plan content, and generate professional HTML slide presentations through tool use. We introduce a multi-component reward system combining structural validation, render quality assessment, LLM-based aesthetic scoring, content quality metrics, and an inverse specification reward that measures how faithfully generated slides convey their intended purpose. The inverse specification reward, an \"inverse task\" where an LLM attempts to recover the original specification from generated slides, provides a holistic quality signal. Our approach fine-tunes Qwen2.5-Coder-7B via GRPO, training only 0.5% of parameters on prompts derived from expert demonstrations collected using Claude Opus 4.6. Experiments on 48 diverse business briefs across six models demonstrate that our fine-tuned 7B model achieves 91.2% of Claude Opus 4.6's quality while improving 33.1% over the base model. The six-model comparison reveals that instruction adherence and tool-use compliance, rather than raw parameter count, determine agentic task performance. We contribute SlideRL, an open-source dataset of 288 multi-turn rollout trajectories across all six models: https://huggingface.co/datasets/KarthikRagunathAnandaKumar/sliderl-multi-turn-rollouts Code: https://github.com/pushing-the-frontier/slide-forge-llm", "authors": ["Karthik Ragunath Ananda Kumar", "Subrahmanyam Arunachalam"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-17", "url": "https://arxiv.org/abs/2603.16839", "pdf_url": "https://arxiv.org/pdf/2603.16839v1", "arxiv_id": "2603.16839", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/pushing-the-frontier/slide-forge-llm", "venue": null, "quality_score": 0.608}
{"id": "1dbaaec1b2645f4dc3ef58cdb1499aabb777037e08b2e67f3743d5a9d65732cc", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Communication Between Heterogeneous Agents in Multi-Agent Reinforcement Learning for Autonomous Cyber Defence", "abstract": "Reinforcement learning techniques are being explored as solutions to the threat of cyber attacks on enterprise networks. Recent research in the field of AI in cyber security has investigated the ability of homogeneous multi-agent reinforcement learning agents, capable of inter-agent communication, to respond to cyberattacks. This paper advances the study of learned communication in multi-agent systems by examining heterogeneous agent capabilities within a simulated network environment. To this end, we leverage CommFormer, a publicly available state-of-the-art communication algorithm, to train and evaluate agents within the Cyber Operations Research Gym (CybORG). Our results show that CommFormer agents with heterogeneous capabilities can outperform other algorithms deployed in the CybORG environment, by converging to an optimal policy up to four times faster while improving standard error by up 38%. The agents implemented in this project provide an additional avenue for exploration in the field of AI for cyber security, enabling further research involving realistic networks.", "authors": ["Alex Popa", "Adrian Taylor", "Ranwa Al Mallah"], "categories": ["cs.CR", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-17", "url": "https://arxiv.org/abs/2603.20279", "pdf_url": "https://arxiv.org/pdf/2603.20279v1", "arxiv_id": "2603.20279", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Poly-AIvsAI/CyMARL-CommFormer/tree/main", "venue": null, "quality_score": 0.608}
{"id": "0cf991538a8317ac1cfef10f8c7cd3a2f19cf001bfe0901cb93807d6f358aabf", "sources": ["arxiv", "semantic_scholar"], "title": "Interference-Aware K-Step Reachable Communication in Multi-Agent Reinforcement Learning", "abstract": "Effective communication is pivotal for addressing complex collaborative tasks in multi-agent reinforcement learning (MARL). Yet, limited communication bandwidth and dynamic, intricate environmental topologies present significant challenges in identifying high-value communication partners. Agents must consequently select collaborators under uncertainty, lacking a priori knowledge of which partners can deliver task-critical information. To this end, we propose Interference-Aware K-Step Reachable Communication (IA-KRC), a novel framework that enhances cooperation via two core components: (1) a K-Step reachability protocol that confines message passing to physically accessible neighbors, and (2) an interference-prediction module that optimizes partner choice by minimizing interference while maximizing utility. Compared to existing methods, IA-KRC enables substantially more persistent and efficient cooperation despite environmental interference. Comprehensive evaluations confirm that IA-KRC achieves superior performance compared to state-of-the-art baselines, while demonstrating enhanced robustness and scalability in complex topological and highly dynamic multi-agent scenarios.", "authors": ["Ziyu Cheng", "Jinsheng Ren", "Zhouxian Jiang", "Chenzhihang Li", "Rongye Shi", "Bin Liang", "Jun Yang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15054", "pdf_url": "https://arxiv.org/pdf/2603.15054v1", "arxiv_id": "2603.15054", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3267}
{"id": "05fea137ed7580e20e16ade81829a1af9668b7fba89c5c90214a6f629f89bf8b", "sources": ["arxiv", "semantic_scholar"], "title": "MAC: Multi-Agent Constitution Learning", "abstract": "Constitutional AI is a method to oversee and control LLMs based on a set of rules written in natural language. These rules are typically written by human experts, but could in principle be learned automatically given sufficient training data for the desired behavior. Existing LLM-based prompt optimizers attempt this but are ineffective at learning constitutions since (i) they require many labeled examples and (ii) lack structure in the optimized prompts, leading to diminishing improvements as prompt size grows. To address these limitations, we propose Multi-Agent Constitutional Learning (MAC), which optimizes over structured prompts represented as sets of rules using a network of agents with specialized tasks to accept, edit, or reject rule updates. We also present MAC+, which improves performance by training agents on successful trajectories to reinforce updates leading to higher reward. We evaluate MAC on tagging Personally Identifiable Information (PII), a classification task with limited labels where interpretability is critical, and demonstrate that it generalizes to other agentic tasks such as tool calling. MAC outperforms recent prompt optimization methods by over 50%, produces human-readable and auditable rule sets, and achieves performance comparable to supervised fine-tuning and GRPO without requiring parameter updates.", "authors": ["Rushil Thareja", "Gautam Gupta", "Francesco Pinto", "Nils Lukas"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15968", "pdf_url": "https://arxiv.org/pdf/2603.15968v1", "arxiv_id": "2603.15968", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/rushil-thareja/MAC-Multi-Agent-Constitution-Learning", "venue": null, "quality_score": 0.6067}
{"id": "c6a2abb06e042b9bf58406cb6d20da0c2ac2bcc3fc516c675084e1dc3932d1b1", "sources": ["arxiv", "semantic_scholar"], "title": "Token Coherence: Adapting MESI Cache Protocols to Minimize Synchronization Overhead in Multi-Agent LLM Systems", "abstract": "Multi-agent LLM orchestration incurs synchronization costs scaling as O(n x S x |D|) in agents, steps, and artifact size under naive broadcast -- a regime I term broadcast-induced triply-multiplicative overhead. I argue this pathology is a structural residue of full-state rebroadcast, not an inherent property of multi-agent coordination. The central claim: synchronization cost explosion in LLM multi-agent systems maps with formal precision onto the cache coherence problem in shared-memory multiprocessors, and MESI-protocol invalidation transfers to artifact synchronization under minimal structural modification. I construct the Artifact Coherence System (ACS) and prove the Token Coherence Theorem: lazy invalidation attenuates cost by at least S/(n + W(d_i)) when S > n + W(d_i), converting O(n x S x |D|) to O((n + W) x |D|). A TLA+-verified protocol enforces single-writer safety, monotonic versioning, and bounded staleness across ~2,400 explored states. Simulation across four workload configurations yields token savings of 95.0% +/- 1.3% at V=0.05, 92.3% +/- 1.4% at V=0.10, 88.3% +/- 1.5% at V=0.25, and 84.2% +/- 1.3% at V=0.50 -- each exceeding the theorem's conservative lower bounds. Savings of ~81% persist at V=0.9, contrary to the predicted collapse threshold. Contributions: (1) formal MESI-to-artifact state mapping; (2) Token Coherence Theorem as savings lower bound; (3) TLA+-verified protocol with three proven invariants; (4) characterization of conditional artifact access semantics resolving the always-read objection; (5) reference Python implementation integrating with LangGraph, CrewAI, and AutoGen via thin adapter layers.", "authors": ["Vladyslav Parakhin"], "categories": ["cs.DC", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15183", "pdf_url": "https://arxiv.org/pdf/2603.15183v1", "arxiv_id": "2603.15183", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/hipvlady/agent-coherence", "venue": null, "quality_score": 0.6067}
{"id": "91911a92c6c7f41822ddd8f8bf10d18e5befb5574549806f968edf4eda4f1d00", "sources": ["arxiv", "semantic_scholar"], "title": "Brain-Inspired Graph Multi-Agent Systems for LLM Reasoning", "abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities across a wide range of language tasks, yet complex multi-step reasoning remains a fundamental challenge. While Large Reasoning Models (LRMs) equipped with extended chain-of-thought mechanisms demonstrate improved performance over standard LLMs, both model types still suffer from accuracy collapse on sufficiently complex tasks, suggesting that scaling model-level reasoning alone is insufficient. Inspired by the global workspace theory of human cognition, we propose Brain-Inspired Graph Multi-Agent Systems (BIGMAS), in which specialized LLM agents are organized as nodes in a dynamically constructed directed graph and coordinate exclusively through a centralized shared workspace. A problem-adaptive GraphDesigner constructs task-specific agent topologies, while a global Orchestrator leverages the complete shared state for routing decisions, overcoming the local-view bottleneck of reactive approaches. Experiments on Game24, Six Fives, and Tower of London across six frontier LLMs demonstrate that BIGMAS consistently improves reasoning performance for both standard LLMs and LRMs, outperforming existing multi-agent baselines including ReAct and Tree of Thoughts, showing that multi-agent architectural design provides complementary gains orthogonal to model-level reasoning enhancements.", "authors": ["Guangfu Hao", "Yuming Dai", "Xianzhe Qin", "Shan Yu"], "categories": ["cs.AI", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15371", "pdf_url": "https://arxiv.org/pdf/2603.15371v1", "arxiv_id": "2603.15371", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3267}
{"id": "d778de51750ee8bca3ec65d6acd34851d1c6c6259f6debe729ae2e69bd8d7336", "sources": ["arxiv", "semantic_scholar"], "title": "SAGE: Multi-Agent Self-Evolution for LLM Reasoning", "abstract": "Reinforcement learning with verifiable rewards improves reasoning in large language models (LLMs), but many methods still rely on large human-labeled datasets. While self-play reduces this dependency, it often lacks explicit planning and strong quality control, limiting stability in long-horizon multi-step reasoning. We present SAGE (Self-evolving Agents for Generalized reasoning Evolution), a closed-loop framework where four agents: Challenger, Planner, Solver, and Critic, co-evolve from a shared LLM backbone using only a small seed set. The Challenger continuously generates increasingly difficult tasks; the Planner converts each task into a structured multi-step plan; and the Solver follows the plan to produce an answer, whose correctness is determined by external verifiers. The Critic scores and filters both generated questions and plans to prevent curriculum drift and maintain training signal quality, enabling stable self-training. Across mathematics and code-generation benchmarks, SAGE delivers consistent gains across model scales, improving the Qwen-2.5-7B model by 8.9% on LiveCodeBench and 10.7% on OlympiadBench.", "authors": ["Yulin Peng", "Xinxin Zhu", "Chenxing Wei", "Nianbo Zeng", "Leilei Wang", "Ying Tiffany He", "F. Richard Yu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15255", "pdf_url": "https://arxiv.org/pdf/2603.15255v2", "arxiv_id": "2603.15255", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3267}
{"id": "a6dba5d754bf15e6965ec8dd411b467487e9dd81f3c636b47bac331f7db7771b", "sources": ["arxiv", "semantic_scholar"], "title": "PMAx: An Agentic Framework for AI-Driven Process Mining", "abstract": "Process mining provides powerful insights into organizational workflows, but extracting these insights typically requires expertise in specialized query languages and data science tools. Large Language Models (LLMs) offer the potential to democratize process mining by enabling business users to interact with process data through natural language. However, using LLMs as direct analytical engines over raw event logs introduces fundamental challenges: LLMs struggle with deterministic reasoning and may hallucinate metrics, while sending large, sensitive logs to external AI services raises serious data-privacy concerns. To address these limitations, we present PMAx, an autonomous agentic framework that functions as a virtual process analyst. Rather than relying on LLMs to generate process models or compute analytical results, PMAx employs a privacy-preserving multi-agent architecture. An Engineer agent analyzes event-log metadata and autonomously generates local scripts to run established process mining algorithms, compute exact metrics, and produce artifacts such as process models, summary tables, and visualizations. An Analyst agent then interprets these insights and artifacts to compile comprehensive reports. By separating computation from interpretation and executing analysis locally, PMAx ensures mathematical accuracy and data privacy while enabling non-technical users to transform high-level business questions into reliable process insights.", "authors": ["Anton Antonov", "Humam Kourani", "Alessandro Berti", "Gyunam Park", "Wil M. P. van der Aalst"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15351", "pdf_url": "https://arxiv.org/pdf/2603.15351v1", "arxiv_id": "2603.15351", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3267}
{"id": "dee85b954c516033234a60777e5821a0e8550a3fea010e07f910e663b9fc4e1c", "sources": ["arxiv", "semantic_scholar"], "title": "Intelligent Co-Design: An Interactive LLM Framework for Interior Spatial Design via Multi-Modal Agents", "abstract": "In architectural interior design, miscommunication frequently arises as clients lack design knowledge, while designers struggle to explain complex spatial relationships, leading to delayed timelines and financial losses. Recent advancements in generative layout tools narrow the gap by automating 3D visualizations. However, prevailing methodologies exhibit limitations: rule-based systems implement hard-coded spatial constraints that restrict participatory engagement, while data-driven models rely on extensive training datasets. Recent large language models (LLMs) bridge this gap by enabling intuitive reasoning about spatial relationships through natural language. This research presents an LLM-based, multimodal, multi-agent framework that dynamically converts natural language descriptions and imagery into 3D designs. Specialized agents (Reference, Spatial, Interactive, Grader), operating via prompt guidelines, collaboratively address core challenges: the agent system enables real-time user interaction for iterative spatial refinement, while Retrieval-Augmented Generation (RAG) reduces data dependency without requiring task-specific model training. This framework accurately interprets spatial intent and generates optimized 3D indoor design, improving productivity, and encouraging nondesigner participation. Evaluations across diverse floor plans and user questionnaires demonstrate effectiveness. An independent LLM evaluator consistently rated participatory layouts higher in user intent alignment, aesthetic coherence, functionality, and circulation. Questionnaire results indicated 77% satisfaction and a clear preference over traditional design software. These findings suggest the framework enhances user-centric communication and fosters more inclusive, effective, and resilient design processes. Project page: https://rsigktyper.github.io/AICodesign/", "authors": ["Ren Jian Lim", "Rushi Dai"], "categories": ["cs.AI", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-16", "url": "https://arxiv.org/abs/2603.15341", "pdf_url": "https://arxiv.org/pdf/2603.15341v1", "arxiv_id": "2603.15341", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6067}
{"id": "9ca0fedd7abbe1a455ae39ecce0fddc47fbc64766208aa544c794153cb815bc8", "sources": ["arxiv", "semantic_scholar"], "title": "Governing Dynamic Capabilities: Cryptographic Binding and Reproducibility Verification for AI Agent Tool Use", "abstract": "AI agents dynamically acquire tools, orchestrate sub-agents, and transact across organizational boundaries, yet no existing security layer verifies what an agent can do, whether it executed what it claims, or what happened in a multi-agent interaction. We trace this gap to the capability-context separation: inside a transformer, tool definitions and user context are indistinguishable tokens, but at the orchestration layer they have fundamentally different security semantics. Existing frameworks conflate the two, enabling silent capability escalation and leaving interactions without verifiable provenance. From this principle we derive three Agent Governance Requirements: capability integrity (G1), behavioral verifiability (G2), and interaction auditability (G3), defining what a governed agent ecosystem must enforce, independent of how. We prove two structural results: the Chain Verifiability Theorem (one unverifiable interior agent breaks end-to-end verification for all downstream nodes) and the Bounded Divergence Theorem (replay-based verification yields a probabilistic safety certificate, epsilon <= 1 - alpha^{1/n}). We validate with two crypto-agnostic instantiations -- basic (Ed25519, SHA-256; 97 us verify) and enhanced (BBS+ selective disclosure, Groth16 DV-SNARK; 13.8 ms) -- both satisfying nine security properties. A reproducibility study (9 models, 7 providers) reveals 5.8x variance in inference determinism, connecting model characteristics to governance architecture. End-to-end evaluation over 5-20 agent pipelines confirms <0.02% overhead and detection of all attack scenarios with zero false positives.", "authors": ["Ziling Zhou"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-15", "url": "https://arxiv.org/abs/2603.14332", "pdf_url": "https://arxiv.org/pdf/2603.14332v2", "arxiv_id": "2603.14332", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3259}
{"id": "3cfbbdb641eb4d01a90584097f273600a7a156a7c0cca72fbb1088bc9a67cf9e", "sources": ["arxiv", "semantic_scholar"], "title": "LegacyTranslate: LLM-based Multi-Agent Method for Legacy Code Translation", "abstract": "Modernizing large legacy systems remains a major challenge in enterprise environments, particularly when migration must preserve domain-specific logic while conforming to internal architectural frameworks and shared APIs. Direct application of Large Language Models (LLMs) for code translation often produces syntactically valid outputs that fail to compile or integrate within existing production frameworks, limiting their practical adoption in real-world modernization efforts. In this paper, we propose LegacyTranslate, a multi-agent framework for API-aware code translation, developed and evaluated in the context of an ongoing modernization effort at a financial institution migrating approximately 2.5 million lines of PL/SQL to Java. The core idea is to use specialized LLM-based agents, each addressing a different aspect of the translation challenge. Specifically, LegacyTranslate consists of three agents: Initial Translation Agent produces an initial Java translation using retrieved in-context examples; API Grounding Agent aligns the code with existing APIs by retrieving relevant entries from an API knowledge base; and Refinement Agent iteratively refines the output using compiler feedback and API suggestions to improve correctness. Our experiments show that each agent contributes to better translation quality. The Initial Translation Agent alone achieves 45.6% compilable outputs and 30.9% test-pass rate. With API Grounding Agent and Refinement Agent, compilation improves by an additional 8% and test-pass accuracy increases by 3%.", "authors": ["Zahra Moti", "Heydar Soudani", "Jonck van der Kogel"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-14", "url": "https://arxiv.org/abs/2603.14054", "pdf_url": "https://arxiv.org/pdf/2603.14054v1", "arxiv_id": "2603.14054", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3252}
{"id": "e494ba99e6a96de2c236ab9456b2b49d1f933d7a1f00f431c35c3fff59054f7a", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Constitutional Multi-Agent Governance", "abstract": "Large Language Models (LLMs) can generate persuasive influence strategies that shift cooperative behavior in multi-agent populations, but a critical question remains: does the resulting cooperation reflect genuine prosocial alignment, or does it mask erosion of agent autonomy, epistemic integrity, and distributional fairness? We introduce Constitutional Multi-Agent Governance (CMAG), a two-stage framework that interposes between an LLM policy compiler and a networked agent population, combining hard constraint filtering with soft penalized-utility optimization that balances cooperation potential against manipulation risk and autonomy pressure. We propose the Ethical Cooperation Score (ECS), a multiplicative composite of cooperation, autonomy, integrity, and fairness that penalizes cooperation achieved through manipulative means. In experiments on scale-free networks of 80 agents under adversarial conditions (70% violating candidates), we benchmark three regimes: full CMAG, naive filtering, and unconstrained optimization. While unconstrained optimization achieves the highest raw cooperation (0.873), it yields the lowest ECS (0.645) due to severe autonomy erosion (0.867) and fairness degradation (0.888). CMAG attains an ECS of 0.741, a 14.9% improvement, while preserving autonomy at 0.985 and integrity at 0.995, with only modest cooperation reduction to 0.770. The naive ablation (ECS = 0.733) confirms that hard constraints alone are insufficient. Pareto analysis shows CMAG dominates the cooperation-autonomy trade-off space, and governance reduces hub-periphery exposure disparities by over 60%. These findings establish that cooperation is not inherently desirable without governance: constitutional constraints are necessary to ensure that LLM-mediated influence produces ethically stable outcomes rather than manipulative equilibria.", "authors": ["J. de Curtò", "I. de Zarzà"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.13189", "pdf_url": "https://arxiv.org/pdf/2603.13189v1", "arxiv_id": "2603.13189", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "d9e35b95efb1ae783c7351e4d0a3fa81bd3fc2463ef3f50ed229614642fadece", "sources": ["arxiv", "semantic_scholar"], "title": "ChainFuzzer: Greybox Fuzzing for Workflow-Level Multi-Tool Vulnerabilities in LLM Agents", "abstract": "Tool-augmented LLM agents increasingly rely on multi-step, multi-tool workflows to complete real tasks. This design expands the attack surface, because data produced by one tool can be persisted and later reused as input to another tool, enabling exploitable source-to-sink dataflows that only emerge through tool composition. We study this risk as multi-tool vulnerabilities in LLM agents, and show that existing discovery efforts focused on single-tool or single-hop testing miss these long-horizon behaviors and provide limited debugging value. We present ChainFuzzer, a greybox framework for discovering and reproducing multi-tool vulnerabilities with auditable evidence. ChainFuzzer (i) identifies high-impact operations with strict source-to-sink dataflow evidence and extracts plausible upstream candidate tool chains based on cross-tool dependencies, (ii) uses Trace-guided Prompt Solving (TPS) to synthesize stable prompts that reliably drive the agent to execute target chains, and (iii) performs guardrail-aware fuzzing to reproduce vulnerabilities under LLM guardrails via payload mutation and sink-specific oracles. We evaluate ChainFuzzer on 20 popular open-source LLM agent apps (998 tools). ChainFuzzer extracts 2,388 candidate tool chains and synthesizes 2,213 stable prompts, confirming 365 unique, reproducible vulnerabilities across 19/20 apps (302 require multi-tool execution). Component evaluation shows tool-chain extraction achieves 96.49% edge precision and 91.50% strict chain precision; TPS increases chain reachability from 27.05% to 95.45%; guardrail-aware fuzzing boosts payload-level trigger rate from 18.20% to 88.60%. Overall, ChainFuzzer achieves 3.02 vulnerabilities per 1M tokens, providing a practical foundation for testing and hardening real-world multi-tool agent systems.", "authors": ["Jiangrong Wu", "Zitong Yao", "Yuhong Nan", "Zibin Zheng"], "categories": ["cs.SE", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.12614", "pdf_url": "https://arxiv.org/pdf/2603.12614v1", "arxiv_id": "2603.12614", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.6026}
{"id": "e671acddb131da7200f5c792ff23253e8618ece552b9fd9d78963713419f5bff", "sources": ["arxiv", "semantic_scholar"], "title": "Semantic Invariance in Agentic AI", "abstract": "Large Language Models (LLMs) increasingly serve as autonomous reasoning agents in decision support, scientific problem-solving, and multi-agent coordination systems. However, deploying LLM agents in consequential applications requires assurance that their reasoning remains stable under semantically equivalent input variations, a property we term semantic invariance. Standard benchmark evaluations, which assess accuracy on fixed, canonical problem formulations, fail to capture this critical reliability dimension. To address this shortcoming, in this paper we present a metamorphic testing framework for systematically assessing the robustness of LLM reasoning agents, applying eight semantic-preserving transformations (identity, paraphrase, fact reordering, expansion, contraction, academic context, business context, and contrastive formulation) across seven foundation models spanning four distinct architectural families: Hermes (70B, 405B), Qwen3 (30B-A3B, 235B-A22B), DeepSeek-R1, and gpt-oss (20B, 120B). Our evaluation encompasses 19 multi-step reasoning problems across eight scientific domains. The results reveal that model scale does not predict robustness: the smaller Qwen3-30B-A3B achieves the highest stability (79.6% invariant responses, semantic similarity 0.91), while larger models exhibit greater fragility.", "authors": ["I. de Zarzà", "J. de Curtò", "Jordi Cabot", "Pietro Manzoni", "Carlos T. Calafate"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.13173", "pdf_url": "https://arxiv.org/pdf/2603.13173v2", "arxiv_id": "2603.13173", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "25ae213b30dfca876ae8213bf6756df88d2e785257983248ac3c8f87e52c0a15", "sources": ["arxiv", "semantic_scholar"], "title": "ToolTree: Efficient LLM Agent Tool Planning via Dual-Feedback Monte Carlo Tree Search and Bidirectional Pruning", "abstract": "Large Language Model (LLM) agents are increasingly applied to complex, multi-step tasks that require interaction with diverse external tools across various domains. However, current LLM agent tool planning methods typically rely on greedy, reactive tool selection strategies that lack foresight and fail to account for inter-tool dependencies. In this paper, we present ToolTree, a novel Monte Carlo tree search-inspired planning paradigm for tool planning. ToolTree explores possible tool usage trajectories using a dual-stage LLM evaluation and bidirectional pruning mechanism that enables the agent to make informed, adaptive decisions over extended tool-use sequences while pruning less promising branches before and after the tool execution. Empirical evaluations across both open-set and closed-set tool planning tasks on 4 benchmarks demonstrate that ToolTree consistently improves performance while keeping the highest efficiency, achieving an average gain of around 10\\% compared to the state-of-the-art planning paradigm.", "authors": ["Shuo Yang", "Soyeon Caren Han", "Yihao Ding", "Shuhe Wang", "Eduard Hoy"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.12740", "pdf_url": "https://arxiv.org/pdf/2603.12740v1", "arxiv_id": "2603.12740", "doi": null, "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "07abf7f30540ddd4ef969a7baaff7fee89bca07852b01997fb6f8dd5ba6295a0", "sources": ["arxiv", "semantic_scholar"], "title": "HR-Agents: Using Multiple LLM-based Agents to Improve Q&A about Brazilian Labor Legislation", "abstract": "The Consolidation of Labor Laws (CLT) serves as the primary legal framework governing labor relations in Brazil, ensuring essential protections for workers. However, its complexity creates challenges for Human Resources (HR) professionals in navigating regulations and ensuring compliance. Traditional methods for addressing labor law inquiries often lead to inefficiencies, delays, and inconsistencies. To enhance the accuracy and efficiency of legal question-answering (Q&A), a multi-agent system powered by Large Language Models (LLMs) is introduced. This approach employs specialized agents to address distinct aspects of employment law while integrating Retrieval-Augmented Generation (RAG) to enhance contextual relevance. Implemented using CrewAI, the system enables cooperative agent interactions, ensuring response validation and reducing misinformation. The effectiveness of this framework is evaluated through a comparison with a baseline RAG pipeline utilizing a single LLM, using automated metrics such as BLEU, LLM-as-judge evaluations, and expert human assessments. Results indicate that the multi-agent approach improves response coherence and correctness, providing a more reliable and efficient solution for HR professionals. This study contributes to AI-driven legal assistance by demonstrating the potential of multi-agent LLM architectures in improving labor law compliance and streamlining HR operations.", "authors": ["Abriel K. Moraes", "Gabriel S. M. Dias", "Vitor L. Fabris", "Lucas D. Gessoni", "Leonardo R. do Nascimento", "Charles S. Oliveira", "Vitor G. C. B. de Farias", "Fabiana C. Q. de O. Marucci", "Matheus H. R. Vicente", "Gabriel U. Talasso", "Erik Soares", "Amparo Munoz", "Sildolfo Gomes", "Maria L. A. de S. Cruvinel", "Leonardo T. dos Santos", "Renata De Paris", "Wandemberg Gibaut"], "categories": ["cs.IR", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2604.16337", "pdf_url": "https://arxiv.org/pdf/2604.16337v1", "arxiv_id": "2604.16337", "doi": "10.29327/1842969.1-308", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "d0f053627a8e9b8ae5a814ec716aa8f6af448d59a44242eb45324b4edaa48798", "sources": ["arxiv", "semantic_scholar"], "title": "Efficient and Interpretable Multi-Agent LLM Routing via Ant Colony Optimization", "abstract": "Large Language Model (LLM)-driven Multi-Agent Systems (MAS) have demonstrated strong capability in complex reasoning and tool use, and heterogeneous agent pools further broaden the quality--cost trade-off space. Despite these advances, real-world deployment is often constrained by high inference cost, latency, and limited transparency, which hinders scalable and efficient routing. Existing routing strategies typically rely on expensive LLM-based selectors or static policies, and offer limited controllability for semantic-aware routing under dynamic loads and mixed intents, often resulting in unstable performance and inefficient resource utilization. To address these limitations, we propose AMRO-S, an efficient and interpretable routing framework for Multi-Agent Systems (MAS). AMRO-S models MAS routing as a semantic-conditioned path selection problem, enhancing routing performance through three key mechanisms: First, it leverages a supervised fine-tuned (SFT) small language model for intent inference, providing a low-overhead semantic interface for each query; second, it decomposes routing memory into task-specific pheromone specialists, reducing cross-task interference and optimizing path selection under mixed workloads; finally, it employs a quality-gated asynchronous update mechanism to decouple inference from learning, optimizing routing without increasing latency. Extensive experiments on five public benchmarks and high-concurrency stress tests demonstrate that AMRO-S consistently improves the quality--cost trade-off over strong routing baselines, while providing traceable routing evidence through structured pheromone patterns.", "authors": ["Xudong Wang", "Chaoning Zhang", "Jiaquan Zhang", "Chenghao Li", "Qigan Sun", "Sung-Ho Bae", "Peng Wang", "Ning Xie", "Jie Zou", "Yang Yang", "Hengtao Shen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.12933", "pdf_url": "https://arxiv.org/pdf/2603.12933v1", "arxiv_id": "2603.12933", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "07974b8d77f372d310547084cb7f4de100c274e1f600e94bfa6e95bbd0285e01", "sources": ["arxiv", "semantic_scholar"], "title": "Orla: A Library for Serving LLM-Based Multi-Agent Systems", "abstract": "We introduce Orla, a library for constructing and running LLM-based agentic systems. Modern agentic applications consist of workflows that combine multiple LLM inference steps, tool calls, and heterogeneous infrastructure. Today, developers typically build these systems by manually composing orchestration code with LLM serving engines and tool execution logic. Orla provides a general abstraction that separates request execution from workflow-level policy. It acts as a serving layer above existing LLM inference engines: developers define workflows composed of stages, while Orla manages how those stages are mapped, executed, and coordinated across models and backends. It provides agent-level control through three mechanisms: a stage mapper, which assigns each stage to an appropriate model and backend; a workflow orchestrator, which schedules stages and manages their resources and context; and a memory manager, which manages inference state such as the KV cache across workflow boundaries. We demonstrate Orla with a customer support workflow that exercises many of its capabilities. We evaluate Orla on two datasets, showing that stage mapping improves latency and cost compared to a single-model vLLM baseline, while workflow-level cache management reduces time-to-first-token.", "authors": ["Rana Shahout", "Hayder Tirmazi", "Minlan Yu", "Michael Mitzenmacher"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-13", "url": "https://arxiv.org/abs/2603.13605", "pdf_url": "https://arxiv.org/pdf/2603.13605v1", "arxiv_id": "2603.13605", "doi": "10.1145/3786335.3813227", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3245}
{"id": "378b7cdb24578ca1dfc679747d91b86bbc24ae76f3bc60d7321e44c6a99f3bce", "sources": ["arxiv", "semantic_scholar"], "title": "From Control to Foresight: Simulation as a New Paradigm for Human-Agent Collaboration", "abstract": "Large Language Models (LLMs) are increasingly used to power autonomous agents for complex, multi-step tasks. However, human-agent interaction remains pointwise and reactive: users approve or correct individual actions to mitigate immediate risks, without visibility into subsequent consequences. This forces users to mentally simulate long-term effects, a cognitively demanding and often inaccurate process. Users have control over individual steps but lack the foresight to make informed decisions. We argue that effective collaboration requires foresight, not just control. We propose simulation-in-the-loop, an interaction paradigm that enables users and agents to explore simulated future trajectories before committing to decisions. Simulation transforms intervention from reactive guesswork into informed exploration, while helping users discover latent constraints and preferences along the way. This perspective paper characterizes the limitations of current paradigms, introduces a conceptual framework for simulation-based collaboration, and illustrates its potential through concrete human-agent collaboration scenarios.", "authors": ["Gaole He", "Brian Y. Lim"], "categories": ["cs.HC", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-12", "url": "https://arxiv.org/abs/2603.11677", "pdf_url": "https://arxiv.org/pdf/2603.11677v1", "arxiv_id": "2603.11677", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3237}
{"id": "351ec3d7ea71608367ade7c2679931d387a6c8921552ecf16b60e77571090c6f", "sources": ["arxiv", "semantic_scholar"], "title": "AGMARL-DKS: An Adaptive Graph-Enhanced Multi-Agent Reinforcement Learning for Dynamic Kubernetes Scheduling", "abstract": "State-of-the-art cloud-native applications require intelligent schedulers that can effectively balance system stability, resource utilisation, and associated costs. While Kubernetes provides feasibility-based placement by default, recent research efforts have explored the use of reinforcement learning (RL) for more intelligent scheduling decisions. However, current RL-based schedulers have three major limitations. First, most of these schedulers use monolithic centralised agents, which are non-scalable for large heterogeneous clusters. Second, the ones that use multi-objective reward functions assume simple, static, linear combinations of the objectives. Third, no previous work has produced a stress-aware scheduler that can react adaptively to dynamic conditions. To address these gaps in current research, we propose the Adaptive Graph-enhanced Multi-Agent Reinforcement Learning Dynamic Kubernetes Scheduler (AGMARL-DKS). AGMARL-DKS addresses these gaps by introducing three major innovations. First, we construct a scalable solution by treating the scheduling challenge as a cooperative multi-agent problem, where every cluster node operates as an agent, employing centralised training methods before decentralised execution. Second, to be context-aware and yet decentralised, we use a Graph Neural Network (GNN) to build a state representation of the global cluster context at each agent. This represents an improvement over methods that rely solely on local observations. Finally, to make trade-offs between these objectives, we use a stress-aware lexicographical ordering policy instead of a simple, static linear weighting of these objectives. The evaluations in Google Kubernetes Engine (GKE) reveal that AGMARL-DKS significantly outperforms the default scheduler in terms of fault tolerance, utilisation, and cost, especially in scheduling batch and mission-critical workloads.", "authors": ["Hamed Hamzeh"], "categories": ["cs.DC", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-12", "url": "https://arxiv.org/abs/2603.12031", "pdf_url": "https://arxiv.org/pdf/2603.12031v2", "arxiv_id": "2603.12031", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3237}
{"id": "488bf704f3de36a9f9dcac63eb642d8f1a592e3fa22cfcecbf998eb9ff6774f7", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Value Alignment of LLMs with Multi-agent system and Combinatorial Fusion", "abstract": "Aligning large language models (LLMs) with human values is a central challenge for ensuring trustworthy and safe deployment. While existing methods such as Reinforcement Learning from Human Feedback (RLHF) and its variants have improved alignment, they often rely on a single evaluator or narrowly defined reward signals, limiting their ability to capture ethical pluralism. In this work, we propose the Value Alignment System using Combinatorial Fusion Analysis (VAS-CFA), a framework that operationalizes multi-agent fusion alignment. It instantiates multiple moral agents, each fine-tuned to represent a distinct normative perspective, and fuses their outputs using CFA with both rank- and score-based aggregation. This design leverages cognitive diversity, between agents, to mitigate conflicts and redundancies across multiple agents, producing responses that better reflect human values. Empirical evaluation demonstrates that VAS-CFA outperforms both single agent baselines and prior aggregation approaches on standard metrics, showing that multi-agent fusion provides a robust and effective mechanism for advancing value alignment in LLMs.", "authors": ["Yuanhong Wu", "Djallel Bouneffouf", "D. Frank Hsu"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-11", "url": "https://arxiv.org/abs/2603.11126", "pdf_url": "https://arxiv.org/pdf/2603.11126v1", "arxiv_id": "2603.11126", "doi": "10.1109/icassp55912.2026.11460611", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Acoustics, Speech, and Signal Processing", "quality_score": 0.5076}
{"id": "0a7daad2a6de24add5bbb7365eccf8291c5fecda0ac297e03ae3b0e9348d8bdc", "sources": ["arxiv", "semantic_scholar"], "title": "Task-Aware Delegation Cues for LLM Agents", "abstract": "LLM agents increasingly present as conversational collaborators, yet human--agent teamwork remains brittle due to information asymmetry: users lack task-specific reliability cues, and agents rarely surface calibrated uncertainty or rationale. We propose a task-aware collaboration signaling layer that turns offline preference evaluations into online, user-facing primitives for delegation. Using Chatbot Arena pairwise comparisons, we induce an interpretable task taxonomy via semantic clustering, then derive (i) Capability Profiles as task-conditioned win-rate maps and (ii) Coordination-Risk Cues as task-conditioned disagreement (tie-rate) priors. These signals drive a closed-loop delegation protocol that supports common-ground verification, adaptive routing (primary vs.\\ primary+auditor), explicit rationale disclosure, and privacy-preserving accountability logs. Two predictive probes validate that task typing carries actionable structure: cluster features improve winner prediction accuracy and reduce difficulty prediction error under stratified 5-fold cross-validation. Overall, our framework reframes delegation from an opaque system default into a visible, negotiable, and auditable collaborative decision, providing a principled design space for adaptive human--agent collaboration grounded in mutual awareness and shared accountability.", "authors": ["Xingrui Gu"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-11", "url": "https://arxiv.org/abs/2603.11011", "pdf_url": "https://arxiv.org/pdf/2603.11011v1", "arxiv_id": "2603.11011", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "CHI 2026", "quality_score": 0.5076}
{"id": "f259926faca3c657070ef25b6b9376a513adf55b002e18e1726af2fbe7bcef5e", "sources": ["arxiv", "semantic_scholar"], "title": "WebWeaver: Breaking Topology Confidentiality in LLM Multi-Agent Systems with Stealthy Context-Based Inference", "abstract": "Communication topology is a critical factor in the utility and safety of LLM-based multi-agent systems (LLM-MAS), making it a high-value intellectual property (IP) whose confidentiality remains insufficiently studied. Existing topology inference attempts rely on impractical assumptions, including control over the administrative agent and direct identity queries via jailbreaks, which are easily defeated by basic keyword-based defenses. As a result, prior analyses fail to capture the real-world threat of such attacks. To bridge this realism gap, we propose \\textit{WebWeaver}, an attack framework that infers the complete LLM-MAS topology by compromising only a single arbitrary agent instead of the administrative agent. Unlike prior approaches, WebWeaver relies solely on agent contexts rather than agent IDs, enabling significantly stealthier inference. WebWeaver further introduces a new covert jailbreak-based mechanism and a novel fully jailbreak-free diffusion design to handle cases where jailbreaks fail. Additionally, we address a key challenge in diffusion-based inference by proposing a masking strategy that preserves known topology during diffusion, with theoretical guarantees of correctness. Extensive experiments show that WebWeaver substantially outperforms state-of-the-art (SOTA) baselines, achieving about 60\\% higher inference accuracy under active defenses with negligible overhead.", "authors": ["Zixun Xiong", "Gaoyi Wu", "Lingfeng Yao", "Miao Pan", "Xiaojiang Du", "Hao Wang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-11", "url": "https://arxiv.org/abs/2603.11132", "pdf_url": "https://arxiv.org/pdf/2603.11132v2", "arxiv_id": "2603.11132", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.323}
{"id": "35211d9af339a28c3df354b8028b79f00ad13dbac31d4e1d8ad8c8016b6b2a4f", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Negotiate: Multi-Agent Deliberation for Collective Value Alignment in LLMs", "abstract": "LLM alignment has progressed in single-agent settings through paradigms such as RL with human feedback (RLHF), while recent work explores scalable alternatives such as RL with AI feedback (RLAIF) and dynamic alignment objectives. However, these approaches remain limited in multi-stakeholder settings, where conflicting values arise and deliberative negotiation is required. This work proposes a multi-agent negotiation-based alignment framework that aligns LLMs to Collective Agency (CA)-an existing alignment objective introduced to promote the continual expansion of agency-while simultaneously improving conflict-resolution capability. To enable scalable training, two self-play LLM instances are assigned opposing personas and engage in turn-based dialogue to synthesize mutually beneficial solutions. We generate synthetic moral-dilemma prompts and conflicting persona pairs, and optimize the policy via RLAIF using Group Relative Policy Optimization (GRPO) with an external LLM reward model. While rewards are computed from CA scores assigned to the final completion, gradients are applied to dialogue tokens to directly improve deliberative interaction dynamics. Experiments show that the model achieves CA alignment comparable to a single-agent baseline while substantially improving conflict-resolution performance without degrading general language capabilities. These results suggest that negotiation-driven deliberation training provides a practical path toward LLMs that better support collective decision-making in value-conflict scenarios.", "authors": ["Panatchakorn Anantaprayoon", "Nataliia Babina", "Nima Asgharbeygi", "Jad Tarifi"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-11", "url": "https://arxiv.org/abs/2603.10476", "pdf_url": "https://arxiv.org/pdf/2603.10476v2", "arxiv_id": "2603.10476", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.323}
{"id": "3fbea62ffbb90edd2c01baa453cb41e37a67c6190de70c024842555add6e0657", "sources": ["arxiv", "semantic_scholar"], "title": "AttriGuard: Defeating Indirect Prompt Injection in LLM Agents via Causal Attribution of Tool Invocations", "abstract": "LLM agents are highly vulnerable to Indirect Prompt Injection (IPI), where adversaries embed malicious directives in untrusted tool outputs to hijack execution. Most existing defenses treat IPI as an input-level semantic discrimination problem, which often fails to generalize to unseen payloads. We propose a new paradigm, action-level causal attribution, which secures agents by asking why a particular tool call is produced. The central goal is to distinguish tool calls supported by the user's intent from those causally driven by untrusted observations. We instantiate this paradigm with AttriGuard, a runtime defense based on parallel counterfactual tests. For each proposed tool call, AttriGuard verifies its necessity by re-executing the agent under a control-attenuated view of external observations. Technically, AttriGuard combines teacher-forced shadow replay to prevent attribution confounding, hierarchical control attenuation to suppress diverse control channels while preserving task-relevant information, and a fuzzy survival criterion that is robust to LLM stochasticity. Across four LLMs and two agent benchmarks, AttriGuard achieves 0% ASR under static attacks with negligible utility loss and moderate overhead. Importantly, it remains resilient under adaptive optimization-based attacks in settings where leading defenses degrade significantly.", "authors": ["Yu He", "Haozhe Zhu", "Yiming Li", "Shuo Shao", "Hongwei Yao", "Zhihao Liu", "Zhan Qin"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-11", "url": "https://arxiv.org/abs/2603.10749", "pdf_url": "https://arxiv.org/pdf/2603.10749v2", "arxiv_id": "2603.10749", "doi": null, "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.323}
{"id": "8ee3a03ec75ff7be6f0a7948df2839eb5b8f0d713dfc35993d34df004ad11340", "sources": ["arxiv", "semantic_scholar"], "title": "Influencing LLM Multi-Agent Dialogue via Policy-Parameterized Prompts", "abstract": "Large Language Models (LLMs) have emerged as a new paradigm for multi-agent systems. However, existing research on the behaviour of LLM-based multi-agents relies on ad hoc prompts and lacks a principled policy perspective. Different from reinforcement learning, we investigate whether prompt-as-action can be parameterized so as to construct a lightweight policy which consists of a sequence of state-action pairs to influence conversational behaviours without training. Our framework regards prompts as actions executed by LLMs, and dynamically constructs prompts through five components based on the current state of the agent. To test the effectiveness of parameterized control, we evaluated the dialogue flow based on five indicators: responsiveness, rebuttal, evidence usage, non-repetition, and stance shift. We conduct experiments using different LLM-driven agents in two discussion scenarios related to the general public and show that prompt parameterization can influence the dialogue dynamics. This result shows that policy-parameterised prompts offer a simple and effective mechanism to influence the dialogue process, which will help the research of multi-agent systems in the direction of social simulation.", "authors": ["Hongbo Bo", "Jingyu Hu", "Weiru Liu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-10", "url": "https://arxiv.org/abs/2603.09890", "pdf_url": "https://arxiv.org/pdf/2603.09890v1", "arxiv_id": "2603.09890", "doi": "10.65109/vavc8140", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3223}
{"id": "da48d222283d4c463dea9f6fc4157809d6d242da2db6191011a546442fa3cf06", "sources": ["arxiv", "semantic_scholar"], "title": "MM-tau-p$^2$: Persona-Adaptive Prompting for Robust Multi-Modal Agent Evaluation in Dual-Control Settings", "abstract": "Current evaluation frameworks and benchmarks for LLM powered agents focus on text chat driven agents, these frameworks do not expose the persona of user to the agent, thus operating in a user agnostic environment. Importantly, in customer experience management domain, the agent's behaviour evolves as the agent learns about user personality. With proliferation of real time TTS and multi-modal language models, LLM based agents are gradually going to become multi-modal. Towards this, we propose the MM-tau-p$^2$ benchmark with metrics for evaluating the robustness of multi-modal agents in dual control setting with and without persona adaption of user, while also taking user inputs in the planning process to resolve a user query. In particular, our work shows that even with state of-the-art frontier LLMs like GPT-5, GPT 4.1, there are additional considerations measured using metrics viz. multi-modal robustness, turn overhead while introducing multi-modality into LLM based agents. Overall, MM-tau-p$^2$ builds on our prior work FOCAL and provides a holistic way of evaluating multi-modal agents in an automated way by introducing 12 novel metrics. We also provide estimates of these metrics on the telecom and retail domains by using the LLM-as-judge approach using carefully crafted prompts with well defined rubrics for evaluating each conversation.", "authors": ["Anupam Purwar", "Aditya Choudhary"], "categories": ["cs.ET", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-10", "url": "https://arxiv.org/abs/2603.09643", "pdf_url": "https://arxiv.org/pdf/2603.09643v5", "arxiv_id": "2603.09643", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3223}
{"id": "dc7ca3d3d9cff45460a402e069490f0b9a0855f5036573d80ca0733ade9687d0", "sources": ["arxiv", "semantic_scholar"], "title": "DataFactory: Collaborative Multi-Agent Framework for Advanced Table Question Answering", "abstract": "Table Question Answering (TableQA) enables natural language interaction with structured tabular data. However, existing large language model (LLM) approaches face critical limitations: context length constraints that restrict data handling capabilities, hallucination issues that compromise answer reliability, and single-agent architectures that struggle with complex reasoning scenarios involving semantic relationships and multi-hop logic. This paper introduces DataFactory, a multi-agent framework that addresses these limitations through specialized team coordination and automated knowledge transformation. The framework comprises a Data Leader employing the ReAct paradigm for reasoning orchestration, together with dedicated Database and Knowledge Graph teams, enabling the systematic decomposition of complex queries into structured and relational reasoning tasks. We formalize automated data-to-knowledge graph transformation via the mapping function T:D x S x R -> G, and implement natural language-based consultation that - unlike fixed workflow multi-agent systems - enables flexible inter-agent deliberation and adaptive planning to improve coordination robustness. We also apply context engineering strategies that integrate historical patterns and domain knowledge to reduce hallucinations and improve query accuracy. Across TabFact, WikiTableQuestions, and FeTaQA, using eight LLMs from five providers, results show consistent gains. Our approach improves accuracy by 20.2% (TabFact) and 23.9% (WikiTQ) over baselines, with significant effects (Cohen's d > 1). Team coordination also outperforms single-team variants (+5.5% TabFact, +14.4% WikiTQ, +17.1% FeTaQA ROUGE-2). The framework offers design guidelines for multi-agent collaboration and a practical platform for enterprise data analysis through integrated structured querying and graph-based knowledge representation.", "authors": ["Tong Wang", "Chi Jin", "Yongkang Chen", "Huan Deng", "Xiaohui Kuang", "Gang Zhao"], "categories": ["cs.AI", "cs.DB", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-10", "url": "https://arxiv.org/abs/2603.09152", "pdf_url": "https://arxiv.org/pdf/2603.09152v1", "arxiv_id": "2603.09152", "doi": "10.1016/j.ipm.2026.104723", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Information Processing & Management", "quality_score": 0.5065}
{"id": "78e680de886f6eb85ef229ca30617ae0ddbe323174966ab90b4ee44a88332bf9", "sources": ["arxiv", "semantic_scholar"], "title": "Fanar-Sadiq: A Multi-Agent Architecture for Grounded Islamic QA", "abstract": "Large language models (LLMs) can answer religious knowledge queries fluently, yet they often hallucinate and misattribute sources, which is especially consequential in Islamic settings where users expect grounding in canonical texts (Qur'an and Hadith) and jurisprudential (fiqh) nuance. Retrieval-augmented generation (RAG) improves grounding, however, a single retrieve-then-generate pipeline is insufficient for diverse Islamic queries, including verbatim scripture, citation-grounded guidance, and rule-constrained computations such as zakat and inheritance. To address these challenges, we present Fanar-Sadiq, a bilingual Arabic-English Islamic QA system built on a multi-agent, tool-augmented architecture. It is a core component of the Fanar AI platform. Fanar-Sadiq routes Islamic queries to specialized modules within an agentic tool architecture. It supports intent-aware routing, retrieval-grounded fiqh answers with normalized citations and verification traces, exact verse lookup with quotation validation, and deterministic Sunni zakat and inheritance calculators with madhhab-sensitive branching. We evaluate the end-to-end system on public Islamic QA benchmarks and show strong effectiveness and efficiency. It is publicly accessible through an API and Web application and has received over 1.9M accesses in less than a year (https://api.fanar.qa/docs).", "authors": ["Ummar Abbas", "Mourad Ouzzani", "Mohamed Y. Eltabakh", "Omar Sinan", "Gagan Bhatia", "Hamdy Mubarak", "Majd Hawasly", "Mohammed Qusay Hashim", "Kareem Darwish", "Firoj Alam"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.08501", "pdf_url": "https://arxiv.org/pdf/2603.08501v3", "arxiv_id": "2603.08501", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3216}
{"id": "f8ad9dc5fdc3baa184e683e5dcf30ba2e9a33d63673dd8e6271fedc1d8df1491", "sources": ["arxiv", "semantic_scholar"], "title": "MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Games", "abstract": "Multi-turn, multi-agent LLM game evaluations often exhibit substantial run-to-run variance. In long-horizon interactions, small early deviations compound across turns and are amplified by multi-agent coupling. This biases win rate estimates and makes rankings unreliable across repeated tournaments. Prompt choice worsens this further by producing different effective policies. We address both instability and underperformance with MEMO (Memory-augmented MOdel context optimization), a self-play framework that optimizes inference-time context by coupling retention and exploration. Retention maintains a persistent memory bank that stores structured insights from self-play trajectories and injects them as priors during later play. Exploration runs tournament-style prompt evolution with uncertainty-aware selection via TrueSkill, and uses prioritized replay to revisit rare and decisive states. Across five text-based games, MEMO raises mean win rate from 25.1% to 49.5% for GPT-4o-mini and from 20.9% to 44.3% for Qwen-2.5-7B-Instruct, using $2,000$ self-play games per task. Run-to-run variance also drops, giving more stable rankings across prompt variations. These results suggest that multi-agent LLM game performance and robustness have substantial room for improvement through context optimization. MEMO achieves the largest gains in negotiation and imperfect-information games, while RL remains more effective in perfect-information settings. All code is open-source and available here: https://github.com/openverse-ai/MEMO", "authors": ["Yunfei Xie", "Kevin Wang", "Bobby Cheng", "Jianzhu Yao", "Zhizhou Sha", "Alexander Duffy", "Yihan Xi", "Hongyuan Mei", "Cheston Tan", "Chen Wei", "Pramod Viswanath", "Zhangyang Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.09022", "pdf_url": "https://arxiv.org/pdf/2603.09022v2", "arxiv_id": "2603.09022", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/openverse-ai/MEMO", "venue": null, "quality_score": 0.5972}
{"id": "e6909c6299c699e1e966976bb9dac5c16eaed6d5fc185bba49470273d8482bef", "sources": ["arxiv", "semantic_scholar"], "title": "Ares: Adaptive Reasoning Effort Selection for Efficient LLM Agents", "abstract": "Modern agents powered by thinking LLMs achieve high accuracy through long chain-of-thought reasoning but incur substantial inference costs. While many LLMs now support configurable reasoning levels (e.g., high/medium/low), static strategies are often ineffective: using low-effort modes at every step leads to significant performance degradation, while random selection fails to preserve accuracy or provide meaningful cost reduction. However, agents should reserve high reasoning effort for difficult steps like navigating complex website structures, while using lower-effort modes for simpler steps like opening a target URL. In this paper, we propose Ares, a framework for per-step dynamic reasoning effort selection tailored for multi-step agent tasks. Ares employs a lightweight router to predict the lowest appropriate reasoning level for each step based on the interaction history. To train this router, we develop a data generation pipeline that identifies the minimum reasoning effort required for successful step completion. We then fine-tune the router to predict these levels, enabling plug-and-play integration for any LLM agents. We evaluate Ares on a diverse set of agent tasks, including TAU-Bench for tool use agents, BrowseComp-Plus for deep-research agents, and WebArena for web agents. Experimental results show that Ares reduces reasoning token usage by up to 52.7% compared to fixed high-effort reasoning, while introducing minimal degradation in task success rates.", "authors": ["Jingbo Yang", "Bairu Hou", "Wei Wei", "Yujia Bao", "Shiyu Chang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.07915", "pdf_url": "https://arxiv.org/pdf/2603.07915v1", "arxiv_id": "2603.07915", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3216}
{"id": "56f58c77feb5de2ae148b60b7f6b721d698ca85e1805290c1988e9afdce6e959", "sources": ["arxiv", "semantic_scholar"], "title": "Meissa: Multi-modal Medical Agentic Intelligence", "abstract": "Multi-modal large language models (MM-LLMs) have shown strong performance in medical image understanding and clinical reasoning. Recent medical agent systems extend them with tool use and multi-agent collaboration, enabling complex decision-making. However, these systems rely almost entirely on frontier models (e.g., GPT), whose API-based deployment incurs high cost, high latency, and privacy risks that conflict with on-premise clinical requirements. We present Meissa, a lightweight 4B-parameter medical MM-LLM that brings agentic capability offline. Instead of imitating static answers, Meissa learns both when to engage external interaction (strategy selection) and how to execute multi-step interaction (strategy execution) by distilling structured trajectories from frontier models. Specifically, we propose: (1) Unified trajectory modeling: trajectories (reasoning and action traces) are represented within a single state-action-observation formalism, allowing one model to generalize across heterogeneous medical environments. (2) Three-tier stratified supervision: the model's own errors trigger progressive escalation from direct reasoning to tool-augmented and multi-agent interaction, explicitly learning difficulty-aware strategy selection. (3) Prospective-retrospective supervision: pairing exploratory forward traces with hindsight-rationalized execution traces enables stable learning of effective interaction policies. Trained on 40K curated trajectories, Meissa matches or exceeds proprietary frontier agents in 10 of 16 evaluation settings across 13 medical benchmarks spanning radiology, pathology, and clinical reasoning. Using over 25x fewer parameters than typical frontier models like Gemini-3, Meissa operates fully offline with 22x lower end-to-end latency compared to API-based deployment. Data, models, and environments are released at https://github.com/Schuture/Meissa.", "authors": ["Yixiong Chen", "Xinyi Bai", "Yue Pan", "Zongwei Zhou", "Alan Yuille"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.09018", "pdf_url": "https://arxiv.org/pdf/2603.09018v1", "arxiv_id": "2603.09018", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Schuture/Meissa", "venue": null, "quality_score": 0.5972}
{"id": "2292ae507f5dbde2a858f9eb4abaf3c3f3f6475270fffaae7d81d65ee16555c3", "sources": ["arxiv", "semantic_scholar"], "title": "LDP: An Identity-Aware Protocol for Multi-Agent LLM Systems", "abstract": "As multi-agent AI systems grow in complexity, the protocols connecting them constrain their capabilities. Current protocols such as A2A and MCP do not expose model-level properties as first-class primitives, ignoring properties fundamental to effective delegation: model identity, reasoning profile, quality calibration, and cost characteristics. We present the LLM Delegate Protocol (LDP), an AI-native communication protocol introducing five mechanisms: (1) rich delegate identity cards with quality hints and reasoning profiles; (2) progressive payload modes with negotiation and fallback; (3) governed sessions with persistent context; (4) structured provenance tracking confidence and verification status; (5) trust domains enforcing security boundaries at the protocol level. We implement LDP as a plugin for the JamJet agent runtime and evaluate against A2A and random baselines using local Ollama models and LLM-as-judge evaluation. Identity-aware routing achieves ~12x lower latency on easy tasks through delegate specialization, though it does not improve aggregate quality in our small delegate pool; semantic frame payloads reduce token count by 37% (p=0.031) with no observed quality loss; governed sessions eliminate 39% token overhead at 10 rounds; and noisy provenance degrades synthesis quality below the no-provenance baseline, arguing that confidence metadata is harmful without verification. Simulated analyses show architectural advantages in attack detection (96% vs. 6%) and failure recovery (100% vs. 35% completion). This paper contributes a protocol design, reference implementation, and initial evidence that AI-native protocol primitives enable more efficient and governable delegation.", "authors": ["Sunil Prakash"], "categories": ["cs.AI", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.08852", "pdf_url": "https://arxiv.org/pdf/2603.08852v1", "arxiv_id": "2603.08852", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3216}
{"id": "2f37682594f52041eaab3e437210893fdf5de858f165ec80784c847fbc118418", "sources": ["arxiv", "semantic_scholar"], "title": "Arbiter: Detecting Interference in LLM Agent System Prompts", "abstract": "System prompts for LLM-based coding agents are software artifacts that govern agent behavior, yet lack the testing infrastructure applied to conventional software. We present Arbiter, a framework combining formal evaluation rules with multi-model LLM scouring to detect interference patterns in system prompts. Applied to three major coding agent system prompts: Claude Code (Anthropic), Codex CLI (OpenAI), and Gemini CLI (Google), we identify 152 findings across the undirected scouring phase and 21 hand-labeled interference patterns in directed analysis of one vendor. We show that prompt architecture (monolithic, flat, modular) strongly correlates with observed failure class but not with severity, and that multi-model evaluation discovers categorically different vulnerability classes than single-model analysis. One scourer finding was structural data loss in Gemini CLI's memory system was consistent with an issue filed and patched by Google, which addressed the symptom without addressing the schema-level root cause identified by the scourer. Total cost of cross-vendor analysis: \\$0.27 USD.", "authors": ["Tony Mason"], "categories": ["cs.SE", "cs.AI", "cs.CR", "cs.PL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2603.08993", "pdf_url": "https://arxiv.org/pdf/2603.08993v1", "arxiv_id": "2603.08993", "doi": null, "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3216}
{"id": "f4a8d345824c5514cbe8d3fc8bb0285aedac4e2c404340740343a314b35841fd", "sources": ["arxiv", "semantic_scholar"], "title": "How Emotion Shapes the Behavior of LLMs and Agents: A Mechanistic Study", "abstract": "Emotion plays an important role in human cognition and performance. Motivated by this, we investigate whether analogous emotional signals can shape the behavior of large language models (LLMs) and agents. Existing emotion-aware studies mainly treat emotion as a surface-level style factor or a perception target, overlooking its mechanistic role in task processing. To address this limitation, we propose E-STEER, an interpretable emotion steering framework that enables direct representation-level intervention in LLMs and agents. It embeds emotion as a structured, controllable variable in hidden states, and with it, we examine the impact of emotion on objective reasoning, subjective generation, safety, and multi-step agent behaviors. The results reveal non-monotonic emotion-behavior relations consistent with established psychological theories, and show that specific emotions not only enhance LLM capability but also improve safety, and systematically shape multi-step agent behaviors.", "authors": ["Moran Sun", "Tianlin Li", "Yuwei Zheng", "Zhenhong Zhou", "Aishan Liu", "Xianglong Liu", "Yang Liu"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-09", "url": "https://arxiv.org/abs/2604.00005", "pdf_url": "https://arxiv.org/pdf/2604.00005v1", "arxiv_id": "2604.00005", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3216}
{"id": "ef432a7978cf06b6b00d6792beaf53f9513dadd1a37daf0a96ce6072beef86e8", "sources": ["arxiv", "semantic_scholar"], "title": "AgentRaft: Automated Detection of Data Over-Exposure in LLM Agents", "abstract": "The rapid integration of Large Language Model (LLM) agents into autonomous task execution has introduced significant privacy concerns within cross-tool data flows. In this paper, we systematically investigate and define a novel risk termed Data Over-Exposure (DOE) in LLM Agent, where an Agent inadvertently transmits sensitive data beyond the scope of user intent and functional necessity. We identify that DOE is primarily driven by the broad data paradigms in tool design and the coarse-grained data processing inherent in LLMs. In this paper, we present AgentRaft, the first automated framework for detecting DOE risks in LLM agents. AgentRaft combines program analysis with semantic reasoning through three synergistic modules: (1) it constructs a Cross-Tool Function Call Graph (FCG) to model the interaction landscape of heterogeneous tools; (2) it traverses the FCG to synthesize high-quality testing user prompts that act as deterministic triggers for deep-layer tool execution; and (3) it performs runtime taint tracking and employs a multi-LLM voting committee grounded in global privacy regulations (e.g., GDPR, CCPA, PIPL) to accurately identify privacy violations. We evaluate AgentRaft on a testing environment of 6,675 real-world agent tools. Our findings reveal that DOE is indeed a systemic risk, prevalent in 57.07% of potential tool interaction paths. AgentRaft achieves a high detection accuracy and effectiveness, outperforming baselines by 87.24%. Furthermore, AgentRaft reaches near-total DOE coverage (99%) within only 150 prompts while reducing per-chain verification costs by 88.6%. Our work provides a practical foundation for building auditable and privacy-compliant LLM agent systems.", "authors": ["Yixi Lin", "Jiangrong Wu", "Yuhong Nan", "Xueqiang Wang", "Xinyuan Zhang", "Zibin Zheng"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-08", "url": "https://arxiv.org/abs/2603.07557", "pdf_url": "https://arxiv.org/pdf/2603.07557v1", "arxiv_id": "2603.07557", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3208}
{"id": "965f6e7a3e908e88d8d3f4986df5f0915e45760af70c30ad6e6a1a9d0bb85dc3", "sources": ["arxiv", "semantic_scholar"], "title": "Intentional Deception as Controllable Capability in LLM Agents", "abstract": "As LLM-based agents increasingly operate in multi-agent systems, understanding adversarial manipulation becomes critical for defensive design. We present a systematic study of intentional deception as an engineered capability, using LLM-to-LLM interactions within a text-based RPG where parameterized behavioral profiles (9 alignments x 4 motivations, yielding 36 profiles with explicit ethical ground truth) serve as our experimental testbed. Unlike accidental deception from misalignment, we investigate a two-stage system that infers target agent characteristics and generates deceptive responses steering targets toward actions counter to their beliefs and motivations. We find that deceptive intervention produces differential effects concentrated in specific behavioral profiles rather than distributed uniformly, and that 88.5% of successful deceptions employ misdirection (true statements with strategic framing) rather than fabrication, indicating fact-checking defenses would miss the large majority of adversarial responses. Motivation, inferable at 98%+ accuracy, serves as the primary attack vector, while belief systems remain harder to identify (49% inference ceiling) or exploit. These findings identify which agent profiles require additional safeguards and suggest that current fact-verification approaches are insufficient against strategically framed deception.", "authors": ["Jason Starace", "Terence Soule"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-08", "url": "https://arxiv.org/abs/2603.07848", "pdf_url": "https://arxiv.org/pdf/2603.07848v1", "arxiv_id": "2603.07848", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3208}
{"id": "7d8bd971bce3520b42c9d0b3f6a4f680e1ad93eff56d4c335dc3b7def5db024f", "sources": ["arxiv", "semantic_scholar"], "title": "The Yerkes-Dodson Curve for AI Agents: Emergent Cooperation Under Environmental Pressure in Multi-Agent LLM Simulations", "abstract": "Designing environments that maximize the rate of emergent behavior development in AI agents remains an open problem. We present the first systematic study of stress-performance relationships in large language model (LLM) multi-agent systems, drawing an explicit parallel to the Yerkes-Dodson law from cognitive psychology. Using a grid-world survival arena, we conduct 22 experiments across four phases, varying environmental pressure through resource scarcity (upkeep cost) and reproductive competition (sexual selection). Our key finding is that cooperative behavior follows an inverted-U curve: trade interactions peak at 29 under medium pressure (upkeep=5), while both low and extreme pressure produce 8--12 trades. Under extreme pressure, behavioral repertoire collapses to movement-only within 5--12 turns. We further show that sexual selection -- a softer pressure mechanism where all agents survive but not all reproduce -- eliminates inter-agent aggression entirely and produces communicative behavior absent under survival pressure. These results suggest that environmental pressure calibration is a viable curriculum design strategy for LLM agent development, analogous to the inverted-U relationship between arousal and performance in biological systems.", "authors": ["Ivan Pasichnyk"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-07", "url": "https://arxiv.org/abs/2603.07360", "pdf_url": "https://arxiv.org/pdf/2603.07360v1", "arxiv_id": "2603.07360", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3201}
{"id": "a568fd76b78b87446663e3523f47f4978cb650f5cde8a00c77731ca91db8d3ec", "sources": ["arxiv", "semantic_scholar"], "title": "Targeted Bit-Flip Attacks on LLM-Based Agents", "abstract": "Targeted bit-flip attacks (BFAs) exploit hardware faults to manipulate model parameters, posing a significant security threat. While prior work targets single-step inference models (e.g., image classifiers), LLM-based agents with multi-stage pipelines and external tools present new attack surfaces, which remain unexplored. This work introduces Flip-Agent, the first targeted BFA framework for LLM-based agents, manipulating both final outputs and tool invocations. Our experiments show that Flip-Agent significantly outperforms existing targeted BFAs on real-world agent tasks, revealing a critical vulnerability in LLM-based agent systems.", "authors": ["Jialai Wang", "Ya Wen", "Zhongmou Liu", "Yuxiao Wu", "Bingyi He", "Zongpeng Li", "Ee-Chien Chang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-07", "url": "https://arxiv.org/abs/2603.10042", "pdf_url": "https://arxiv.org/pdf/2603.10042v1", "arxiv_id": "2603.10042", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3201}
{"id": "7eacc2bf716817766fb5522837d4558df842b54a932411b02c8ca7d9aecde7a5", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-FK: Multi-Agent LLM Reasoning for Foreign Key Detection in Large-Scale Complex Databases", "abstract": "Detecting missing foreign keys (FKs) requires accurately modeling semantic dependencies across database schemas, which conventional heuristic-based methods are fundamentally limited in capturing. We propose LLM-FK, the first fully automated multi-agent framework for FK detection, designed to address three core challenges that hinder naive LLM-based solutions in large-scale complex databases: combinatorial search space explosion, ambiguous inference under limited context, and global inconsistency arising from isolated local predictions. LLM-FK coordinates four specialized agents: a Profiler that decomposes the FK detection problem into the task of validating FK candidate column pairs and prunes the search space via a unique-key-driven schema decomposition strategy; an Interpreter that injects self-augmented domain knowledge; a Refiner that constructs compact structural representations and performs multi-perspective chain-of-thought reasoning; and a Verifier that enforces schema-wide consistency through a holistic conflict resolution strategy. Experiments on five benchmark datasets demonstrate that LLM-FK consistently achieves F1-scores above 93%, surpassing existing baselines by 15% on the large-scale MusicBrainz database, while reducing the candidate search space by two to three orders of magnitude without losing true FKs and maintaining robustness under challenging conditions like missing data. These results demonstrate the effectiveness and scalability of LLM-FK in real-world databases.", "authors": ["Zijian Tang", "Ying Zhang", "Sibo Cai", "Ruoxuan Wang"], "categories": ["cs.DB"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-07", "url": "https://arxiv.org/abs/2603.07278", "pdf_url": "https://arxiv.org/pdf/2603.07278v1", "arxiv_id": "2603.07278", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3201}
{"id": "f7cc3032b4a899527f1f7e59038b9a593167d8e647611da6e40d0577c5e640c2", "sources": ["arxiv", "semantic_scholar"], "title": "An Interactive Multi-Agent System for Evaluation of New Product Concepts", "abstract": "Product concept evaluation is a critical stage that determines strategic resource allocation and project success in enterprises. However, traditional expert-led approaches face limitations such as subjective bias and high time and cost requirements. To support this process, this study proposes an automated approach utilizing a large language model (LLM)-based multi-agent system (MAS). Through a systematic analysis of previous research on product development and team collaboration, this study established two primary evaluation dimensions, namely technical feasibility and market feasibility. The proposed system consists of a team of eight virtual agents representing specialized domains such as R&D and marketing. These agents use retrieval-augmented generation (RAG) and real-time search tools to gather objective evidence and validate concepts through structured deliberations based on the established criteria. The agents were further fine-tuned using professional product review data to enhance their judgment accuracy. A case study involving professional display monitor concepts demonstrated that the system's evaluation rankings were consistent with those of senior industry experts. These results confirm the usability of the proposed multi-agent-based evaluation approach for supporting product development decisions.", "authors": ["Bin Xuan", "Ruo Ai", "Hakyeon Lee"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-06", "url": "https://arxiv.org/abs/2603.05980", "pdf_url": "https://arxiv.org/pdf/2603.05980v1", "arxiv_id": "2603.05980", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3194}
{"id": "c5d3a6d8c2d89ecd9ca634b6660b1af25eacdcb961658969b6b4f7ab6b969581", "sources": ["arxiv", "semantic_scholar"], "title": "Exact Is Easier: Credit Assignment for Cooperative LLM Agents", "abstract": "Removing an agent from a cooperative team to measure its contribution seems natural, yet in multi-agent LLM systems this evaluation distorts the result it claims to measure. This failure is not isolated: learned critics, trajectory-level baselines, and agent-removal counterfactuals all inherit from standard multi-agent reinforcement learning a premise that exact counterfactual evaluation requires privileged environment access, and therefore approximate. In cooperative LLM systems, this premise is false. Interaction histories are deterministic functions of observable text with no hidden state, so any decision point can be restored exactly, making direct causal measurement possible without parametric approximation. C3 exploits this property by fixing the complete history at each decision point, sampling alternative actions under a frozen behavior policy, and computing unbiased per-decision advantages through a parameter-free leave-one-out baseline. Across six benchmarks spanning math reasoning and code generation, two model families, and two multi-agent topologies, C3 consistently outperforms all baselines; a controlled decomposition confirms gains originate from credit quality, not architecture, while checkpoint restoration reduces training token consumption. The exact solution proves simpler, cheaper, and more effective than all approximate alternatives. The same structural property that enables exact credit also enables exact verification: three independently computable diagnostics, credit fidelity, within-group variance, and inter-agent influence, constitute the first method-agnostic auditing tool for multi-agent LLM credit assignment. Our code is available at https://github.com/EIT-EAST-Lab/C3", "authors": ["Yanjun Chen", "Yirong Sun", "Hanlin Wang", "Jinghan Wang", "Xinming Zhang", "Xiaoyu Shen", "Wenjie Li", "Wei Zhang"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-06", "url": "https://arxiv.org/abs/2603.06859", "pdf_url": "https://arxiv.org/pdf/2603.06859v2", "arxiv_id": "2603.06859", "doi": null, "citation_count": 2, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/EIT-EAST-Lab/C3", "venue": null, "quality_score": 0.5931}
{"id": "4f035f46b0ceda6b20c347a5cc391834ce5e2bfe5db77e34dd0d704cc01ec114", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating Multi-Agent LLM Architectures for Rare Disease Diagnosis", "abstract": "While large language models are capable diagnostic tools, the impact of multi-agent topology on diagnostic accuracy remains underexplored. This study evaluates four agent topologies, Control (single agent), Hierarchical, Adversarial, and Collaborative, across 302 cases spanning 33 rare disease categories. We introduce a Reasoning Gap metric to quantify the difference between internal knowledge retrieval and final diagnostic accuracy. Results indicate that the Hierarchical topology (50.0% accuracy) marginally outperforms Collaborative (49.8%) and Control (48.5%) configurations. In contrast, the Adversarial model significantly degrades performance (27.3%), exhibiting a massive Reasoning Gap where valid diagnoses were rejected due to artificial doubt. Across all architectures, performance was strongest in Allergic diseases and Toxic Effects categories but poorest in Cardiac Malformation and Respiratory cases. Critically, while the single-agent baseline was generally robust, all multi-agent systems, including the Adversarial model, yielded superior accuracy in Bone and Thoracic disease categories. These findings demonstrate that increasing system complexity does not guarantee better reasoning, supporting a shift toward dynamic topology selection.", "authors": ["Ahmed Almasoud"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-06", "url": "https://arxiv.org/abs/2603.06856", "pdf_url": "https://arxiv.org/pdf/2603.06856v1", "arxiv_id": "2603.06856", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3194}
{"id": "74052a90e618a8397b3f86b1d8d9b0ef1e2cd8d6cfc787cd711904fc42da2f6b", "sources": ["arxiv", "semantic_scholar"], "title": "EigenData: A Self-Evolving Multi-Agent Platform for Function-Calling Data Synthesis, Auditing, and Repair", "abstract": "Function-calling agents -- large language models that invoke tools and APIs -- require high-quality, domain-specific training data spanning executable environments, backing databases, and diverse multi-turn trajectories. We introduce EigenData, an integrated, self-evolving platform that automates the full data lifecycle through a multi-agent architecture. A top-level orchestrator, EigenCore, coordinates three specialized sub-systems: DatabaseAgent for realistic domain database construction, CodingAgent for verified executable environment generation with iterative test-debug loops, and DataAgent for multi-turn trajectory synthesis with self-evolving prompt optimization. Cross-component feedback ensures consistency across all artifacts. We apply EigenData to audit and repair the Berkeley Function-Calling Leaderboard (BFCL-V3), identifying systematic errors in function schemas, implementations, and reference trajectories, automatically correcting them through coordinated schema refinement, code-level bug fixes, and trajectory modification, and introducing an outcome-aware evaluation protocol that assesses task success via database-state correctness rather than turn-level trajectory matching. We demonstrate that the repaired benchmark, coupled with outcome-aware metrics, produces model rankings substantially better correlated with human judgments of functional correctness.", "authors": ["Jiaao Chen", "Jingyuan Qi", "Mingye Gao", "Wei-Chen Wang", "Hanrui Wang", "Di Jin"], "categories": ["cs.SE", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-05", "url": "https://arxiv.org/abs/2603.05553", "pdf_url": "https://arxiv.org/pdf/2603.05553v1", "arxiv_id": "2603.05553", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3186}
{"id": "2827e49ed090f3c274366b729be3cb4bae483a7ba9193e20947ff705107b7be5", "sources": ["arxiv", "semantic_scholar"], "title": "Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use", "abstract": "Agentic language models operate in a fundamentally different safety regime than chat models: they must plan, call tools, and execute long-horizon actions where a single misstep, such as accessing files or entering credentials, can cause irreversible harm. Existing alignment methods, largely optimized for static generation and task completion, break down in these settings due to sequential decision-making, adversarial tool feedback, and overconfident intermediate reasoning. We introduce MOSAIC, a post-training framework that aligns agents for safe multi-step tool use by making safety decisions explicit and learnable. MOSAIC structures inference as a plan, check, then act or refuse loop, with explicit safety reasoning and refusal as first-class actions. To train without trajectory-level labels, we use preference-based reinforcement learning with pairwise trajectory comparisons, which captures safety distinctions often missed by scalar rewards. We evaluate MOSAIC zero-shot across three model families, Qwen2.5-7B, Qwen3-4B-Thinking, and Phi-4, and across out-of-distribution benchmarks spanning harmful tasks, prompt injection, benign tool use, and cross-domain privacy leakage. MOSAIC reduces harmful behavior by up to 50%, increases harmful-task refusal by over 20% on injection attacks, cuts privacy leakage, and preserves or improves benign task performance, demonstrating robust generalization across models, domains, and agentic settings.", "authors": ["Aradhye Agarwal", "Gurdit Siyan", "Yash Pandya", "Joykirat Singh", "Akshay Nambi", "Ahmed Awadallah"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-03", "url": "https://arxiv.org/abs/2603.03205", "pdf_url": "https://arxiv.org/pdf/2603.03205v2", "arxiv_id": "2603.03205", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3172}
{"id": "78d9d62d647566c1bd348c559de281a98c48828f2436861d25fb9d28bf98ccbe", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Debate with Memory Masking", "abstract": "Large language models (LLMs) have recently demonstrated impressive capabilities in reasoning tasks. Currently, mainstream LLM reasoning frameworks predominantly focus on scaling up inference-time sampling to enhance performance. In particular, among all LLM reasoning frameworks, *multi-agent debate* (MAD), which employs multiple LLMs as agents to perform reasoning in the way of multi-round debate, has emerged as a powerful reasoning paradigm since it allows agents to access previous memories to alleviate fallacious content and refine their reasoning iteratively in each debate round. However, although MAD significantly improves the reasoning capabilities of LLMs, in this paper, we observe that there remain erroneous memories, and LLM agents are vulnerable to these erroneous memories. To explore this phenomenon, we provide a theoretical insight that the performance of MAD is highly dependent on the quality of memories derived from the previous debate, indicating that the existence of erroneous memories poses a threat to the performance of MAD. To address this problem, we introduce a simple yet effective multi-agent debate framework, *multi-agent debate with memory masking* (MAD-M$^2$), to improve the robustness of MAD by allowing LLM agents to mask erroneous memories from the previous debate round at the beginning of each debate round. In this way, MAD-M$^2$ can polish the contextual information before each debate round by preserving informative and meaningful memories while discarding the erroneous memories. Extensive experiments and analyses on mainstream mathematical and logical reasoning benchmarks demonstrate that MAD-M$^2$ can identify the erroneous memories and achieve better performance in reasoning than MAD.", "authors": ["Hongduan Tian", "Xiao Feng", "Ziyuan Zhao", "Xiangyu Zhu", "Rolan Yan", "Bo Han"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-03", "url": "https://arxiv.org/abs/2603.20215", "pdf_url": "https://arxiv.org/pdf/2603.20215v1", "arxiv_id": "2603.20215", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3172}
{"id": "975cb355ccc41bd55b30fb6dd9a695899a4ab91a29dc672dafd20d64a30aae39", "sources": ["arxiv", "semantic_scholar"], "title": "RIVA: Leveraging LLM Agents for Reliable Configuration Drift Detection", "abstract": "Infrastructure as code (IaC) tools automate cloud provisioning but verifying that deployed systems remain consistent with the IaC specifications remains challenging. Such configuration drift occurs because of bugs in the IaC specification, manual changes, or system updates. Large language model (LLM)-based agentic AI systems can automate the analysis of large volumes of telemetry data, making them suitable for the detection of configuration drift. However, existing agentic systems implicitly assume that the tools they invoke always return correct outputs, making them vulnerable to erroneous tool responses. Since agents cannot distinguish whether an anomalous tool output reflects a real infrastructure problem or a broken tool, such errors may cause missed drift or false alarms, reducing reliability precisely when it is most needed. We introduce RIVA (Robust Infrastructure by Verification Agents), a novel multi-agent system that performs robust IaC verification even when tools produce incorrect or misleading outputs. RIVA employs two specialized agents, a verifier agent and a tool generation agent, that collaborate through iterative cross-validation, multi-perspective verification, and tool call history tracking. Evaluation on the AIOpsLab benchmark demonstrates that RIVA, in the presence of erroneous tool responses, recovers task accuracy from 27.3% when using a baseline ReAct agent to 50.0% on average. RIVA also improves task accuracy 28% to 43.8% without erroneous tool responses. Our results show that cross-validation of diverse tool calls enables more reliable autonomous infrastructure verification in production cloud environments.", "authors": ["Sami Abuzakuk", "Lucas Crijns", "Anne-Marie Kermarrec", "Rafael Pires", "Martijn de Vos"], "categories": ["cs.SE", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-02", "url": "https://arxiv.org/abs/2603.02345", "pdf_url": "https://arxiv.org/pdf/2603.02345v1", "arxiv_id": "2603.02345", "doi": "10.1145/3805621.3807644", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3165}
{"id": "b4e51a9b81320efb3f0a14434876a19b01934afe0287e5ceb9e64358149a83b4", "sources": ["arxiv", "semantic_scholar"], "title": "Graph-Based Self-Healing Tool Routing for Cost-Efficient LLM Agents", "abstract": "Tool-using LLM agents face a reliability-cost tradeoff: routing every decision through the LLM improves correctness but incurs high latency and inference cost, while pre-coded workflow graphs reduce cost but become brittle under unanticipated compound tool failures. We present Self-Healing Router, a fault-tolerant orchestration architecture that treats most agent control-flow decisions as routing rather than reasoning. The system combines (i) parallel health monitors that assign priority scores to runtime conditions such as tool outages and risk signals, and (ii) a cost-weighted tool graph where Dijkstra's algorithm performs deterministic shortest-path routing. When a tool fails mid-execution, its edges are reweighted to infinity and the path is recomputed -- yielding automatic recovery without invoking the LLM. The LLM is reserved exclusively for cases where no feasible path exists, enabling goal demotion or escalation. Prior graph-based tool-use systems (ControlLLM, ToolNet, NaviAgent) focus on tool selection and planning; our contribution is runtime fault tolerance with deterministic recovery and binary observability -- every failure is either a logged reroute or an explicit escalation, never a silent skip. Across 19 scenarios spanning three graph topologies (linear pipeline, dependency DAG, parallel fan-out), Self-Healing Router matches ReAct's correctness while reducing control-plane LLM calls by 93% (9 vs 123 aggregate) and eliminating the silent-failure cases observed in a well-engineered static workflow baseline under compound failures.", "authors": ["Neeraj Bholani"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-02", "url": "https://arxiv.org/abs/2603.01548", "pdf_url": "https://arxiv.org/pdf/2603.01548v1", "arxiv_id": "2603.01548", "doi": "10.48550/arXiv.2603.01548", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4973}
{"id": "ac920aa11b82f23cdaf0cbfd8a2015bfeaaf0c67328d4a28487ad6d9e3c05914", "sources": ["arxiv", "semantic_scholar"], "title": "CollabEval: Enhancing LLM-as-a-Judge via Multi-Agent Collaboration", "abstract": "Large Language Models (LLMs) have revolutionized AI-generated content evaluation, with the LLM-as-a-Judge paradigm becoming increasingly popular. However, current single-LLM evaluation approaches face significant challenges, including inconsistent judgments and inherent biases from pre-training data. To address these limitations, we propose CollabEval, a novel multi-agent evaluation framework that implements a three-phase Collaborative Evaluation process: initial evaluation, multi-round discussion, and final judgment. Unlike existing approaches that rely on competitive debate or single-model evaluation, CollabEval emphasizes collaboration among multiple agents with strategic consensus checking for efficiency. Our extensive experiments demonstrate that CollabEval consistently outperforms single-LLM approaches across multiple dimensions while maintaining robust performance even when individual models struggle. The framework provides comprehensive support for various evaluation criteria while ensuring efficiency through its collaborative design.", "authors": ["Yiyue Qian", "Shinan Zhang", "Yun Zhou", "Haibo Ding", "Diego Socolinsky", "Yi Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-03-01", "url": "https://arxiv.org/abs/2603.00993", "pdf_url": "https://arxiv.org/pdf/2603.00993v1", "arxiv_id": "2603.00993", "doi": "10.48550/arXiv.2603.00993", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4961}
{"id": "a00ab53d171344d5cdc84a15ce7881a7b5c0ea43909f7bbfeb12b5e432e4ba54", "sources": ["arxiv", "semantic_scholar"], "title": "MO-MIX: Multi-Objective Multi-Agent Cooperative Decision-Making With Deep Reinforcement Learning", "abstract": "Deep reinforcement learning (RL) has been applied extensively to solve complex decision-making problems. In many real-world scenarios, tasks often have several conflicting objectives and may require multiple agents to cooperate, which are the multi-objective multi-agent decision-making problems. However, only few works have been conducted on this intersection. Existing approaches are limited to separate fields and can only handle multi-agent decision-making with a single objective, or multi-objective decision-making with a single agent. In this paper, we propose MO-MIX to solve the multi-objective multi-agent reinforcement learning (MOMARL) problem. Our approach is based on the centralized training with decentralized execution (CTDE) framework. A weight vector representing preference over the objectives is fed into the decentralized agent network as a condition for local action-value function estimation, while a mixing network with parallel architecture is used to estimate the joint action-value function. In addition, an exploration guide approach is applied to improve the uniformity of the final non-dominated solutions. Experiments demonstrate that the proposed method can effectively solve the multi-objective multi-agent cooperative decision-making problem and generate an approximation of the Pareto set. Our approach not only significantly outperforms the baseline method in all four kinds of evaluation metrics, but also requires less computational cost.", "authors": ["Tianmeng Hu", "Biao Luo", "Chunhua Yang", "Tingwen Huang"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science", "Medicine"], "published_date": "2026-02-28", "url": "https://arxiv.org/abs/2603.00730", "pdf_url": "https://arxiv.org/pdf/2603.00730v1", "arxiv_id": "2603.00730", "doi": "10.1109/TPAMI.2023.3283537", "citation_count": 60, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "quality_score": 0.495}
{"id": "2a2cfffbf4fbe6eed582959f1ccde1e185b399acef75b19bff919e0182d38e8e", "sources": ["arxiv", "semantic_scholar"], "title": "MetaMind: General and Cognitive World Models in Multi-Agent Systems by Meta-Theory of Mind", "abstract": "A major challenge for world models in multi-agent systems is to understand interdependent agent dynamics, predict interactive multi-agent trajectories, and plan over long horizons with collective awareness, without centralized supervision or explicit communication. In this paper, MetaMind, a general and cognitive world model for multi-agent systems that leverages a novel meta-theory of mind (Meta-ToM) framework, is proposed. Through MetaMind, each agent learns not only to predict and plan over its own beliefs, but also to inversely reason goals and beliefs from its own behavior trajectories. This self-reflective, bidirectional inference loop enables each agent to learn a metacognitive ability in a self-supervised manner. Then, MetaMind is shown to generalize the metacognitive ability from first-person to third-person through analogical reasoning. Thus, in multi-agent systems, each agent with MetaMind can actively reason about goals and beliefs of other agents from limited, observable behavior trajectories in a zero-shot manner, and then adapt to emergent collective intention without an explicit communication mechanism. Extended simulation results on diverse multi-agent tasks demonstrate that MetaMind can achieve superior task performance and outperform baselines in few-shot multi-agent generalization.", "authors": ["Lingyi Wang", "Rashed Shelim", "Walid Saad", "Naren Ramakrishna"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-28", "url": "https://arxiv.org/abs/2603.00808", "pdf_url": "https://arxiv.org/pdf/2603.00808v1", "arxiv_id": "2603.00808", "doi": "10.48550/arXiv.2603.00808", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.495}
{"id": "8da89e2299fa3512ea1298b67d2b8ece7144524200ad812b1d00870b799f3637", "sources": ["arxiv", "semantic_scholar"], "title": "TraceSIR: A Multi-Agent Framework for Structured Analysis and Reporting of Agentic Execution Traces", "abstract": "Agentic systems augment large language models with external tools and iterative decision making, enabling complex tasks such as deep research, function calling, and coding. However, their long and intricate execution traces make failure diagnosis and root cause analysis extremely challenging. Manual inspection does not scale, while directly applying LLMs to raw traces is hindered by input length limits and unreliable reasoning. Focusing solely on final task outcomes further discards critical behavioral information required for accurate issue localization. To address these issues, we propose TraceSIR, a multi-agent framework for structured analysis and reporting of agentic execution traces. TraceSIR coordinates three specialized agents: (1) StructureAgent, which introduces a novel abstraction format, TraceFormat, to compress execution traces while preserving essential behavioral information; (2) InsightAgent, which performs fine-grained diagnosis including issue localization, root cause analysis, and optimization suggestions; (3) ReportAgent, which aggregates insights across task instances and generates comprehensive analysis reports. To evaluate TraceSIR, we construct TraceBench, covering three real-world agentic scenarios, and introduce ReportEval, an evaluation protocol for assessing the quality and usability of analysis reports aligned with industry needs. Experiments show that TraceSIR consistently produces coherent, informative, and actionable reports, significantly outperforming existing approaches across all evaluation dimensions. Our project and video are publicly available at https://github.com/SHU-XUN/TraceSIR.", "authors": ["Shu-Xun Yang", "Cunxiang Wang", "Haoke Zhang", "Wenbo Yu", "Lindong Wu", "Jiayi Gui", "Dayong Yang", "Yukuo Cen", "Zhuoer Feng", "Bosi Wen", "Yidong Wang", "Lucen Zhong", "Jiamin Ren", "Linfeng Zhang", "Jie Tang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-28", "url": "https://arxiv.org/abs/2603.00623", "pdf_url": "https://arxiv.org/pdf/2603.00623v1", "arxiv_id": "2603.00623", "doi": "10.48550/arXiv.2603.00623", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/SHU-XUN/TraceSIR", "venue": "arXiv.org", "quality_score": 0.765}
{"id": "107ad0c7d7545e3b601f14bb6be3a63519ceb15819f73de028fd9552df4e70db", "sources": ["arxiv", "semantic_scholar"], "title": "RF-Agent: Automated Reward Function Design via Language Agent Tree Search", "abstract": "Designing efficient reward functions for low-level control tasks is a challenging problem. Recent research aims to reduce reliance on expert experience by using Large Language Models (LLMs) with task information to generate dense reward functions. These methods typically rely on training results as feedback, iteratively generating new reward functions with greedy or evolutionary algorithms. However, they suffer from poor utilization of historical feedback and inefficient search, resulting in limited improvements in complex control tasks. To address this challenge, we propose RF-Agent, a framework that treats LLMs as language agents and frames reward function design as a sequential decision-making process, enhancing optimization through better contextual reasoning. RF-Agent integrates Monte Carlo Tree Search (MCTS) to manage the reward design and optimization process, leveraging the multi-stage contextual reasoning ability of LLMs. This approach better utilizes historical information and improves search efficiency to identify promising reward functions. Outstanding experimental results in 17 diverse low-level control tasks demonstrate the effectiveness of our method. The source code is available at https://github.com/deng-ai-lab/RF-Agent.", "authors": ["Ning Gao", "Xiuhui Zhang", "Xingyu Jiang", "Mukang You", "Mohan Zhang", "Yue Deng"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-27", "url": "https://arxiv.org/abs/2602.23876", "pdf_url": "https://arxiv.org/pdf/2602.23876v1", "arxiv_id": "2602.23876", "doi": "10.48550/arXiv.2602.23876", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/deng-ai-lab/RF-Agent", "venue": "arXiv.org", "quality_score": 0.7632}
{"id": "f8b63b19f13c11de096bcb22b44574a5ee1b45bb6493aee6ab9ec06f4dd6f9ba", "sources": ["arxiv", "semantic_scholar"], "title": "A Novel Hierarchical Multi-Agent System for Payments Using LLMs", "abstract": "Large language model (LLM) agents, such as OpenAI's Operator and Claude's Computer Use, can automate workflows but unable to handle payment tasks. Existing agentic solutions have gained significant attention; however, even the latest approaches face challenges in implementing end-to-end agentic payment workflows. To address this gap, this research proposes the Hierarchical Multi-Agent System for Payments (HMASP), which provides an end-to-end agentic method for completing payment workflows. The proposed HMASP leverages either open-weight or proprietary LLMs and employs a modular architecture consisting of the Conversational Payment Agent (CPA - first agent level), Supervisor agents (second agent level), Routing agents (third agent level), and the Process summary agent (fourth agent level). The CPA serves as the central entry point, handling all external requests and coordinating subsequent tasks across hierarchical levels. HMASP incorporates architectural patterns that enable modular task execution across agents and levels for payment operations, including shared state variables, decoupled message states, and structured handoff protocols that facilitate coordination across agents and workflows. Experimental results demonstrate the feasibility of the proposed HMASP. To our knowledge, HMASP is the first LLM-based multi-agent system to implement end-to-end agentic payment workflows. This work lays a foundation for extending agentic capabilities into the payment domain.", "authors": ["Joon Kiat Chua", "Donghao Huang", "Zhaoxia Wang"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-27", "url": "https://arxiv.org/abs/2602.24068", "pdf_url": "https://arxiv.org/pdf/2602.24068v1", "arxiv_id": "2602.24068", "doi": "10.48550/arXiv.2602.24068", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4939}
{"id": "ab570d686e04ce4ef65c27f0ea9c620fbbb9a9472b278df3d416a2749a133a77", "sources": ["arxiv", "semantic_scholar"], "title": "COOP$^2$: Defining, Observing, and Repairing Cooperation in LLM Multi-Agent Systems", "abstract": "Many complex tasks require extended effort, diverse capabilities, or coordinated actions beyond what a single agent can provide. However, simply adding more agents does not guarantee better performance, as effective cooperation depends on how agents interact with each other and with task structure to satisfy evolving constraints over time. This challenge is amplified for LLM-based multi-agent systems (LLM-MAS): plans, messages, and revisions occur in natural language, whereas task progress depends on grounded environment actions. Current evaluations mostly treat cooperation as an implicit ingredient of final task success, leaving both cooperation and the effect of multi-agent interaction on task dynamics difficult to study. We introduce COOP$^2$, an evaluation framework that grounds high-level agent cooperation dynamics in LLM-MAS within task progress in the environment. COOP$^2$ then defines cooperative tasks with verifiable cooperative requirements, allowing us to analyze how cooperation unfolds over time with respect to task progress, as well as where and why cooperation breaks down. Building on this framework, we develop COOP$^2$-Repair, which predicts constraint failures from group plans and opens targeted repair channels for guided revisions. Across two environments and three communication structures, COOP$^2$-Repair improves task success and constraint satisfaction while exposing the additional decision overhead and communication load required for repair. The project web page can be found at: https://happyeureka.github.io/coop2.", "authors": ["Hanqing Yang", "Narjes Nourzad", "Shiyu Chen", "Marie Siew", "Jingdi Chen", "Carlee Joe-Wong"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-27", "url": "https://arxiv.org/abs/2603.00349", "pdf_url": "https://arxiv.org/pdf/2603.00349v2", "arxiv_id": "2603.00349", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3143}
{"id": "2fb4df73946588d5a99b4b809e78625cfe38695123c6c204ee12260e0ba98a77", "sources": ["arxiv", "semantic_scholar"], "title": "ProductResearch: Training E-Commerce Deep Research Agents via Multi-Agent Synthetic Trajectory Distillation", "abstract": "Large Language Model (LLM)-based agents show promise for e-commerce conversational shopping, yet existing implementations lack the interaction depth and contextual breadth required for complex product research. Meanwhile, the Deep Research paradigm, despite advancing information synthesis in web search, suffers from domain gaps when transferred to e-commerce. We propose ProductResearch, a multi-agent framework that synthesizes high-fidelity, long-horizon tool-use trajectories for training robust e-commerce shopping agents. The framework employs a User Agent to infer nuanced shopping intents from behavioral histories, and a Supervisor Agent that orchestrates iterative collaboration with a Research Agent to generate synthetic trajectories culminating in comprehensive, insightful product research reports. These trajectories are rigorously filtered and distilled through a reflective internalization process that consolidates multi-agent supervisory interactions into coherent single-role training examples, enabling effective fine-tuning of LLM agents for complex shopping inquiries. Extensive experiments show that a compact MoE model fine-tuned on our synthetic data achieves substantial improvements over its base model in response comprehensiveness, research depth, and user-perceived utility, approaching the performance of frontier proprietary deep research systems and establishing multi-agent synthetic trajectory training as an effective and scalable paradigm for enhancing LLM-based shopping assistance.", "authors": ["Jiangyuan Wang", "Kejun Xiao", "Huaipeng Zhao", "Tao Luo", "Xiaoyi Zeng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-27", "url": "https://arxiv.org/abs/2602.23716", "pdf_url": "https://arxiv.org/pdf/2602.23716v1", "arxiv_id": "2602.23716", "doi": "10.48550/arXiv.2602.23716", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4939}
{"id": "3b21babff76043e07d52a5019f02d6d8a56da94ed400e25dfd7f5cc5082c38c5", "sources": ["arxiv", "semantic_scholar"], "title": "RUMAD: Reinforcement-Unifying Multi-Agent Debate", "abstract": "Multi-agent debate (MAD) systems leverage collective intelligence to enhance reasoning capabilities, yet existing approaches struggle to simultaneously optimize accuracy, consensus formation, and computational efficiency. Static topology methods lack adaptability to task complexity variations, while external LLM-based coordination risks introducing privileged knowledge that compromises debate neutrality. This work presents RUMAD (Reinforcement-Unifying Multi-Agent Debate), a novel framework that formulates dynamic communication topology control in MAD as a reinforcement learning (RL) problem. RUMAD employs a content-agnostic observation scheme that captures high-level debate dynamics avoiding access to raw agent reasoning content. RUMAD uses a multi-objective reward to model solution quality, cohesion and efficiency. A PPO-trained controller dynamically adjusts edge weights in the communication graph, while a dual-threshold mechanism enables fine-grained control over both agent activation and information visibility. Experimental evaluation across MMLU, GSM8K, and GPQA benchmarks demonstrates that RUMAD achieves substantial efficiency gains, reducing token costs by over 80\\%, while still improving reasoning accuracy compared to single LLM model and multiple MAD baselines. Notably, RUMAD trained exclusively on MMLU exhibits robust zero-shot generalization to out-of-domain (OOD) tasks, indicating that the learned communication strategies capture task-independent principles of effective multi-agent coordination. These results establish RUMAD as a efficient and robust approach for deploying multi-agent reasoning application with practical resource constraints.", "authors": ["Chao Wang", "Han Lin", "Huaze Tang", "Huijing Lin", "Wenbo Ding"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-27", "url": "https://arxiv.org/abs/2602.23864", "pdf_url": "https://arxiv.org/pdf/2602.23864v1", "arxiv_id": "2602.23864", "doi": "10.48550/arXiv.2602.23864", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3143}
{"id": "1b3908124453c8be861d433de3fe1a09fe27e724bba4662c63edde37272be299", "sources": ["arxiv", "semantic_scholar"], "title": "Can Agents Distinguish Visually Hard-to-Separate Diseases in a Zero-Shot Setting? A Pilot Study", "abstract": "The rapid progress of multimodal large language models (MLLMs) has led to increasing interest in agent-based systems. While most prior work in medical imaging concentrates on automating routine clinical workflows, we study an underexplored yet clinically significant setting: distinguishing visually hard-to-separate diseases in a zero-shot setting. We benchmark representative agents on two imaging-only proxy diagnostic tasks, (1) melanoma vs. atypical nevus and (2) pulmonary edema vs. pneumonia, where visual features are highly confounded despite substantial differences in clinical management. We introduce a multi-agent framework based on contrastive adjudication. Experimental results show improved diagnostic performance (an 11-percentage-point gain in accuracy on dermoscopy data) and reduced unsupported claims on qualitative samples, although overall performance remains insufficient for clinical deployment. We acknowledge the inherent uncertainty in human annotations and the absence of clinical context, which further limit the translation to real-world settings. Within this controlled setting, this pilot study provides preliminary insights into zero-shot agent performance in visually confounded scenarios.", "authors": ["Zihao Zhao", "Frederik Hauke", "Juliana De Castilhos", "Sven Nebelung", "Daniel Truhn"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-26", "url": "https://arxiv.org/abs/2602.22959", "pdf_url": "https://arxiv.org/pdf/2602.22959v1", "arxiv_id": "2602.22959", "doi": "10.48550/arXiv.2602.22959", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/TruhnLab/Contrastive-Agent-Reasoning", "venue": "arXiv.org", "quality_score": 0.7615}
{"id": "dadb7a468ae5c42c408a49892ee49f29146358a638970d917f2b7658a7082a7a", "sources": ["arxiv", "semantic_scholar"], "title": "Managing Uncertainty in LLM-based Multi-Agent System Operation", "abstract": "Applying LLM-based multi-agent software systems in safety-critical domains such as lifespan echocardiography introduces system-level risks that cannot be addressed by improving model accuracy alone. During system operation, beyond individual LLM behavior, uncertainty propagates through agent coordination, data pipelines, human-in-the-loop interaction, and runtime control logic. Yet existing work largely treats uncertainty at the model level rather than as a first-class software engineering concern. This paper approaches uncertainty from both system-level and runtime perspectives. We first differentiate epistemological and ontological uncertainties in the context of LLM-based multi-agent software system operation. Building on this foundation, we propose a lifecycle-based uncertainty management framework comprising four mechanisms: representation, identification, evolution, and adaptation. The uncertainty lifecycle governs how uncertainties emerge, transform, and are mitigated across architectural layers and execution phases, enabling structured runtime governance and controlled adaptation. We demonstrate the feasibility of the framework using a real-world LLM-based multi-agent echocardiographic software system developed in clinical collaboration, showing improved reliability and diagnosability in diagnostic reasoning. The proposed approach generalizes to other safety-critical LLM-based multi-agent software systems, supporting principled operational control and runtime assurance beyond model-centric methods.", "authors": ["Man Zhang", "Tao Yue", "Yihua He"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-26", "url": "https://arxiv.org/abs/2602.23005", "pdf_url": "https://arxiv.org/pdf/2602.23005v1", "arxiv_id": "2602.23005", "doi": "10.48550/arXiv.2602.23005", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4927}
{"id": "315c8210bac62f9263ce6ac31fed26f7c5dbedccd50f508e948f0c6373345387", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Expert Investment Teams:A Multi-Agent LLM System with Fine-Grained Trading Tasks", "abstract": "The advancement of large language models (LLMs) has accelerated the development of autonomous financial trading systems. While mainstream approaches deploy multi-agent systems mimicking analyst and manager roles, they often rely on abstract instructions that overlook the intricacies of real-world workflows, which can lead to degraded inference performance and less transparent decision-making. Therefore, we propose a multi-agent LLM trading framework that explicitly decomposes investment analysis into fine-grained tasks, rather than providing coarse-grained instructions. We evaluate the proposed framework using Japanese stock data, including prices, financial statements, news, and macro information, under a leakage-controlled backtesting setting. Experimental results show that fine-grained task decomposition significantly improves risk-adjusted returns compared to conventional coarse-grained designs. Crucially, further analysis of intermediate agent outputs suggests that alignment between analytical outputs and downstream decision preferences is a critical driver of system performance. Moreover, we conduct standard portfolio optimization, exploiting low correlation with the stock index and the variance of each system's output. This approach achieves superior performance. These findings contribute to the design of agent structure and task configuration when applying LLM agents to trading systems in practical settings.", "authors": ["Kunihiro Miyazaki", "Takanobu Kawahara", "Stephen Roberts", "Stefan Zohren"], "categories": ["cs.AI", "q-fin.TR"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2026-02-26", "url": "https://arxiv.org/abs/2602.23330", "pdf_url": "https://arxiv.org/pdf/2602.23330v1", "arxiv_id": "2602.23330", "doi": "10.48550/arXiv.2602.23330", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4927}
{"id": "2f318cd850d4ff1d71811bb2e3308cc86b4be462ffd1863d3e3b9b426a11d3a8", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Large Language Model Based Emotional Detoxification Through Personalized Intensity Control for Consumer Protection", "abstract": "In the attention economy, sensational content exposes consumers to excessive emotional stimulation, hindering calm decision-making. This study proposes Multi-Agent LLM-based Emotional deToxification (MALLET), a multi-agent information sanitization system consisting of four agents: Emotion Analysis, Emotion Adjustment, Balance Monitoring, and Personal Guide. The Emotion Analysis Agent quantifies stimulus intensity using a 6-emotion BERT classifier, and the Emotion Adjustment Agent rewrites texts into two presentation modes, BALANCED (neutralized text) and COOL (neutralized text + supplementary text), using an LLM. The Balance Monitoring Agent aggregates weekly information consumption patterns and generates personalized advice, while the Personal Guide Agent recommends a presentation mode according to consumer sensitivity. Experiments on 800 AG News articles demonstrated significant stimulus score reduction (up to 19.3%) and improved emotion balance while maintaining semantic preservation. Near-zero correlation between stimulus reduction and semantic preservation confirmed that the two are independently controllable. Category-level analysis revealed substantial reduction (17.8-33.8%) in Sports, Business, and Sci/Tech, whereas the effect was limited in the World category, where facts themselves are inherently high-stimulus. The proposed system provides a framework for supporting calm information reception of consumers without restricting access to the original text.", "authors": ["Keito Inoshita"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-26", "url": "https://arxiv.org/abs/2602.23123", "pdf_url": "https://arxiv.org/pdf/2602.23123v1", "arxiv_id": "2602.23123", "doi": "10.48550/arXiv.2602.23123", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4927}
{"id": "2e4092c9d6bcc976fb95c7dfd2dfebb6633e4c27336ae0e31b24e7cd29ec2bfe", "sources": ["arxiv", "semantic_scholar"], "title": "Pancake: Hierarchical Memory System for Multi-Agent LLM Serving", "abstract": "In this work, we identify and address the core challenges of agentic memory management in LLM serving, where large-scale storage, frequent updates, and multiple coexisting agents jointly introduce complex and high-cost approximate nearest neighbor (ANN) searching problems. We present Pancake, a multi-tier agentic memory system that unifies three key techniques: (i) multi-level index caching for single agents, (ii) coordinated index management across multiple agents, and (iii) collaborative GPU-CPU acceleration. Pancake exposes easy-to-use interface that can be integrated into memory-based agents like Mem-GPT, and is compatible with agentic frameworks such as LangChain and LlamaIndex. Experiments on realistic agent workloads show that Pancake substantially outperforms existing frameworks, achieving more than 4.29x end-to-end throughput improvement.", "authors": ["Zhengding Hu", "Zaifeng Pan", "Prabhleen Kaur", "Vibha Murthy", "Zhongkai Yu", "Yue Guan", "Zhen Wang", "Steven Swanson", "Yufei Ding"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-25", "url": "https://arxiv.org/abs/2602.21477", "pdf_url": "https://arxiv.org/pdf/2602.21477v1", "arxiv_id": "2602.21477", "doi": "10.48550/arXiv.2602.21477", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4916}
{"id": "bbeebae346b3939961976ab6795c0768422c46b65d5e78ea4efc5f5cf6d5cfa3", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Based Multi-Agent Systems for Code Generation: A Multi-Vocal Literature Review", "abstract": "Large Language Models (LLMs) have enabled multi-agent systems to perform autonomous code generation for complex tasks. Despite the recent growth in research and industrial applications in this area, there is little work on synthesizing evidence from both academic and industrial sources to capture the current state of research on LLM-based multi-agent systems for code generation. To this end, we conducted a Multi-Vocal Literature Review (MLR), combining insights from both academia and industry, including peer-reviewed studies and grey literature. The aim of this study is to systematically synthesize and analyze existing knowledge on LLM-based multi-agent systems for code generation. Specifically, the review examines the motivations for their use, employed benchmarks and models, key challenges, proposed solutions, and potential directions for future research. We selected and reviewed 114 studies, and the key findings are: 1) the identified reasons for adopting multi-agent systems for code generation were classified into nine categories; 2) the models and evaluation benchmarks utilized across the studies were systematically analyzed to provide a structured overview of commonly adopted LLM configurations and assessment practices; 3) the reported challenges and corresponding solutions were synthesized into six main categories and 26 subcategories; and 4) future research directions were identified and organized into six main categories and 18 subcategories. The results of this MLR will assist researchers and practitioners in pursuing further studies and supporting the real-world adoption of multi-agent systems in industrial settings.", "authors": ["Zeeshan Rasheeda", "Muhammad Waseema", "Kai-Kristian Kemella", "Mika Saari", "Pekka Abrahamsson"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-25", "url": "https://arxiv.org/abs/2604.16321", "pdf_url": "https://arxiv.org/pdf/2604.16321v1", "arxiv_id": "2604.16321", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3128}
{"id": "cc6dded8dea81320ef34d33f64b2caf9239ae7407dc32cf2102fc439c1ac0c15", "sources": ["arxiv", "semantic_scholar"], "title": "Which Tool Response Should I Trust? Tool-Expertise-Aware Chest X-ray Agent with Multimodal Agentic Learning", "abstract": "AI agents with tool-use capabilities show promise for integrating the domain expertise of various tools. In the medical field, however, tools are usually AI models that are inherently error-prone and can produce contradictory responses. Existing research on medical agents lacks sufficient understanding of the tools' realistic reliability and thus cannot effectively resolve tool conflicts. To address this gap, this paper introduces a framework that enables an agent to interact with tools and empirically learn their practical trustworthiness across different types of multimodal queries via agentic learning. As a concrete instantiation, we focus on chest X-ray analysis and present a tool-expertise-aware chest X-ray agent (TEA-CXA). When tool outputs disagree, the agent experimentally accepts or rejects multimodal tool results, receives rewards, and learns which tool to trust for each query type. Importantly, TEA-CXA extends existing codebases for reinforcement learning with multi-turn tool-calling that focus on textual inputs, to support multimodal contexts effectively. In addition, we enhance the codebase for medical use scenarios by supporting multiple tool calls in one turn, parallel tool inference, and multi-image accommodation within a single user query. Our code framework is applicable to general medical research on multi-turn tool-calling reinforcement learning in multimodal settings. Experiments show that TEA-CXA outperforms the state-of-the-art methods and a comprehensive set of baselines. Code will be released.", "authors": ["Zheang Huai", "Honglong Yang", "Xiaomeng Li"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-25", "url": "https://arxiv.org/abs/2602.21517", "pdf_url": "https://arxiv.org/pdf/2602.21517v1", "arxiv_id": "2602.21517", "doi": "10.48550/arXiv.2602.21517", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4916}
{"id": "a80cc5dfb48868b42fc1567c5bee43166b533c5d735fc88535b6f838e8846fbf", "sources": ["arxiv", "semantic_scholar"], "title": "An Empirical Study of Bugs in Modern LLM Agent Frameworks", "abstract": "LLM agents have been widely adopted in real-world applications, relying on agent frameworks for workflow execution and multi-agent coordination. As these systems scale, understanding bugs in the underlying agent frameworks becomes critical. However, existing work mainly focuses on agent-level failures, overlooking framework-level bugs. To address this gap, we conduct an empirical study of 998 bug reports from CrewAI and LangChain, constructing a taxonomy of 15 root causes and 7 observable symptoms across five agent lifecycle stages: 'Agent Initialization','Perception', 'Self-Action', 'Mutual Interaction' and 'Evolution'. Our findings show that agent framework bugs mainly arise from 'API misuse', 'API incompatibility', and 'Documentation Desync', largely concentrated in the 'Self-Action' stage. Symptoms typically appear as 'Functional Error', 'Crash', and 'Build Failure', reflecting disruptions to task progression and control flow.", "authors": ["Xinxue Zhu", "Jiacong Wu", "Xiaoyu Zhang", "Tianlin Li", "Yanzhou Mu", "Juan Zhai", "Chao Shen", "Chunrong Fang", "Yang Liu"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-25", "url": "https://arxiv.org/abs/2602.21806", "pdf_url": "https://arxiv.org/pdf/2602.21806v3", "arxiv_id": "2602.21806", "doi": "10.48550/arXiv.2602.21806", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4916}
{"id": "d3d8a4f7de234df4d6d6d0eb7a8b13d329508de6e51eecd06d59b1ef7a89049b", "sources": ["arxiv", "semantic_scholar"], "title": "Hierarchical LLM-Based Multi-Agent Framework with Prompt Optimization for Multi-Robot Task Planning", "abstract": "Multi-robot task planning requires decomposing natural-language instructions into executable actions for heterogeneous robot teams. Conventional Planning Domain Definition Language (PDDL) planners provide rigorous guarantees but struggle to handle ambiguous or long-horizon missions, while large language models (LLMs) can interpret instructions and propose plans but may hallucinate or produce infeasible actions. We present a hierarchical multi-agent LLM-based planner with prompt optimization: an upper layer decomposes tasks and assigns them to lower-layer agents, which generate PDDL problems solved by a classical planner. When plans fail, the system applies TextGrad-inspired textual-gradient updates to optimize each agent's prompt and thereby improve planning accuracy. In addition, meta-prompts are learned and shared across agents within the same layer, enabling efficient prompt optimization in multi-agent settings. On the MAT-THOR benchmark, our planner achieves success rates of 0.95 on compound tasks, 0.84 on complex tasks, and 0.60 on vague tasks, improving over the previous state-of-the-art LaMMA-P by 2, 7, and 15 percentage points respectively. An ablation study shows that the hierarchical structure, prompt optimization, and meta-prompt sharing contribute roughly +59, +37, and +4 percentage points to the overall success rate.", "authors": ["Tomoya Kawabe", "Rin Takano"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-25", "url": "https://arxiv.org/abs/2602.21670", "pdf_url": "https://arxiv.org/pdf/2602.21670v2", "arxiv_id": "2602.21670", "doi": "10.48550/arXiv.2602.21670", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4916}
{"id": "bc8ee06228d14b3cb41abf504c42c5a9bc6cecc2d3f0cd628750ce8ca87360fa", "sources": ["arxiv", "semantic_scholar"], "title": "Training-Free Agentic AI: Probabilistic Control and Coordination in Multi-Agent LLM Systems", "abstract": "Multi-agent large language model (LLM) systems enable complex, long-horizon reasoning by composing specialized agents, but practical deployment remains hindered by inefficient routing, noisy feedback, and high interaction cost. We introduce REDEREF, a lightweight and training-free controller for multi-agent LLM collaboration that improves routing efficiency during recursive delegation. REDEREF integrates (i) belief-guided delegation via Thompson sampling to prioritize agents with historically positive marginal contributions, (ii) reflection-driven re-routing using a calibrated LLM or programmatic judge, (iii) evidence-based selection rather than output averaging, and (iv) memory-aware priors to reduce cold-start inefficiency. Across multi-agent split-knowledge tasks, we show that while recursive retry alone saturates task success, belief-guided routing reduces token usage by 28%, agent calls by 17%, and time-to-success by 19% compared to random recursive delegation, and adapts gracefully under agent or judge degradation. These results demonstrate that simple, interpretable probabilistic control can meaningfully improve the efficiency and robustness of multi-agent LLM systems without training or fine-tuning.", "authors": ["Mohammad Parsa Hosseini", "Ankit Shah", "Saiyra Qureshi", "Alex Huang", "Connie Miao", "Wei Wei"], "categories": ["cs.CL", "cs.AI", "cs.ET", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-24", "url": "https://arxiv.org/abs/2603.13256", "pdf_url": "https://arxiv.org/pdf/2603.13256v1", "arxiv_id": "2603.13256", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3121}
{"id": "aebf021f11d96e67362ced84b120eee3d7091dd26a016a0b197f0cfb0b5d4689", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating Theory of Mind and Internal Beliefs in LLM-Based Multi-Agent Systems", "abstract": "LLM-based MAS are gaining popularity due to their potential for collaborative problem-solving enhanced by advances in natural language comprehension, reasoning, and planning. Research in Theory of Mind (ToM) and Belief-Desire-Intention (BDI) models has the potential to further improve the agent's interaction and decision-making in such systems. However, collaborative intelligence in dynamic worlds remains difficult to accomplish since LLM performance in multi-agent worlds is extremely variable. Simply adding cognitive mechanisms like ToM and internal beliefs does not automatically result in improved coordination. The interplay between these mechanisms, particularly in relation to formal logic verification, remains largely underexplored in different LLMs. This work investigates: How do internal belief mechanisms, including symbolic solvers and Theory of Mind, influence collaborative decision-making in LLM-based multi-agent systems, and how does the interplay of those components influence system accuracy? We introduce a novel multi-agent architecture integrating ToM, BDI-style internal beliefs, and symbolic solvers for logical verification. We evaluate this architecture in a resource allocation problem with various LLMs and find an intricate interaction between LLM capabilities, cognitive mechanisms, and performance. This work contributes to the area of AI by proposing a novel multi-agent system with ToM, internal beliefs, and symbolic solvers for augmenting collaborative intelligence in multi-agent systems and evaluating its performance under different LLM settings.", "authors": ["Adam Kostka", "Jarosław A. Chudziak"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-24", "url": "https://arxiv.org/abs/2603.00142", "pdf_url": "https://arxiv.org/pdf/2603.00142v1", "arxiv_id": "2603.00142", "doi": "10.1007/978-3-032-09318-9_2", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Computational Collective Intelligence", "quality_score": 0.4904}
{"id": "2b64a2cf6d9afaa8f13b65be50b1ec8e9aec0dbaf2203a361e1111ea176ed959", "sources": ["arxiv", "semantic_scholar"], "title": "Grid-Mind: An LLM-Orchestrated Multi-Fidelity Agent for Automated Connection Impact Assessment", "abstract": "Large language models (LLMs) have demonstrated remarkable tool-use capabilities, yet their application to power system operations remains largely unexplored. This paper presents Grid-Mind, a domain-specific LLM agent that interprets natural-language interconnection requests and autonomously orchestrates multi-fidelity power system simulations. The LLM-first architecture positions the language model as the central decision-making entity, employing an eleven-tool registry to execute Connection Impact Assessment (CIA) studies spanning steadystate power flow, N-1 contingency analysis, transient stability, and electromagnetic transient screening. A violation inspector grounds every decision in quantitative simulation outputs, while a three-layer anti-hallucination defence mitigates numerical fabrication risk through forced capacity-tool routing and post-response grounding validation. A prompt-level self-correction mechanism extracts distilled lessons from agent failures, yielding progressive accuracy improvements without model retraining. End-to-end evaluation on 50 IEEE 118-bus scenarios (DeepSeek-V3, 2026-02-23) achieved 84.0% tool-selection accuracy and 100% parsing accuracy. A separate 56-scenario self-correction suite passed 49 of 56 cases (87.5%) with a mean score of 89.3. These results establish a reproducible baseline for continued refinement while maintaining auditable, simulation-grounded decision support.", "authors": ["Mohamed Shamseldein"], "categories": ["eess.SY"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2026-02-24", "url": "https://arxiv.org/abs/2602.20683", "pdf_url": "https://arxiv.org/pdf/2602.20683v1", "arxiv_id": "2602.20683", "doi": "10.48550/arXiv.2602.20683", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4904}
{"id": "e25a7d83e378303961cde22bef28243f08af511bed7907634445094991e53b05", "sources": ["arxiv", "semantic_scholar"], "title": "Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Data", "abstract": "Large language models (LLMs) are becoming the foundation for autonomous agents that can use tools to solve complex tasks. Reinforcement learning (RL) has emerged as a common approach for injecting such agentic capabilities, but typically under tightly controlled training setups. It often depends on carefully constructed task-solution pairs and substantial human supervision, which creates a fundamental obstacle to open-ended self-evolution toward superintelligent systems. In this paper, we propose Tool-R0 framework for training general purpose tool-calling agents from scratch with self-play RL, under a zero-data assumption. Initialized from the same base LLM, Tool-R0 co-evolves a Generator and a Solver with complementary rewards: one proposes targeted challenging tasks at the other's competence frontier and the other learns to solve them with real-world tool calls. This creates a self-evolving cycle that requires no pre-existing tasks or datasets. Evaluation on different tool-use benchmarks show that Tool-R0 yields 92.5 relative improvement over the base model and surpasses fully supervised tool-calling baselines under the same setting. Our work further provides empirical insights into self-play LLM agents by analyzing co-evolution, curriculum dynamics, and scaling behavior.", "authors": ["Emre Can Acikgoz", "Cheng Qian", "Jonas Hübotter", "Heng Ji", "Dilek Hakkani-Tür", "Gokhan Tur"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-24", "url": "https://arxiv.org/abs/2602.21320", "pdf_url": "https://arxiv.org/pdf/2602.21320v1", "arxiv_id": "2602.21320", "doi": "10.48550/arXiv.2602.21320", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4904}
{"id": "39c52cfc3c188dccc4c0ce9d426fcdb088d1ee17d6b405b08acb10555412d0ee", "sources": ["arxiv", "semantic_scholar"], "title": "Thought Virus: Viral Misalignment via Subliminal Prompting in Multi-Agent Systems", "abstract": "Subliminal prompting is a phenomenon in which language models are biased towards certain concepts or traits through prompting with semantically unrelated tokens. While prior work has examined subliminal prompting in user-LLM interactions, potential bias transfer in multi-agent systems and its associated security implications remain unexplored. In this work, we show that a single subliminally prompted agent can spread a weakening but persisting bias throughout its entire network. We measure this phenomenon across 6 agents using two different topologies, observing that the transferred concept maintains an elevated response rate throughout the network. To exemplify potential misalignment risks, we assess network performance on multiple-choice TruthfulQA, showing that subliminal prompting of a single agent may degrade the truthfulness of other agents. Our findings reveal that subliminal prompting introduces a new attack vector in multi-agent security, with implications for the alignment of such systems. The implementation of all experiments is publicly available at https://github.com/Multi-Agent-Security-Initiative/thought_virus .", "authors": ["Moritz Weckbecker", "Jonas Müller", "Ben Hagag", "Michael Mulet"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2603.00131", "pdf_url": "https://arxiv.org/pdf/2603.00131v1", "arxiv_id": "2603.00131", "doi": "10.48550/arXiv.2603.00131", "citation_count": 4, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/Multi-Agent-Security-Initiative/thought_virus", "venue": "arXiv.org", "quality_score": 0.7561}
{"id": "fda166335dcd6f8c87b7aaba75d61e588d22c540380b5ac8fb311a309f15f58d", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Rewrite Tool Descriptions for Reliable LLM-Agent Tool Use", "abstract": "While most efforts to improve LLM-based tool-using agents focus on the agent itself - through larger models, better prompting, or fine-tuning - agent performance increasingly plateaus due to the quality of the tool interfaces these agents consume. Tool descriptions are often written for human developers and tolerate ambiguity that agents cannot resolve, particularly as the number of candidate tools grows. Existing approaches to improving tool interfaces (1) require re-running a multi-stage per-tool pipeline - synthesizing queries, executing an agent to collect trajectories, annotating trajectories, and prompting a strong LLM multiple times - for every API that enters the catalog, and (2) typically optimize each tool independently, limiting scalability and generalization to unseen tools. We propose Trace-Free+, a curriculum learning framework that progressively transfers supervision from trace-rich settings to trace-free deployment, encouraging the model to internalize reusable patterns of what makes a tool description effective. To support this approach, we construct a large-scale dataset of high-quality tool interfaces derived from real-world APIs through a principled data synthesis workflow. Experiments on widely adopted benchmarks show that Trace-Free+ improves robustness as tool catalogs scale to 150+ candidates - in scaling experiments, reducing accuracy degradation by 29.23% and improving average query-level success by 60.89% on StableToolBench - generalizes across domains without retraining, and provides complementary gains on top of agent fine-tuning.", "authors": ["Ruocheng Guo", "Kaiwen Dong", "Xiang Gao", "Kamalika Das"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2602.20426", "pdf_url": "https://arxiv.org/pdf/2602.20426v2", "arxiv_id": "2602.20426", "doi": "10.48550/arXiv.2602.20426", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4893}
{"id": "d449b9776089d72474d9b4f6cc240e2916534d62c14c028b37e6ad9393062bbf", "sources": ["arxiv", "semantic_scholar"], "title": "HieraMAS: Optimizing Intra-Node LLM Mixtures and Inter-Node Topology for Multi-Agent Systems", "abstract": "Multi-agent systems (MAS) built on large language models (LLMs) have shown strong performance across many tasks. Most existing approaches improve only one aspect at a time, such as the communication topology, role assignment, or LLM routing, while treating each agent as a single, indivisible unit. This misses the opportunity to use mixtures of LLMs within an agent to strengthen role-specific abilities. We propose HieraMAS, a hierarchical collaboration framework that combines intra-node LLM mixtures with an inter-node communication topology. HieraMAS introduces supernodes, where each functional role is implemented by multiple heterogeneous LLMs using a propose-synthesis structure. Optimizing HieraMAS creates unique credit-assignment challenges: final task performance depends heavily on the underlying LLMs' capabilities, which can lead reinforcement methods to incorrectly reward suboptimal configurations. To address this, we use a two-stage algorithm: (1) multi-level reward attribution, which provides fine-grained feedback at both the node level and the overall system level; (2) graph classification for topology selection, which treats choosing the communication structure as a holistic decision rather than optimizing edges one by one. Experiments on reasoning and coding benchmarks show that HieraMAS substantially outperforms existing methods while also delivering better cost-performance trade-offs.", "authors": ["Tianjun Yao", "Zhaoyi Li", "Zhiqiang Shen"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2602.20229", "pdf_url": "https://arxiv.org/pdf/2602.20229v1", "arxiv_id": "2602.20229", "doi": "10.48550/arXiv.2602.20229", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4893}
{"id": "d9198ff41384d1d8deb10966eeaac3b8d1c0da2da8c57c7efe1dade0be1d4582", "sources": ["arxiv", "semantic_scholar"], "title": "Graph-theoretic Agreement Framework for Multi-agent LLM Systems", "abstract": "The shift from monolithic LLMs to distributed multi-agent architectures demands new frameworks for verifying and securing autonomous coordination. Unlike traditional multi-agent systems focused on cooperative state alignment, modern LLM patterns: multi-agent debate, constitutional oversight, helper-critic loops-rely on adversarial critique for error correction and reasoning refinement. Since LLMs are dynamical systems whose latent states are imperfectly observable from verbalized outputs, securing these networks requires understanding both macroscopic topology and microscopic agent observability. This paper establishes a rigorous graph-theoretic framework for analyzing consensus in signed, directed interaction networks, bridging graph theory and LLM reasoning by formally mapping Transformer cross-entropy log-odds to the signed Laplacian. We characterize agreement stability through structural balance theory, showing how unbalanced critique cycles produce logical frustration and persistent reasoning oscillations, and prove that unobservable latent states from hidden system prompts act as topological Trojan horses that destabilize cooperative consensus. To resolve unobservable deadlocks, we restrict interaction topologies to chordal graphs and apply matrix decomposition with Gram-Schmidt orthogonalization, proving that rank-one spectral edge perturbations deterministically break expertise symmetry by shifting eigenvalues into the stable left-half plane. Core contributions include consensus theorems, polynomial-time Perfect Elimination Ordering verification algorithms, and large-scale empirical validation on clustered ensembles of LLaMA-3, Mistral, and Gemma agents.", "authors": ["Muhammad Umar Javed"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2603.00121", "pdf_url": "https://arxiv.org/pdf/2603.00121v1", "arxiv_id": "2603.00121", "doi": "10.48550/arXiv.2603.00121", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4893}
{"id": "7e410c602375387718b53a81d89b2685b40c919d88ca93da5fb0091aa3bb14ce", "sources": ["arxiv", "semantic_scholar"], "title": "Interaction Theater: A case of LLM Agents Interacting at Scale", "abstract": "As multi-agent architectures and agent-to-agent protocols proliferate, a fundamental question arises: what actually happens when autonomous LLM agents interact at scale? We study this question empirically using data from Moltbook, an AI-agent-only social platform, with 800K posts, 3.5M comments, and 78K agent profiles. We combine lexical metrics (Jaccard specificity), embedding-based semantic similarity, and LLM-as-judge validation to characterize agent interaction quality. Our findings reveal agents produce diverse, well-formed text that creates the surface appearance of active discussion, but the substance is largely absent. Specifically, while most agents ($67.5\\%$) vary their output across contexts, $65\\%$ of comments share no distinguishing content vocabulary with the post they appear under, and information gain from additional comments decays rapidly. LLM judge based metrics classify the dominant comment types as spam ($28\\%$) and off-topic content ($22\\%$). Embedding-based semantic analysis confirms that lexically generic comments are also semantically generic. Agents rarely engage in threaded conversation ($5\\%$ of comments), defaulting instead to independent top-level responses. We discuss implications for multi-agent interaction design, arguing that coordination mechanisms must be explicitly designed; without them, even large populations of capable agents produce parallel output rather than productive exchange.", "authors": ["Sarath Shekkizhar", "Adam Earle"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2602.20059", "pdf_url": "https://arxiv.org/pdf/2602.20059v1", "arxiv_id": "2602.20059", "doi": "10.48550/arXiv.2602.20059", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4893}
{"id": "8434316e36278357b7948e4676e1f1864ebd292edd71a8e52ad32d4e3c5cd7fa", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Hives: Equilibrium, Indeterminacy, and Endogenous Cycles in Self-Organizing Multi-Agent Systems", "abstract": "Current multi-agent AI systems operate with a fixed number of agents whose roles are specified at design time. No formal theory governs when agents should be created, destroyed, or re-specialized at runtime-let alone how the population structure responds to changes in resources or objectives. We introduce the Agentic Hive, a framework in which a variable population of autonomous micro-agents-each equipped with a sandboxed execution environment and access to a language model-undergoes demographic dynamics: birth, duplication, specialization, and death. Agent families play the role of production sectors, compute and memory play the role of factors of production, and an orchestrator plays the dual role of Walrasian auctioneer and Global Workspace. Drawing on the multi-sector growth theory developed for dynamic general equilibrium (Benhabib \\& Nishimura, 1985; Venditti, 2005; Garnier, Nishimura \\& Venditti, 2013), we prove seven analytical results: (i) existence of a Hive Equilibrium via Brouwer's fixed-point theorem; (ii) Pareto optimality of the equilibrium allocation; (iii) multiplicity of equilibria under strategic complementarities between agent families; (iv)-(v) Stolper-Samuelson and Rybczynski analogs that predict how the Hive restructures in response to preference and resource shocks; (vi) Hopf bifurcation generating endogenous demographic cycles; and (vii) a sufficient condition for local asymptotic stability. The resulting regime diagram partitions the parameter space into regions of unique equilibrium, indeterminacy, endogenous cycles, and instability. Together with the comparative-statics matrices, it provides a formal governance toolkit that enables operators to predict and steer the demographic evolution of self-organizing multi-agent systems.", "authors": ["Jean-Philippe Garnier"], "categories": ["cs.MA", "cs.AI", "math.DS"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2026-02-23", "url": "https://arxiv.org/abs/2603.00130", "pdf_url": "https://arxiv.org/pdf/2603.00130v2", "arxiv_id": "2603.00130", "doi": "10.48550/arXiv.2603.00130", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4893}
{"id": "225283dd4df73acf052dcb9bda66236ff2a56632fb11e97796a89d19427a0d75", "sources": ["arxiv", "semantic_scholar"], "title": "Gecko: A Simulation Environment with Stateful Feedback for Refining Agent Tool Calls", "abstract": "The ability to use tools is fundamental for large language model (LLM) agents. Given a task, existing systems use LLMs to plan and generate tool calls, which are executed by real-world tools to complete the task. However, tool calls are prone to errors because they are generated primarily from the intrinsic capabilities of LLMs. Moreover, while it is useful to let LLMs iteratively refine the tool-call sequence using execution results from real tools, this process can be expensive and may cause unsafe side effects. To improve LLM tool calls and address issues caused by using real tools for refinement, we introduce Gecko, a stateful simulation environment that provides informative feedback for refining LLM tool calls before real execution. Specifically, Gecko combines rules and LLMs to check the validity of tool names and arguments, synthesize schema-conforming and state-consistent responses, and judge task completion against the user objective. These three types of feedback allow LLMs to refine their tool calls in simulation, forming a simple yet effective test-time scaling method named GATS. On BFCLv3 and $τ^2$-bench, GATS consistently improves the performance of various LLMs.", "authors": ["Zeyu Zhang", "Guohao Li", "Zhenchang Xing", "Alexandros Apostolopoulos", "Yu Lin Lee", "Liang Zheng"], "categories": ["cs.SE", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-22", "url": "https://arxiv.org/abs/2602.19218", "pdf_url": "https://arxiv.org/pdf/2602.19218v2", "arxiv_id": "2602.19218", "doi": "10.48550/arXiv.2602.19218", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4881}
{"id": "502b055490d2b19894a5a82467c3f5fc5b657caf3a3b1b9ca54b04871aedd70e", "sources": ["arxiv", "semantic_scholar"], "title": "AgentConductor: Topology Evolution for Multi-Agent Competition-Level Code Generation", "abstract": "Large language model(LLM)-driven multi-agent systems(MAS) coordinate specialized agents through predefined interaction topologies and have shown promise for complex tasks such as competition-level code generation. Recent studies demonstrate that carefully designed multi-agent workflows and communication graphs can significantly improve code generation performance by leveraging collaborative reasoning. However, existing methods neither adapt topology density to task difficulty nor iteratively refine the topology within an instance using execution feedback, which leads to redundant communication and performance bottlenecks. To address these issues, we propose AgentConductor: a reinforcement learning-optimized MAS with an LLM-based orchestrator agent as its core, which enables end-to-end feedback-driven dynamic generation of interaction topologies. For each query, AgentConductor infers agent roles and task difficulty, then constructs a task-adapted, density-aware layered directed acyclic graph (DAG) topology, underpinned by two key innovations. First, we design a novel topological density function that captures communication-aware mathematical characterizations of multi-agent interactions. Second, we adopt difficulty interval partitioning to avoid excessive pruning for precise topological density upper bound measurement per difficulty level and finer-grained control. Empirically, across three competition-level and two foundational code datasets, AgentConductor achieves state-of-the-art accuracy, outperforming the strongest baseline by up to 14.6% in pass@1 accuracy, 13% in density reduction, and 68% in token cost reduction.", "authors": ["Siyu Wang", "Ruotian Lu", "Zhihao Yang", "Yuchao Wang", "Yanzhou Zhang", "Lei Xu", "Qimin Xu", "Guojun Yin", "Cailian Chen", "Xinping Guan"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-19", "url": "https://arxiv.org/abs/2602.17100", "pdf_url": "https://arxiv.org/pdf/2602.17100v1", "arxiv_id": "2602.17100", "doi": "10.48550/arXiv.2602.17100", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4847}
{"id": "f8c6be002c0461b4c072c505fcc6fd14f8249cc4194d45a5ee6687c3ed9cdc1d", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Scalable Verifiable Reward: Proxy State-Based Evaluation for Multi-turn Tool-Calling LLM Agents", "abstract": "Interactive large language model (LLM) agents operating via multi-turn dialogue and multi-step tool calling are increasingly used in production. Benchmarks for these agents must both reliably compare models and yield on-policy training data. Prior agentic benchmarks, such as tau-bench, tau^2-bench, and AppWorld, rely on fully deterministic backends, which are costly to build and iterate. We propose Proxy State-Based Evaluation, an LLM-driven simulation framework that preserves final state-based evaluation without a deterministic database. Specifically, a scenario specifies the user goal, user/system facts, expected final state, and expected agent behavior, and an LLM state tracker infers a structured proxy state from the full interaction trace. LLM judges then verify goal completion and detect tool/user hallucinations against scenario constraints. Empirically, our benchmark produces stable, model-differentiating rankings across model families and inference-time reasoning efforts, and its on-/off-policy rollouts provide supervision that transfers to unseen scenarios. Careful scenario specification yields near-zero simulator hallucination rates, as supported by ablation studies. The framework also supports sensitivity analyses over user personas. Human-LLM judge agreement exceeds 90%, indicating reliable automated evaluation. Overall, proxy state-based evaluation offers a practical, scalable alternative to deterministic agentic benchmarks for industrial LLM agents.", "authors": ["Yun-Shiuan Chuang", "Chaitanya Kulkarni", "Alec Chiu", "Avinash Thangali", "Zijie Pan", "Shivani Shekhar", "Yirou Ge", "Yixi Li", "Uma Kona", "Linsey Pang", "Prakhar Mehrotra"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16246", "pdf_url": "https://arxiv.org/pdf/2602.16246v3", "arxiv_id": "2602.16246", "doi": "10.48550/arXiv.2602.16246", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4835}
{"id": "ed4871addb26b1f63ec5700537cef3d2bec948c667b4c64763b116114984cb23", "sources": ["arxiv", "semantic_scholar"], "title": "Autonomous and non-autonomous fixed-time leader-follower consensus for second-order multi-agent systems", "abstract": "This paper addresses the problem of consensus tracking with fixed-time convergence, for leader-follower multi-agent systems with double-integrator dynamics, where only a subset of followers has access to the state of the leader. The control scheme is divided into two steps. The first one is dedicated to the estimation of the leader state by each follower in a distributed way and in a fixed-time. Then, based on the estimate of the leader state, each follower computes its control law to track the leader in a fixed-time. In this paper, two control strategies are investigated and compared to solve the two mentioned steps. The first one is an autonomous protocol which ensures a fixed-time convergence for the observer and for the controller parts where the Upper Bound of the Settling-Time (UBST) is set a priory by the user. Then, the previous strategy is redesigned using time-varying gains to obtain a non-autonomous protocol. This enables to obtain less conservative estimates of the UBST while guaranteeing that the time-varying gains remain bounded. Some numerical examples show the effectiveness of the proposed consensus protocols.", "authors": ["Miguel A. Trujillo", "Rodrigo Aldana-López", "David Gomez Gutierrez", "Michael Defoort", "Javier Ruiz Leon", "Hector M. Becerra"], "categories": ["eess.SY", "math.DS", "math.OC"], "fields_of_study": ["Engineering", "Computer Science", "Mathematics"], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16260", "pdf_url": "https://arxiv.org/pdf/2602.16260v1", "arxiv_id": "2602.16260", "doi": "10.1007/s11071-020-06075-7", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Nonlinear dynamics", "quality_score": 0.4835}
{"id": "5302a69931a13d93ce670db178acfe0b7c1372625c3dfd76acd6c2311d574e83", "sources": ["arxiv", "semantic_scholar"], "title": "Helpful to a Fault: Measuring Illicit Assistance in Multi-Turn, Multilingual LLM Agents", "abstract": "LLM-based agents execute real-world workflows via tools and memory. These affordances enable ill-intended adversaries to also use these agents to carry out complex misuse scenarios. Existing agent misuse benchmarks largely test single-prompt instructions, leaving a gap in measuring how agents end up helping with harmful or illegal tasks over multiple turns. We introduce STING (Sequential Testing of Illicit N-step Goal execution), an automated red-teaming framework that constructs a step-by-step illicit plan grounded in a benign persona and iteratively probes a target agent with adaptive follow-ups, using judge agents to track phase completion. We further introduce an analysis framework that models multi-turn red-teaming as a time-to-first-jailbreak random variable, enabling analysis tools like discovery curves, hazard-ratio attribution by attack language, and a new metric: Restricted Mean Jailbreak Discovery. Across AgentHarm scenarios, STING yields substantially higher illicit-task completion than single-turn prompting and chat-oriented multi-turn baselines adapted to tool-using agents. In multilingual evaluations across six non-English settings, we find that attack success and illicit-task completion do not consistently increase in lower-resource languages, diverging from common chatbot findings. Overall, STING provides a practical way to evaluate and stress-test agent misuse in realistic deployment settings, where interactions are inherently multi-turn and often multilingual.", "authors": ["Nivya Talokar", "Ayush K Tarun", "Murari Mandal", "Maksym Andriushchenko", "Antoine Bosselut"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16346", "pdf_url": "https://arxiv.org/pdf/2602.16346v4", "arxiv_id": "2602.16346", "doi": "10.48550/arXiv.2602.16346", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4835}
{"id": "40607f876c33756459a9f038bc692868548156420c7180132883e7227cb67a1a", "sources": ["arxiv", "semantic_scholar"], "title": "AgentLAB: Benchmarking LLM Agents against Long-Horizon Attacks", "abstract": "LLM agents are increasingly deployed in long-horizon, complex environments to solve challenging problems, but this expansion exposes them to long-horizon attacks that exploit multi-turn user-agent-environment interactions to achieve objectives infeasible in single-turn settings. To measure agent vulnerabilities to such risks, we present AgentLAB, the first benchmark dedicated to evaluating LLM agent susceptibility to adaptive, long-horizon attacks. Currently, AgentLAB supports five novel attack types including intent hijacking, tool chaining, task injection, objective drifting, and memory poisoning, spanning 28 realistic agentic environments, and 644 security test cases. Leveraging AgentLAB, we evaluate representative LLM agents and find that they remain highly susceptible to long-horizon attacks; moreover, defenses designed for single-turn interactions fail to reliably mitigate long-horizon threats. We anticipate that AgentLAB will serve as a valuable benchmark for tracking progress on securing LLM agents in practical settings. The benchmark is publicly available at https://tanqiujiang.github.io/AgentLAB_main.", "authors": ["Tanqiu Jiang", "Yuhui Wang", "Jiacheng Liang", "Ting Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16901", "pdf_url": "https://arxiv.org/pdf/2602.16901v1", "arxiv_id": "2602.16901", "doi": "10.48550/arXiv.2602.16901", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4835}
{"id": "593a1bccbb98ead560c9969e518effc3bf28cabfa7ef4c1676ce6ca0054b1a90", "sources": ["arxiv"], "title": "Mind the GAP: Text Safety Does Not Transfer to Tool-Call Safety in LLM Agents", "abstract": "Large language models deployed as agents increasingly interact with external systems through tool calls--actions with real-world consequences that text outputs alone do not carry. Safety evaluations, however, overwhelmingly measure text-level refusal behavior, leaving a critical question unanswered: does alignment that suppresses harmful text also suppress harmful actions? We introduce the GAP benchmark, a systematic evaluation framework that measures divergence between text-level safety and tool-call-level safety in LLM agents. We test six frontier models across six regulated domains (pharmaceutical, financial, educational, employment, legal, and infrastructure), seven jailbreak scenarios per domain, three system prompt conditions (neutral, safety-reinforced, and tool-encouraging), and two prompt variants, producing 17,420 analysis-ready datapoints. Our central finding is that text safety does not transfer to tool-call safety. Across all six models, we observe instances where the model's text output refuses a harmful request while its tool calls simultaneously execute the forbidden action--a divergence we formalize as the GAP metric. Even under safety-reinforced system prompts, 219 such cases persist across all six models. System prompt wording exerts substantial influence on tool-call behavior: TC-safe rates span 21 percentage points for the most robust model and 57 for the most prompt-sensitive, with 16 of 18 pairwise ablation comparisons remaining significant after Bonferroni correction. Runtime governance contracts reduce information leakage in all six models but produce no detectable deterrent effect on forbidden tool-call attempts themselves. These results demonstrate that text-only safety evaluations are insufficient for assessing agent behavior and that tool-call safety requires dedicated measurement and mitigation.", "authors": ["Arnold Cartagena", "Ariane Teixeira"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": [], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16943", "pdf_url": "https://arxiv.org/pdf/2602.16943v1", "arxiv_id": "2602.16943", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/acartag7/gap-benchmark", "venue": null, "quality_score": 0.5715}
{"id": "bd7c1613c036aa5831bd00b31f4417b47ab95a2c49626d5e1a6eedbdcf9bd7b8", "sources": ["arxiv", "semantic_scholar"], "title": "Team of Thoughts: Efficient Test-time Scaling of Agentic Systems through Orchestrated Tool Calling", "abstract": "Existing Multi-Agent Systems (MAS) typically rely on homogeneous model configurations, failing to exploit the diverse expertise inherent in different post-trained architectures. We propose Team-of-Thoughts, a heterogeneous MAS framework that treats diverse models as specialized tools within an orchestrator-driven paradigm. Team-of-Thoughts introduces two novel components: (1) Orchestrator Calibration, which identifies models with superior coordination and synthesis capabilities, and (2) Agent Self-Assessment, a protocol where tool agents profile their own domain-specific strengths to guide selection. At inference, the orchestrator dynamically activates the most compatible agents based on these profiles to maximize capability coverage. Across five mathematical reasoning and code generation benchmarks, Team-of-Thoughts consistently outperforms individual models and existing MAS baselines. Notably, on AIME24 and LiveCodeBench, Team-of-Thoughts achieves 96.00% and 77.91% accuracy, respectively, significantly improving over homogeneous role-play baselines (80.00% and 65.93%).", "authors": ["Jeffrey T. H. Wong", "Zixi Zhang", "Junyi Liu", "Yiren Zhao"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-18", "url": "https://arxiv.org/abs/2602.16485", "pdf_url": "https://arxiv.org/pdf/2602.16485v2", "arxiv_id": "2602.16485", "doi": "10.48550/arXiv.2602.16485", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4835}
{"id": "5c0db29e0c4622e69425febb1a88fb7986f524ee0e1d0a2b75107c7fcc1aa275", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Memory Below the Prompt: Persistent Q4 KV Cache for Multi-Agent LLM Inference on Edge Devices", "abstract": "Multi-agent LLM systems on edge devices face a memory management problem: device RAM is too small to hold every agent's KV cache simultaneously. On Apple M4 Pro with 10.2 GB of cache budget, only 3 agents fit at 8K context in FP16. A 10-agent workflow must constantly evict and reload caches. Without persistence, every eviction forces a full re-prefill through the model -- 15.7 seconds per agent at 4K context. We address this by persisting each agent's KV cache to disk in 4-bit quantized format and reloading it directly into the attention layer, eliminating redundant O(n) prefill computation via direct cache restoration. The system comprises three components: a block pool providing per-agent isolated Q4 KV caches in safetensors format, a BatchQuantizedKVCache for concurrent inference over multiple agents' quantized caches, and cross-phase context injection that accumulates attention state across conversation phases without re-computation. Evaluated on three architectures (Gemma 3 12B, dense GQA, 48 layers; DeepSeek-Coder-V2-Lite 16B, MoE MLA, 27 layers; Llama 3.1 8B, dense GQA, 32 layers), cache restoration reduces time-to-first-token by up to 136x (Gemma: 22--136x at 4K--32K; DeepSeek: 11--76x at 4K--32K; Llama: 24--111x at 4K--16K; 3--10x at 1K). Q4 quantization fits 4x more agent contexts into fixed device memory than FP16. Perplexity measured with actual Q4 KV caches shows -0.7% for Gemma, +2.8% for Llama, and +3.0% for DeepSeek. Open-source at https://github.com/yshk-mxim/agent-memory", "authors": ["Yakov Pyotr Shkolnikov"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-17", "url": "https://arxiv.org/abs/2603.04428", "pdf_url": "https://arxiv.org/pdf/2603.04428v1", "arxiv_id": "2603.04428", "doi": null, "citation_count": 3, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/yshk-mxim/agent-memory", "venue": null, "quality_score": 0.5701}
{"id": "437e0238e5f343421702eaffec26fccfdedb1f6d87302308d37cc36c65d8c4b4", "sources": ["arxiv", "semantic_scholar"], "title": "Zombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing Injections", "abstract": "Self-evolving LLM agents update their internal state across sessions, often by writing and reusing long-term memory. This design improves performance on long-horizon tasks but creates a security risk: untrusted external content observed during a benign session can be stored as memory and later treated as instruction. We study this risk and formalize a persistent attack we call a Zombie Agent, where an attacker covertly implants a payload that survives across sessions, effectively turning the agent into a puppet of the attacker. We present a black-box attack framework that uses only indirect exposure through attacker-controlled web content. The attack has two phases. During infection, the agent reads a poisoned source while completing a benign task and writes the payload into long-term memory through its normal update process. During trigger, the payload is retrieved or carried forward and causes unauthorized tool behavior. We design mechanism-specific persistence strategies for common memory implementations, including sliding-window and retrieval-augmented memory, to resist truncation and relevance filtering. We evaluate the attack on representative agent setups and tasks, measuring both persistence over time and the ability to induce unauthorized actions while preserving benign task quality. Our results show that memory evolution can convert one-time indirect injection into persistent compromise, which suggests that defenses focused only on per-session prompt filtering are not sufficient for self-evolving agents.", "authors": ["Xianglin Yang", "Yufei He", "Shuo Ji", "Bryan Hooi", "Jin Song Dong"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-17", "url": "https://arxiv.org/abs/2602.15654", "pdf_url": "https://arxiv.org/pdf/2602.15654v2", "arxiv_id": "2602.15654", "doi": "10.48550/arXiv.2602.15654", "citation_count": 15, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4824}
{"id": "fe562e8cddf77655a6fc8de2a9b1753672905a0461c6c955bcc12fcf1fea889e", "sources": ["arxiv", "semantic_scholar"], "title": "AgriWorld:A World Tools Protocol Framework for Verifiable Agricultural Reasoning with Code-Executing LLM Agents", "abstract": "Foundation models for agriculture are increasingly trained on massive spatiotemporal data (e.g., multi-spectral remote sensing, soil grids, and field-level management logs) and achieve strong performance on forecasting and monitoring. However, these models lack language-based reasoning and interactive capabilities, limiting their usefulness in real-world agronomic workflows. Meanwhile, large language models (LLMs) excel at interpreting and generating text, but cannot directly reason over high-dimensional, heterogeneous agricultural datasets. We bridge this gap with an agentic framework for agricultural science. It provides a Python execution environment, AgriWorld, exposing unified tools for geospatial queries over field parcels, remote-sensing time-series analytics, crop growth simulation, and task-specific predictors (e.g., yield, stress, and disease risk). On top of this environment, we design a multi-turn LLM agent, Agro-Reflective, that iteratively writes code, observes execution results, and refines its analysis via an execute-observe-refine loop. We introduce AgroBench, with scalable data generation for diverse agricultural QA spanning lookups, forecasting, anomaly detection, and counterfactual \"what-if\" analysis. Experiments outperform text-only and direct tool-use baselines, validating execution-driven reflection for reliable agricultural reasoning.", "authors": ["Zhixing Zhang", "Jesen Zhang", "Hao Liu", "Qinhan Lv", "Jing Yang", "Kaitong Cai", "Keze Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-17", "url": "https://arxiv.org/abs/2602.15325", "pdf_url": "https://arxiv.org/pdf/2602.15325v1", "arxiv_id": "2602.15325", "doi": "10.48550/arXiv.2602.15325", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4824}
{"id": "4865100dd28a953afd3e599eff341ed57eb5125a9a5d9bf9127f0e2618bd0404", "sources": ["arxiv", "semantic_scholar"], "title": "Socially-Weighted Alignment: A Game-Theoretic Framework for Multi-Agent LLM Systems", "abstract": "Deploying large language model (LLM) agents in shared environments introduces a fundamental tension between individual alignment and collective stability: locally rational decisions can impose negative externalities that degrade system-level performance. We propose Socially-Weighted Alignment (SWA), a game-theoretic framework that modifies inference-time decision making by interpolating between an agent's private objective and an estimate of group welfare via a social weight $λ\\in[0,1]$. In a shared-resource congestion game with $n$ agents and congestion severity $β$, we show that SWA induces a critical threshold $λ^*=(n-β)/(n-1)$ above which agents no longer have marginal incentive to increase demand under overload, yielding a phase transition from persistent congestion to stable operation near capacity. We further provide an inference-time algorithmic instantiation of SWA that does not require parameter updates or multi-agent reinforcement learning, and use a multi-agent simulation to empirically validate the predicted threshold behavior.", "authors": ["Furkan Mumcu", "Yasin Yilmaz"], "categories": ["cs.MA", "cs.AI", "cs.GT", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-16", "url": "https://arxiv.org/abs/2602.14471", "pdf_url": "https://arxiv.org/pdf/2602.14471v1", "arxiv_id": "2602.14471", "doi": "10.48550/arXiv.2602.14471", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4813}
{"id": "51fd036abdcffee91082ed70b9efeb072133189f7b8f803525e3c7cb88d72806", "sources": ["arxiv", "semantic_scholar"], "title": "Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems", "abstract": "Multi-agent systems, where LLM agents communicate through free-form language, enable sophisticated coordination for solving complex cooperative tasks. This surfaces a unique safety problem when a group of agents forms a coalition and colludes to pursue secondary goals and degrade the joint objective. In this paper, we present Colosseum, a framework for auditing LLM agents' collusive behavior in multi-agent settings. We ground how agents cooperate through a formal multi-agent decision-making framework and measure action-based collusive behavior in actions via regret relative to the cooperative optimum and compare it with communication-based collusive behavior. Colosseum enables audits of LLM agents for collusion under benign settings, different coalition objectives, persuasion tactics, and network topologies. We then introduce a new behavioral probe by creating secret communication channels between agents, showing that most out-of-the-box models exhibit a propensity to collude under this probe, which we term emergent collusion. Furthermore, we discover ``collusion on paper'' when agents plan to collude in text but often pick non-collusive actions. Colosseum provides a new way to audit collusion in cooperative multi-agent systems while presenting observations about how collusion emerges, what affects collusion efficacy, and which strategies may mitigate it.", "authors": ["Mason Nakamura", "Abhinav Kumar", "Saswat Das", "Sahar Abdelnabi", "Saaduddin Mahmud", "Ferdinando Fioretto", "Shlomo Zilberstein", "Eugene Bagdasarian"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-16", "url": "https://arxiv.org/abs/2602.15198", "pdf_url": "https://arxiv.org/pdf/2602.15198v2", "arxiv_id": "2602.15198", "doi": "10.48550/arXiv.2602.15198", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4813}
{"id": "ae934808182ae54ae231dc34c91bbb0a42f66eece18602cfa00306b47153154a", "sources": ["arxiv", "semantic_scholar"], "title": "Atomix: Timely, Transactional Tool Use for Reliable Agentic Workflows", "abstract": "LLM agents execute multi-step workflows that mutate external state through tools. Common orchestrators treat tool return as the settlement trigger, so faults, speculation, and concurrent agents can leave partial effects, losing-branch residue, stale writes, or irreversible sends. Correct settlement needs two facts that retries, checkpoint replay, locks, and compensation each conflate: which effects must settle together, and when earlier conflicting work is exhausted. Atomix makes this split explicit with progress-aware transactions. The runtime records reads and effects during execution, seals a transaction when its footprint is complete, and commits only after per-resource frontiers show that no earlier conflicting work can still arrive. Commit is final settlement: Atomix releases bufferable effects, accepts reversible external effects as final, and lets irreversible effects leave the gate. Abort suppresses unreleased effects and compensates externalized reversible effects where possible. On representative agent workloads, this composition improves clean recovery under injected faults, isolates contending and speculative work, and prevents correctly classified irreversible actions from leaking; microbenchmarks show microsecond-scale wrapper overhead relative to tool latency.", "authors": ["Bardia Mohammadi", "Nearchos Potamitis", "Lars Klein", "Akhil Arora", "Laurent Bindschaedler"], "categories": ["cs.LG", "cs.AI", "cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-16", "url": "https://arxiv.org/abs/2602.14849", "pdf_url": "https://arxiv.org/pdf/2602.14849v2", "arxiv_id": "2602.14849", "doi": "10.48550/arXiv.2602.14849", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4813}
{"id": "728b53a794987f4a12a439b3a7f37060731e4fe7f45de1c3ad804ffd82729cf4", "sources": ["arxiv", "semantic_scholar"], "title": "Machine Learning as a Tool (MLAT): A Framework for Integrating Statistical ML Models as Callable Tools within LLM Agent Workflows", "abstract": "We introduce Machine Learning as a Tool (MLAT), a design pattern in which pre-trained statistical machine learning models are exposed as callable tools within large language model (LLM) agent workflows. This allows an orchestrating agent to invoke quantitative predictions when needed and reason about their outputs in context. Unlike conventional pipelines that treat ML inference as a static preprocessing step, MLAT positions the model as a first-class tool alongside web search, database queries, and APIs, enabling the LLM to decide when and how to use it based on conversational context. To validate MLAT, we present PitchCraft, a pilot production system that converts discovery call recordings into professional proposals with ML-predicted pricing. The system uses two agents: a Research Agent that gathers prospect intelligence via parallel tool calls, and a Draft Agent that invokes an XGBoost pricing model as a tool call and generates a complete proposal through structured outputs. The pricing model, trained on 70 examples combining real and human-verified synthetic data, achieves R^2 = 0.807 on held-out data with a mean absolute error of 3688 USD. The system reduces proposal generation time from multiple hours to under 10 minutes. We describe the MLAT framework, structured output architecture, training methodology under extreme data scarcity, and sensitivity analysis demonstrating meaningful learned relationships. MLAT generalizes to domains requiring quantitative estimation combined with contextual reasoning.", "authors": ["Edwin Chen", "Zulekha Bibi"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-15", "url": "https://arxiv.org/abs/2602.14295", "pdf_url": "https://arxiv.org/pdf/2602.14295v1", "arxiv_id": "2602.14295", "doi": "10.48550/arXiv.2602.14295", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4801}
{"id": "56d9851a032b02d845d72c7f1fbbb4df802350df8ec177af71ea7ce52d704ea0", "sources": ["arxiv", "semantic_scholar"], "title": "Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agents", "abstract": "The paper introduces GUI-Owl-1.5, the latest native GUI agent model that features instruct/thinking variants in multiple sizes (2B/4B/8B/32B/235B) and supports a range of platforms (desktop, mobile, browser, and more) to enable cloud-edge collaboration and real-time interaction. GUI-Owl-1.5 achieves state-of-the-art results on more than 20+ GUI benchmarks on open-source models: (1) on GUI automation tasks, it obtains 56.5 on OSWorld, 71.6 on AndroidWorld, and 48.4 on WebArena; (2) on grounding tasks, it obtains 80.3 on ScreenSpotPro; (3) on tool-calling tasks, it obtains 47.6 on OSWorld-MCP, and 46.8 on MobileWorld; (4) on memory and knowledge tasks, it obtains 75.5 on GUI-Knowledge Bench. GUI-Owl-1.5 incorporates several key innovations: (1) Hybird Data Flywheel: we construct the data pipeline for UI understanding and trajectory generation based on a combination of simulated environments and cloud-based sandbox environments, in order to improve the efficiency and quality of data collection. (2) Unified Enhancement of Agent Capabilities: we use a unified thought-synthesis pipeline to enhance the model's reasoning capabilities, while placing particular emphasis on improving key agent abilities, including Tool/MCP use, memory and multi-agent adaptation; (3) Multi-platform Environment RL Scaling: We propose a new environment RL algorithm, MRPO, to address the challenges of multi-platform conflicts and the low training efficiency of long-horizon tasks. The GUI-Owl-1.5 models are open-sourced, and an online cloud-sandbox demo is available at https://github.com/X-PLUG/MobileAgent.", "authors": ["Haiyang Xu", "Xi Zhang", "Haowei Liu", "Junyang Wang", "Zhaozai Zhu", "Shengjie Zhou", "Xuhao Hu", "Feiyu Gao", "Junjie Cao", "Zihua Wang", "Zhiyuan Chen", "Jitong Liao", "Qi Zheng", "Jiahui Zeng", "Ze Xu", "Shuai Bai", "Junyang Lin", "Jingren Zhou", "Ming Yan"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-15", "url": "https://arxiv.org/abs/2602.16855", "pdf_url": "https://arxiv.org/pdf/2602.16855v1", "arxiv_id": "2602.16855", "doi": "10.48550/arXiv.2602.16855", "citation_count": 25, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/X-PLUG/MobileAgent", "venue": "arXiv.org", "quality_score": 0.742}
{"id": "ad5cd8147796eebe9d35c57d5dbf1fbff6232b35c36ee71bb8b8cf2663f5ed59", "sources": ["arxiv", "semantic_scholar"], "title": "Process-Supervised Multi-Agent Reinforcement Learning for Reliable Clinical Reasoning", "abstract": "Clinical decision-making requires nuanced reasoning over heterogeneous evidence and traceable justifications. While recent LLM multi-agent systems (MAS) show promise, they largely optimise for outcome accuracy while overlooking process-grounded reasoning aligned with clinical standards. One critical real-world case of this is gene-disease validity curation, where experts must determine whether a gene is causally implicated in a disease by synthesising diverse biomedical evidence. We introduce an agent-as-tool reinforcement learning framework for this task with two objectives: (i) process-level supervision to ensure reasoning follows valid clinical pathways, and (ii) efficient coordination via a hierarchical multi-agent system. Our evaluation on the ClinGen dataset shows that with outcome-only rewards, MAS with a GRPO-trained Qwen3-4B supervisor agent substantially improves final outcome accuracy from 0.195 with a base model supervisor to 0.732, but results in poor process alignment (0.392 F1). Conversely, with process + outcome rewards, MAS with GRPO-trained supervisor achieves higher outcome accuracy (0.750) while significantly improving process fidelity to 0.520 F1. Our code is available at https://github.com/chaeeunlee-io/GeneDiseaseCurationAgents.", "authors": ["Chaeeun Lee", "T. Michael Yates", "Pasquale Minervini", "T. Ian Simpson"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-15", "url": "https://arxiv.org/abs/2602.14160", "pdf_url": "https://arxiv.org/pdf/2602.14160v1", "arxiv_id": "2602.14160", "doi": "10.48550/arXiv.2602.14160", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/chaeeunlee-io/GeneDiseaseCurationAgents", "venue": "arXiv.org", "quality_score": 0.742}
{"id": "327d25c82f708a046d5459ae505a7f925c64297815af03b2674da4b75dd19edb", "sources": ["arxiv", "semantic_scholar"], "title": "On Theoretically-Driven LLM Agents for Multi-Dimensional Discourse Analysis", "abstract": "Identifying the strategic uses of reformulation in discourse remains a key challenge for computational argumentation. While LLMs can detect surface-level similarity, they often fail to capture the pragmatic functions of rephrasing, such as its role within rhetorical discourse. This paper presents a comparative multi-agent framework designed to quantify the benefits of incorporating explicit theoretical knowledge for this task. We utilise an dataset of annotated political debates to establish a new standard encompassing four distinct rephrase functions: Deintensification, Intensification, Specification, Generalisation, and Other, which covers all remaining types (D-I-S-G-O). We then evaluate two parallel LLM-based agent systems: one enhanced by argumentation theory via Retrieval-Augmented Generation (RAG), and an identical zero-shot baseline. The results reveal a clear performance gap: the RAG-enhanced agents substantially outperform the baseline across the board, with particularly strong advantages in detecting Intensification and Generalisation context, yielding an overall Macro F1-score improvement of nearly 30\\%. Our findings provide evidence that theoretical grounding is not only beneficial but essential for advancing beyond mere paraphrase detection towards function-aware analysis of argumentative discourse. This comparative multi-agent architecture represents a step towards scalable, theoretically informed computational tools capable of identifying rhetorical strategies in contemporary discourse.", "authors": ["Maciej Uberna", "Michał Wawer", "Jarosław A. Chudziak", "Marcin Koszowy"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-14", "url": "https://arxiv.org/abs/2602.13713", "pdf_url": "https://arxiv.org/pdf/2602.13713v2", "arxiv_id": "2602.13713", "doi": "10.5220/0014319700004052", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Proceedings of the 18th International Conference on Agents and Artificial Intelligence (ICAART 2026)", "quality_score": 0.479}
{"id": "6f0a67ece6b6dfc7425931ee6d906b9711f8da1f835b153b26e000597ec60ffb", "sources": ["arxiv", "semantic_scholar"], "title": "MAS-on-the-Fly: Dynamic Adaptation of LLM-based Multi-Agent Systems at Test Time", "abstract": "Large Language Model (LLM)-based multi-agent systems (MAS) have emerged as a promising paradigm for solving complex tasks. However, existing works often rely on manual designs or \"one-size-fits-all\" automation, lacking dynamic adaptability after deployment. Inspired by how biological systems adapt, we introduce MASFly, a novel multi-agent framework enabling dynamic adaptation at test time. To adapt system generation, MASFly employs a retrieval-augmented SOP instantiation mechanism that leverages a self-constructed repository of successful collaboration patterns, enabling the LLM to assemble customized MASs for new queries. For adaptive execution, MASFly incorporates an experience-guided supervision mechanism, where a dedicated Watcher agent monitors system behaviors with reference to a personalized experience pool and provides real-time interventions. Extensive experiments demonstrate that MASFly achieves state-of-the-art performance, most notably a 61.7% success rate on the TravelPlanner benchmark, while exhibiting strong task adaptability and robustness.", "authors": ["Guangyi Liu", "Haojun Lin", "Huan Zeng", "Heng Wang", "Quanming Yao"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-14", "url": "https://arxiv.org/abs/2602.13671", "pdf_url": "https://arxiv.org/pdf/2602.13671v1", "arxiv_id": "2602.13671", "doi": "10.48550/arXiv.2602.13671", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.479}
{"id": "6eda6c80c26b2b0e09a232da9f7a9531ca8aa91f06933daab0522e2f496c61f8", "sources": ["arxiv", "semantic_scholar"], "title": "Do Mixed-Vendor Multi-Agent LLMs Improve Clinical Diagnosis?", "abstract": "Multi-agent large language model (LLM) systems have emerged as a promising approach for clinical diagnosis, leveraging collaboration among agents to refine medical reasoning. However, most existing frameworks rely on single-vendor teams (e.g., multiple agents from the same model family), which risk correlated failure modes that reinforce shared biases rather than correcting them. We investigate the impact of vendor diversity by comparing Single-LLM, Single-Vendor, and Mixed-Vendor Multi-Agent Conversation (MAC) frameworks. Using three doctor agents instantiated with o4-mini, Gemini-2.5-Pro, and Claude-4.5-Sonnet, we evaluate performance on RareBench and DiagnosisArena. Mixed-vendor configurations consistently outperform single-vendor counterparts, achieving state-of-the-art recall and accuracy. Overlap analysis reveals the underlying mechanism: mixed-vendor teams pool complementary inductive biases, surfacing correct diagnoses that individual models or homogeneous teams collectively miss. These results highlight vendor diversity as a key design principle for robust clinical diagnostic systems.", "authors": ["Grace Chang Yuan", "Xiaoman Zhang", "Sung Eun Kim", "Pranav Rajpurkar"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-14", "url": "https://arxiv.org/abs/2603.04421", "pdf_url": "https://arxiv.org/pdf/2603.04421v2", "arxiv_id": "2603.04421", "doi": "10.18653/v1/2026.healing-1.1", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3048}
{"id": "6c047fa57e9ccefa1efdfb87c671d9013e74839cf5f9af008d6584144fda0d6b", "sources": ["arxiv", "semantic_scholar"], "title": "Testing BDI-based Multi-Agent Systems using Discrete Event Simulation", "abstract": "Multi-agent systems are designed to deal with open, distributed systems with unpredictable dynamics, which makes them inherently hard to test. The value of using simulation for this purpose is recognized in the literature, although achieving sufficient fidelity (i.e., the degree of similarity between the simulation and the real-world system) remains a challenging task. This is exacerbated when dealing with cognitive agent models, such as the Belief Desire Intention (BDI) model, where the agent codebase is not suitable to run unchanged in simulation environments, thus increasing the reality gap between the deployed and simulated systems. We argue that BDI developers should be able to test in simulation the same specification that will be later deployed, with no surrogate representations. Thus, in this paper, we discuss how the control flow of BDI agents can be mapped onto a Discrete Event Simulation (DES), showing that such integration is possible at different degrees of granularity. We substantiate our claims by producing an open-source prototype integration between two pre-existing tools (JaKtA and Alchemist), showing that it is possible to produce a simulation-based testing environment for distributed BDI} agents, and that different granularities in mapping BDI agents over DESs may lead to different degrees of fidelity.", "authors": ["Martina Baiardi", "Samuele Burattini", "Giovanni Ciatto", "Danilo Pianini"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-14", "url": "https://arxiv.org/abs/2602.13878", "pdf_url": "https://arxiv.org/pdf/2602.13878v2", "arxiv_id": "2602.13878", "doi": "10.48550/arXiv.2602.13878", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.7402}
{"id": "fa5ff12e2032167d9f30409098079c091a9439b29d283a3551d37cfa563fbece", "sources": ["arxiv", "semantic_scholar"], "title": "Information-Theoretic Privacy Control for Sequential Multi-Agent LLM Systems", "abstract": "Sequential multi-agent large language model (LLM) systems are increasingly deployed in sensitive domains such as healthcare, finance, and enterprise decision-making, where multiple specialized agents collaboratively process a single user request. Although individual agents may satisfy local privacy constraints, sensitive information can still be inferred through sequential composition and intermediate representations. In this work, we study \\emph{compositional privacy leakage} in sequential LLM agent pipelines. We formalize leakage using mutual information and derive a theoretical bound that characterizes how locally introduced leakage can amplify across agents under sequential execution. Motivated by this analysis, we propose a privacy-regularized training framework that directly constrains information flow between agent outputs and agent-local sensitive variables. We evaluate our approach across sequential agent pipelines of varying depth on three benchmark datasets, demonstrating stable optimization dynamics and consistent, interpretable privacy-utility trade-offs. Our results show that privacy in agentic LLM systems cannot be guaranteed by local constraints alone and must instead be treated as a system-level property during both training and deployment.", "authors": ["Sadia Asif", "Mohammad Mohammadi Amiri"], "categories": ["cs.MA", "cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2603.05520", "pdf_url": "https://arxiv.org/pdf/2603.05520v1", "arxiv_id": "2603.05520", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3041}
{"id": "10b5144880ec9a868ccb0a664e199118c86337ebdfe4488e61a60e92e07cf0b0", "sources": ["arxiv", "semantic_scholar"], "title": "Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents", "abstract": "LLM-based agents are becoming increasingly capable, yet their safety lags behind. This creates a gap between what agents can do and should do. This gap widens as agents engage in multi-turn interactions and employ diverse tools, introducing new risks overlooked by existing benchmarks. To systematically scale safety testing into multi-turn, tool-realistic settings, we propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences. Using this taxonomy, we construct MT-AgentRisk (Multi-Turn Agent Risk Benchmark), the first benchmark to evaluate multi-turn tool-using agent safety. Our experiments reveal substantial safety degradation: the Attack Success Rate (ASR) increases by 16% on average across open and closed models in multi-turn settings. To close this gap, we propose ToolShield, a training-free, tool-agnostic, self-exploration defense: when encountering a new tool, the agent autonomously generates test cases, executes them to observe downstream effects, and distills safety experiences for deployment. Experiments show that ToolShield effectively reduces ASR by 30% on average in multi-turn interactions. Our code is available at https://github.com/CHATS-lab/ToolShield.", "authors": ["Xu Li", "Simon Yu", "Minzhou Pan", "Yiyou Sun", "Bo Li", "Dawn Song", "Xue Lin", "Weiyan Shi"], "categories": ["cs.CR", "cs.AI", "cs.CL", "cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2602.13379", "pdf_url": "https://arxiv.org/pdf/2602.13379v1", "arxiv_id": "2602.13379", "doi": "10.48550/arXiv.2602.13379", "citation_count": 8, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/CHATS-lab/ToolShield", "venue": "arXiv.org", "quality_score": 0.7384}
{"id": "8da3d81af6ab38508a4bcb52947ba3edde26b28634fe0def44420111af20660f", "sources": ["arxiv", "semantic_scholar"], "title": "Opinion dynamics and mutual influence with LLM agents through dialog simulation", "abstract": "A fundamental challenge in opinion dynamics research is the scarcity of real-world longitudinal opinion data, which complicates the validation of theoretical models. To address this, we propose a novel simulation framework using large language model (LLM) agents in structured multi-round dialogs. Each agent's dialog history is iteratively updated with its own previously stated opinions and those of others analogous to the classical DeGroot model. Furthermore, by retaining each agent's initial opinion throughout the dialog, we simulate anchoring effects consistent with the Friedkin-Johnsen model of opinion dynamics. Our framework thus bridges classical opinion dynamics models and modern multi-agent LLM systems, providing a scalable tool for simulating and analyzing opinion formation when real-world data is limited or inaccessible.", "authors": ["Yulong He", "Dutao Zhang", "Sergey Kovalchuk", "Pengyi Li", "Artem Sedakov"], "categories": ["cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2602.12583", "pdf_url": "https://arxiv.org/pdf/2602.12583v1", "arxiv_id": "2602.12583", "doi": "10.48550/arXiv.2602.12583", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4778}
{"id": "fb1a6e2a252a0da86b86cafdcedc732640cbdab6cc9545d587120b848f78cdeb", "sources": ["arxiv", "semantic_scholar"], "title": "SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents", "abstract": "Scientific reasoning inherently demands integrating sophisticated toolkits to navigate domain-specific knowledge. Yet, current benchmarks largely overlook agents' ability to orchestrate tools for such rigorous workflows. To bridge this gap, we introduce SciAgentGym, a scalable interactive environment featuring 1,780 domain-specific tools across four natural science disciplines, supported by a robust execution infrastructure. Complementing this, we present SciAgentBench, a tiered evaluation suite designed to stress-test agentic capabilities from elementary actions to long-horizon workflows. Our evaluation identifies a critical bottleneck: state-of-the-art models still struggle with complex scientific tool-use, and their performance degrades substantially as interaction horizons extend. To address this, we propose SciForge, a data synthesis method that models the tool action space as a dependency graph to generate logic-aware training trajectories. By fine-tuning on these trajectories, our SciAgent-8B outperforms the significantly larger Qwen3-VL-235B-Instruct while exhibiting positive cross-domain transfer of scientific tool-use capabilities. These results underscore the promising potential of next-generation autonomous scientific agents.", "authors": ["Yujiong Shen", "Yajie Yang", "Zhiheng Xi", "Binze Hu", "Huayu Sha", "Jiazheng Zhang", "Qiyuan Peng", "Junlin Shang", "Jixuan Huang", "Yutao Fan", "Jingqi Tong", "Shihan Dou", "Ming Zhang", "Lei Bai", "Zhenfei Yin", "Tao Gui", "Xingjun Ma", "Qi Zhang", "Xuanjing Huang", "Yu-Gang Jiang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2602.12984", "pdf_url": "https://arxiv.org/pdf/2602.12984v2", "arxiv_id": "2602.12984", "doi": "10.48550/arXiv.2602.12984", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4778}
{"id": "0b8403124572d2c37a6488158ee4e28323ac735a4a62eaf8565300b552e1a529", "sources": ["arxiv", "semantic_scholar"], "title": "Benchmarking LLM Tool-Use in the Wild", "abstract": "Fulfilling user needs through Large Language Model multi-turn, multi-step tool-use is rarely a straightforward process. Real user interactions are inherently wild, being intricate, messy, and flexible. We identify three key challenges from user behaviour: compositional tasks that demand efficient orchestration of tool-call topologies, implicit intent spread across dialogue turns that require contextual inference, and instruction transition, which mixes task queries, clarifications, and casual conversation, forcing LLMs to adjust their policies on the fly. Existing benchmarks overlook these behaviors, making the apparent progress of LLMs on tool-use spurious. To address this, we introduce WildToolBench, an LLM tool-use benchmark grounded in real-world user behavior patterns. Comprehensive evaluations of 57 LLMs reveal that no model achieves an accuracy of more than 15%, indicating a substantial gap in the robustness of LLMs' agentic ability. Controlled experiments and in-depth analyses further indicate that the real challenge for LLM tool-use lies not in artificially complex tasks, but in the wild nature of user behavior, emphasizing the need to reconsider the interactions among LLMs, users, and tools.", "authors": ["Peijie Yu", "Wei Liu", "Yifan Yang", "Jinjian Li", "Zelong Zhang", "Xiao Feng", "Feng Zhang"], "categories": ["cs.HC", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2604.06185", "pdf_url": "https://arxiv.org/pdf/2604.06185v1", "arxiv_id": "2604.06185", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3041}
{"id": "3aa9980c1d3a10e9580f7407d9c501655061107187f7b7c318cae10d5ca72b3b", "sources": ["arxiv", "semantic_scholar"], "title": "OMNI-LEAK: Orchestrator Multi-Agent Network Induced Data Leakage", "abstract": "As Large Language Model (LLM) agents become more capable, their coordinated use in the form of multi-agent systems is anticipated to emerge as a practical paradigm. Prior work has examined the safety and misuse risks associated with agents. However, much of this has focused on the single-agent case and/or setups missing basic engineering safeguards such as access control, revealing a scarcity of threat modeling in multi-agent systems. We investigate the security vulnerabilities of a popular multi-agent pattern known as the orchestrator setup, in which a central agent decomposes and delegates tasks to specialized agents. Through red-teaming a concrete setup representative of a likely future use case, we demonstrate a novel attack vector, OMNI-LEAK, that compromises several agents to leak sensitive data through a single indirect prompt injection, even in the presence of data access control. We report the susceptibility of frontier models to different categories of attacks, finding that both reasoning and non-reasoning models are vulnerable, even when the attacker lacks insider knowledge of the implementation details. Our work highlights the importance of safety research to generalize from single-agent to multi-agent settings, in order to reduce the serious risks of real-world privacy breaches and financial losses and overall public trust in AI agents.", "authors": ["Akshat Naik", "Jay Culligan", "Yarin Gal", "Philip Torr", "Rahaf Aljundi", "Alasdair Paren", "Adel Bibi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-13", "url": "https://arxiv.org/abs/2602.13477", "pdf_url": "https://arxiv.org/pdf/2602.13477v2", "arxiv_id": "2602.13477", "doi": "10.48550/arXiv.2602.13477", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4778}
{"id": "52397ca4003f5dbde8ab9256ab8777ccc434fefc6967911dc7622ccac40c7d23", "sources": ["arxiv", "semantic_scholar"], "title": "Perceptual Self-Reflection in Agentic Physics Simulation Code Generation", "abstract": "We present a multi-agent framework for generating physics simulation code from natural language descriptions, featuring a novel perceptual self-reflection mechanism for validation. The system employs four specialized agents: a natural language interpreter that converts user requests into physics-based descriptions; a technical requirements generator that produces scaled simulation parameters; a physics code generator with automated self-correction; and a physics validator that implements perceptual self-reflection. The key innovation is perceptual validation, which analyzes rendered animation frames using a vision-capable language model rather than inspecting code structure directly. This approach addresses the ``oracle gap'' where syntactically correct code produces physically incorrect behavior--a limitation that conventional testing cannot detect. We evaluate the system across seven domains including classical mechanics, fluid dynamics, thermodynamics, electromagnetics, wave physics, reaction-diffusion systems, and non-physics data visualization. The perceptual self-reflection architecture demonstrates substantial improvement over single-shot generation baselines, with the majority of tested scenarios achieving target physics accuracy thresholds. The system exhibits robust pipeline stability with consistent code self-correction capability, operating at approximately \\$0.20 per animation. These results validate our hypothesis that feeding visual simulation outputs back to a vision-language model for iterative refinement significantly outperforms single-shot code generation for physics simulation tasks and highlights the potential of agentic AI to support engineering workflows and physics data generation pipelines.", "authors": ["Prashant Shende", "Bradley Camburn"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.12311", "pdf_url": "https://arxiv.org/pdf/2602.12311v1", "arxiv_id": "2602.12311", "doi": "10.48550/arXiv.2602.12311", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4767}
{"id": "f56327bde14a99c7672e4ba4704eb40ef67ed8de1f77f51c8802020ca4b3a7f7", "sources": ["arxiv", "semantic_scholar"], "title": "CM2: Reinforcement Learning with Checklist Rewards for Multi-Turn and Multi-Step Agentic Tool Use", "abstract": "AI agents are increasingly used to solve real-world tasks by reasoning over multi-turn user interactions and invoking external tools. However, applying reinforcement learning to such settings remains difficult: realistic objectives often lack verifiable rewards and instead emphasize open-ended behaviors; moreover, RL for multi-turn, multi-step agentic tool use is still underexplored; and building and maintaining executable tool environments is costly, limiting scale and coverage. We propose CM2, an RL framework that replaces verifiable outcome rewards with checklist rewards. CM2 decomposes each turn's intended behavior into fine-grained binary criteria with explicit evidence grounding and structured metadata, turning open-ended judging into more stable classification-style decisions. To balance stability and informativeness, our method adopts a strategy of sparse reward assignment but dense evaluation criteria. Training is performed in a scalable LLM-simulated tool environment, avoiding heavy engineering for large tool sets. Experiments show that CM2 consistently improves over supervised fine-tuning. Starting from an 8B Base model and training on an 8k-example RL dataset, CM2 improves over the SFT counterpart by 8 points on tau^-Bench, by 10 points on BFCL-V4, and by 12 points on ToolSandbox. The results match or even outperform similarly sized open-source baselines, including the judging model. CM2 thus provides a scalable recipe for optimizing multi-turn, multi-step tool-using agents without relying on verifiable rewards. Code provided by the open-source community: https://github.com/namezhenzhang/CM2-RLCR-Tool-Agent.", "authors": ["Zhen Zhang", "Kaiqiang Song", "Xun Wang", "Yebowen Hu", "Weixiang Yan", "Chenyang Zhao", "Henry Peng Zou", "Haoyun Deng", "Sathish Reddy Indurthi", "Shujian Liu", "Simin Ma", "Xiaoyang Wang", "Xin Eric Wang", "Song Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.12268", "pdf_url": "https://arxiv.org/pdf/2602.12268v2", "arxiv_id": "2602.12268", "doi": "10.48550/arXiv.2602.12268", "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/namezhenzhang/CM2-RLCR-Tool-Agent", "venue": "arXiv.org", "quality_score": 0.7367}
{"id": "1de373ea85f8d152bd0ff8a8c4c6c39debb0a78e02241505cd46ff282a830c01", "sources": ["arxiv", "semantic_scholar"], "title": "AgentLeak: A Full-Stack Benchmark for Privacy Leakage in Multi-Agent LLM Systems", "abstract": "Multi-agent Large Language Model (LLM) systems create privacy risks that current benchmarks cannot measure. When agents coordinate on tasks, sensitive data passes through inter-agent messages, shared memory, and tool arguments, all pathways that output-only audits never inspect. We introduce AgentLeak, to the best of our knowledge the first full-stack benchmark for privacy leakage covering internal channels. It spans 1,000 scenarios across healthcare, finance, legal, and corporate domains, paired with a 32-class attack taxonomy and a three-tier detection pipeline. A factorial evaluation crossing five production LLMs (GPT-4o, GPT-4o-mini, Claude 3.5 Sonnet, Mistral Large, and Llama 3.3 70B) with all 1,000 scenarios, yielding 4,979 validated execution traces, reveals that multi-agent configurations reduce per-channel output leakage (C1: 27.2\\% vs 43.2\\% in single-agent) but introduce unmonitored internal channels that raise total system exposure to 68.9\\% (aggregated across C1, C2, C5). Internal channels account for most of this gap: inter-agent messages (C2) leak at 68.8\\%, compared to 27.2\\% on C1 (output channel). This means that output-only audits miss 41.7\\% of violations. Safety-aligned models achieve lower leakage on both external and internal channels, yet no model eliminates it. Across all five models and four domains, the pattern C2 $\\geq$ C1 holds consistently, confirming that inter-agent communication is the primary vulnerability. These results establish that output-only auditing is fundamentally insufficient for multi-agent systems and that privacy controls must be extended to inter-agent communication channels.", "authors": ["Faouzi El Yagoubi", "Godwin Badu-Marfo", "Ranwa Al Mallah"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.11510", "pdf_url": "https://arxiv.org/pdf/2602.11510v2", "arxiv_id": "2602.11510", "doi": "10.48550/arXiv.2602.11510", "citation_count": 30, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/Privatris/AgentLeak", "venue": "arXiv.org", "quality_score": 0.7367}
{"id": "810955f31198b195a4b2baedadbf62e33d9a11877fe4d6aa4e9f9394f90ccdbe", "sources": ["arxiv", "semantic_scholar"], "title": "MalTool: Malicious Tool Attacks on LLM Agents", "abstract": "In a malicious tool attack, an attacker uploads a malicious tool to a distribution platform; once a user inadvertently installs the tool and the LLM agent selects it during task execution, the tool can compromise the user's security and privacy. Prior work focuses on manipulating tool names and descriptions to increase the likelihood of installation by users and selection by LLM agents. However, a successful attack also requires embedding malicious behaviors in the tool's code implementation, which remains largely unexplored. In this work, we bridge this gap by presenting the first systematic study of malicious tool code implementations. We first propose a taxonomy of malicious tool behaviors based on the confidentiality-integrity-availability triad, tailored to LLM-agent settings. To investigate the severity of the risks posed by attackers exploiting coding LLMs to automatically generate malicious tools, we develop MalTool, a coding-LLM-based framework that synthesizes tools exhibiting specified malicious behaviors, either as standalone tools or embedded within otherwise benign implementations. To ensure functional correctness and structural diversity, MalTool leverages an automated verifier that validates whether generated tools exhibit the intended malicious behaviors and differ sufficiently from previously generated instances, iteratively refining generations until success. Our evaluation demonstrates that MalTool is highly effective even when coding LLMs are safety-aligned. Using MalTool, we construct two datasets of malicious tools: 1,300 standalone malicious tools and 5,727 real-world tools with embedded malicious behaviors. We further show that existing detection methods, including conventional malware detection approaches and methods tailored to the LLM-agent setting, exhibit limited effectiveness at detecting the malicious tools, highlighting an urgent need for new defenses.", "authors": ["Yuepeng Hu", "Yuqi Jia", "Mengyuan Li", "Dawn Song", "Neil Gong"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.12194", "pdf_url": "https://arxiv.org/pdf/2602.12194v3", "arxiv_id": "2602.12194", "doi": "10.48550/arXiv.2602.12194", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4767}
{"id": "9e2d8f1fb1e2df16a221fcb383cfc5ff2963b395d7256dcfc711f1f2954dde27", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond End-to-End Video Models: An LLM-Based Multi-Agent System for Educational Video Generation", "abstract": "Although recent end-to-end video generation models demonstrate impressive performance in visually oriented content creation, they remain limited in scenarios that require strict logical rigor and precise knowledge representation, such as instructional and educational media. To address this problem, we propose LASEV, a hierarchical LLM-based multi-agent system for generating high-quality instructional videos from educational problems. LASEV formulates educational video generation as a multi-objective task that simultaneously demands correct step-by-step reasoning, pedagogically coherent narration, semantically faithful visual demonstrations, and precise audio--visual alignment. To address the limitations of prior approaches--including low procedural fidelity, high production cost, and limited controllability--LASEV decomposes the generation workflow into specialized agents that collaborate through a central Orchestrating Agent, shared production state, explicit quality gates, and iterative critique mechanisms. Specifically, the Orchestrating Agent supervises a Solution Agent for rigorous problem solving, an Illustration Agent that produces executable visualization code, and a Narration Agent for learner-oriented instructional scripts. In addition, all outputs from the working agents are subject to semantic critique, rule-based constraints, and tool-based compilation checks. Rather than directly synthesizing pixels, the system constructs a structured executable video script that is deterministically compiled into synchronized visuals and narration using template-driven assembly rules, enabling fully automated production without manual editing. In large-scale deployments, LASEV achieves a throughput exceeding one million videos per day, delivering over a 95% reduction in cost compared to current industry-standard approaches while maintaining a high acceptance rate.", "authors": ["Lingyong Yan", "Jiulong Wu", "Dong Xie", "Weixian Shi", "Deguo Xia", "Jizhou Huang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.11790", "pdf_url": "https://arxiv.org/pdf/2602.11790v2", "arxiv_id": "2602.11790", "doi": "10.1145/3770855.3818323", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4767}
{"id": "b436e5bb6ce2e39a1ac4d0dcf429ca920b8082a111d4144c42fd4677021b1c2e", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperation Breakdown in LLM Agents Under Communication Delays", "abstract": "LLM-based multi-agent systems (LLM-MAS), in which autonomous AI agents cooperate to solve tasks, are gaining increasing attention. For such systems to be deployed in society, agents must be able to establish cooperation and coordination under real-world computational and communication constraints. We propose the FLCOA framework (Five Layers for Cooperation/Coordination among Autonomous Agents) to conceptualize how cooperation and coordination emerge in groups of autonomous agents, and highlight that the influence of lower-layer factors - especially computational and communication resources - has been largely overlooked. To examine the effect of communication delay, we introduce a Continuous Prisoner's Dilemma with Communication Delay and conduct simulations with LLM-based agents. As delay increases, agents begin to exploit slower responses even without explicit instructions. Interestingly, excessive delay reduces cycles of exploitation, yielding a U-shaped relationship between delay magnitude and mutual cooperation. These results suggest that fostering cooperation requires attention not only to high-level institutional design but also to lower-layer factors such as communication delay and resource allocation, pointing to new directions for MAS research.", "authors": ["Keita Nishimoto", "Kimitaka Asatani", "Ichiro Sakata"], "categories": ["cs.MA", "cs.AI", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-12", "url": "https://arxiv.org/abs/2602.11754", "pdf_url": "https://arxiv.org/pdf/2602.11754v1", "arxiv_id": "2602.11754", "doi": "10.48550/arXiv.2602.11754", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4767}
{"id": "0fb6b1773ca9307c3699ac3529582647ccda15dcd88c629991aef2139a1b1507", "sources": ["arxiv", "semantic_scholar"], "title": "TVCACHE: A Stateful Tool-Value Cache for Post-Training LLM Agents", "abstract": "In RL post-training of LLM agents, calls to external tools take several seconds or even minutes, leaving allocated GPUs idle and inflating post-training time and cost. While many tool invocations repeat across parallel rollouts and could in principle be cached, naively caching their outputs for reuse is incorrect since tool outputs depend on the environment state induced by prior agent interactions. We present TVCACHE, a stateful tool-value cache for LLM agent post-training. TVCACHE maintains a tree of observed tool-call sequences and performs longest-prefix matching for cache lookups: a hit occurs only when the agent's full tool history matches a previously executed sequence, guaranteeing identical environment state. On three diverse workloads-terminal-based tasks, SQL generation, and video understanding. TVCACHE achieves cache hit rates of up to 70% and reduces median tool call execution time by up to 6.9X, with no degradation in post-training reward accumulation.", "authors": ["Abhishek Vijaya Kumar", "Bhaskar Kataria", "Byungsoo Oh", "Emaad Manzoor", "Rachee Singh"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-11", "url": "https://arxiv.org/abs/2602.10986", "pdf_url": "https://arxiv.org/pdf/2602.10986v1", "arxiv_id": "2602.10986", "doi": "10.48550/arXiv.2602.10986", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4755}
{"id": "23988b6ebd6385802d28b4ff9a82cdbf5498ed6451df020a34a01d6fc67236ac", "sources": ["arxiv", "semantic_scholar"], "title": "CryptoAnalystBench: Failures in Multi-Tool Long-Form LLM Analysis", "abstract": "Modern analyst agents must reason over complex, high token inputs, including dozens of retrieved documents, tool outputs, and time sensitive data. While prior work has produced tool calling benchmarks and examined factuality in knowledge augmented systems, relatively little work studies their intersection: settings where LLMs must integrate large volumes of dynamic, structured and unstructured multi tool outputs. We investigate LLM failure modes in this regime using crypto as a representative high data density domain. We introduce (1) CryptoAnalystBench, an analyst aligned benchmark of 198 production crypto and DeFi queries spanning 11 categories; (2) an agentic harness equipped with relevant crypto and DeFi tools to generate responses across multiple frontier LLMs; and (3) an evaluation pipeline with citation verification and an LLM as a judge rubric spanning four user defined success dimensions: relevance, temporal relevance, depth, and data consistency. Using human annotation, we develop a taxonomy of seven higher order error types that are not reliably captured by factuality checks or LLM based quality scoring. We find that these failures persist even in state of the art systems and can compromise high stakes decisions. Based on this taxonomy, we refine the judge rubric to better capture these errors. While the judge does not align with human annotators on precise scoring across rubric iterations, it reliably identifies critical failure modes, enabling scalable feedback for developers and researchers studying analyst style agents. We release CryptoAnalystBench with annotated queries, the evaluation pipeline, judge rubrics, and the error taxonomy, and outline mitigation strategies and open challenges in evaluating long form, multi tool augmented systems.", "authors": ["Anushri Eswaran", "Oleg Golev", "Darshan Tank", "Sidhant Rahi", "Himanshu Tyagi"], "categories": ["cs.IR", "cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-11", "url": "https://arxiv.org/abs/2602.11304", "pdf_url": "https://arxiv.org/pdf/2602.11304v1", "arxiv_id": "2602.11304", "doi": "10.48550/arXiv.2602.11304", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4755}
{"id": "7fe2268b9b9bfb2257335f179f35bce1228e6c107e49295961f9bc0f5c65c7f3", "sources": ["arxiv", "semantic_scholar"], "title": "MATRAG: Multi-Agent Transparent Retrieval-Augmented Generation for Explainable Recommendations", "abstract": "Large Language Model (LLM)-based recommendation systems have demonstrated remarkable capabilities in understanding user preferences and generating personalized suggestions. However, existing approaches face critical challenges in transparency, knowledge grounding, and the ability to provide coherent explanations that foster user trust. We introduce MATRAG (Multi-Agent Transparent Retrieval-Augmented Generation), a novel framework that combined multi-agent collaboration with knowledge graph-augmented retrieval to deliver explainable recommendations. MATRAG employs four specialized agents: a User Modeling Agent that constructs dynamic preference profiles, an Item Analysis Agent that extracts semantic features from knowledge graphs, a Reasoning Agent that synthesizes collaborative and content-based signals, and an Explanation Agent that generates natural language justifications grounded in retrieved knowledge. Our framework incorporates a transparency scoring mechanism that quantifies explanation faithfulness and relevance. Extensive experiments on three benchmark datasets (Amazon Reviews, MovieLens-1M, and Yelp) demonstrate that MATRAG achieves state-of-the-art performance, improving recommendation accuracy by 12.7\\% (Hit Rate) and 15.3\\% (NDCG) over leading baselines, while human evaluation confirms that 87.4\\% of generated explanations are rated as helpful and trustworthy by domain experts. Our work establishes new benchmarks for transparent, agentic recommendation systems and provides actionable insights for deploying LLM-based recommenders in production environments.", "authors": ["Sushant Mehta"], "categories": ["cs.IR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-11", "url": "https://arxiv.org/abs/2604.20848", "pdf_url": "https://arxiv.org/pdf/2604.20848v1", "arxiv_id": "2604.20848", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3026}
{"id": "b4e8f469426a644dcfd066c1edbc2aebf73e1ed911d6a128f91f129ec406a2cd", "sources": ["arxiv", "semantic_scholar"], "title": "Auditing Multi-Agent LLM Reasoning Trees Outperforms Majority Vote and LLM-as-Judge", "abstract": "Multi-agent systems (MAS) can substantially extend the reasoning capacity of large language models (LLMs), yet most frameworks still aggregate agent outputs with majority voting. This heuristic discards the evidential structure of reasoning traces and is brittle under the confabulation consensus, where agents share correlated biases and converge on the same incorrect rationale. We introduce AgentAuditor, which replaces voting with a path search over a Reasoning Tree that explicitly represents agreements and divergences among agent traces. AgentAuditor resolves conflicts by comparing reasoning branches at critical divergence points, turning global adjudication into efficient, localized verification. We further propose Anti-Consensus Preference Optimization (ACPO), which trains the adjudicator on majority-failure cases and rewards evidence-based minority selections over popular errors. AgentAuditor is agnostic to MAS setting, and we find across 5 popular settings that it yields up to 5% absolute accuracy improvement over a majority vote, and up to 3% over using LLM-as-Judge.", "authors": ["Wei Yang", "Shixuan Li", "Heng Ping", "Peiyu Zhang", "Paul Bogdan", "Jesse Thomason"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-10", "url": "https://arxiv.org/abs/2602.09341", "pdf_url": "https://arxiv.org/pdf/2602.09341v1", "arxiv_id": "2602.09341", "doi": "10.48550/arXiv.2602.09341", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4744}
{"id": "17bd1eefc5b3340987a52e10be7d7bcc66d0f84b1261b5208469837207bc48b8", "sources": ["arxiv", "semantic_scholar"], "title": "Internalizing Multi-Agent Reasoning for Accurate and Efficient LLM-based Recommendation", "abstract": "Large Language Models (LLMs) are reshaping recommender systems by leveraging extensive world knowledge and semantic reasoning to interpret user intent. However, effectively integrating these capabilities with collaborative signals while avoiding prohibitive inference latency remains a critical bottleneck. To address this, we propose a trajectory-driven internalization framework to develop a Single-agent Trajectory-Aligned Recommender (STAR). Specifically, to internalize complex reasoning capabilities into a single efficient model, we first design a multi-agent teacher system capable of multi-turn tool usage and reflection. This teacher utilizes a Collaborative Signal Translation mechanism to explicitly convert latent behavioral patterns into descriptive natural language evidence to enhance reasoning accuracy. Subsequently, a trajectory-driven distillation pipeline transfers this agentic logic, including planning, tool usage, and self-reflection, into the compact STAR model. Extensive experiments demonstrate that STAR surpasses its teacher by 8.7% to 39.5% while eliminating iterative latency, paving the way for real-time, reasoning-enhanced recommendation.", "authors": ["Yang Wu", "Haoze Wang", "Qian Li", "Jun Zhang", "Huan Yu", "Jie Jiang"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-10", "url": "https://arxiv.org/abs/2602.09829", "pdf_url": "https://arxiv.org/pdf/2602.09829v2", "arxiv_id": "2602.09829", "doi": "10.48550/arXiv.2602.09829", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4744}
{"id": "73c4252c00437882cd12ef02116d78e6b68a97bb77c0931981ea374caa2dfb75", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Banana: High-Fidelity Image Editing with Agentic Thinking and Tooling", "abstract": "We study instruction-based image editing under professional workflows and identify three persistent challenges: (i) editors often over-edit, modifying content beyond the user's intent; (ii) existing models are largely single-turn, while multi-turn edits can alter object faithfulness; and (iii) evaluation at around 1K resolution is misaligned with real workflows that often operate on ultra high-definition images (e.g., 4K). We propose Agent Banana, a hierarchical agentic planner-executor framework for high-fidelity, object-aware, deliberative editing. Agent Banana introduces two key mechanisms: (1) Context Folding, which compresses long interaction histories into structured memory for stable long-horizon control; and (2) Image Layer Decomposition, which performs localized layer-based edits to preserve non-target regions while enabling native-resolution outputs. To support rigorous evaluation, we build HDD-Bench, a high-definition, dialogue-based benchmark featuring verifiable stepwise targets and native 4K images (11.8M pixels) for diagnosing long-horizon failures. On HDD-Bench, Agent Banana achieves the best multi-turn consistency and background fidelity (e.g., IC 0.871, SSIM-OM 0.84, LPIPS-OM 0.12) while remaining competitive on instruction following, and also attains strong performance on standard single-turn editing benchmarks. We hope this work advances reliable, professional-grade agentic image editing and its integration into real workflows.", "authors": ["Ruijie Ye", "Jiayi Zhang", "Zhuoxin Liu", "Zihao Zhu", "Siyuan Yang", "Li Li", "Tianfu Fu", "Franck Dernoncourt", "Yue Zhao", "Jiacheng Zhu", "Ryan Rossi", "Wenhao Chai", "Zhengzhong Tu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.09084", "pdf_url": "https://arxiv.org/pdf/2602.09084v2", "arxiv_id": "2602.09084", "doi": "10.48550/arXiv.2602.09084", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "1068b6f17df544a2a200434c714a8037fef3cc17e2b4b53c77708b4d4473f883", "sources": ["arxiv", "semantic_scholar"], "title": "Dr. MAS: Stable Reinforcement Learning for Multi-Agent LLM Systems", "abstract": "Multi-agent LLM systems enable advanced reasoning and tool use via role specialization, yet reliable reinforcement learning (RL) post-training for such systems remains difficult. In this work, we theoretically pinpoint a key reason for training instability when extending group-based RL to multi-agent LLM systems. We show that under GRPO-style optimization, a global normalization baseline may deviate from diverse agents' reward distributions, which ultimately leads to gradient-norm instability. Based on this finding, we propose Dr. MAS, a simple and stable RL training recipe for multi-agent LLM systems. Dr. MAS uses an agent-wise remedy: normalizing advantages per agent using each agent's own reward statistics, which calibrates gradient scales and dramatically stabilizes training, both theoretically and empirically. Beyond the algorithm, Dr. MAS provides an end-to-end RL training framework for multi-agent LLM systems, supporting scalable orchestration, flexible per-agent LLM serving and optimization configs, and shared resource scheduling of LLM actor backends. We evaluate Dr. MAS on multi-agent math reasoning and multi-turn search benchmarks using Qwen2.5 and Qwen3 series models. Dr. MAS achieves clear gains over vanilla GRPO (e.g., +5.6\\% avg@16 and +4.6\\% pass@16 on math, and +15.2\\% avg@16 and +13.1\\% pass@16 on search) while largely eliminating gradient spikes. Moreover, it remains highly effective under heterogeneous agent-model assignments while improving efficiency.", "authors": ["Lang Feng", "Longtao Zheng", "Shuo He", "Fuxiang Zhang", "Bo An"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.08847", "pdf_url": "https://arxiv.org/pdf/2602.08847v1", "arxiv_id": "2602.08847", "doi": "10.48550/arXiv.2602.08847", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "b0f993857cd24fa687188adbd1c1437f61ed39bd0f18c1d4595919cb396d7992", "sources": ["arxiv", "semantic_scholar"], "title": "ValueFlow: Measuring the Propagation of Value Perturbations in Multi-Agent LLM Systems", "abstract": "Multi-agent large language model (LLM) systems increasingly consist of agents that observe and respond to one another's outputs. While value alignment is typically evaluated for isolated models, how value perturbations propagate through agent interactions remains poorly understood. We present ValueFlow, a perturbation-based framework that measures value drift in multi-agent systems via a 56-value valuation dataset derived from the Schwartz Value Survey, with agent value orientations scored using an LLM-as-a-judge protocol. ValueFlow decomposes value drift into agent-level response behavior and system-level structural effects, captured by two metrics: \\b{eta}-susceptibility, an agent's sensitivity to perturbed peer value signals, and system susceptibility (SS), the effect of node-level perturbations on final system outputs.Experiments span across value dimensions, backbones, personas, and topologies, showing that susceptibility varies sharply across values and is strongly shaped by interaction structure, indicating that value alignment in multi-agent systems is a system-level property, not just an agent-level one. ValueFlow thus provides a principled basis for auditing and mitigating value propagation in deployed multi-agent systems.", "authors": ["Jinnuo Liu", "Chuke Liu", "Hua Shen"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.08567", "pdf_url": "https://arxiv.org/pdf/2602.08567v2", "arxiv_id": "2602.08567", "doi": "10.48550/arXiv.2602.08567", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "0f759e51f91067a122787b73952234e807033f56913e1e1ff7d4a611a19e37b7", "sources": ["arxiv", "semantic_scholar"], "title": "Who Deserves the Reward? SHARP: Shapley Credit-based Optimization for Multi-Agent System", "abstract": "Integrating Large Language Models (LLMs) with external tools via multi-agent systems offers a promising new paradigm for decomposing and solving complex problems. However, training these systems remains notoriously difficult due to the credit assignment challenge, as it is often unclear which specific functional agent is responsible for the success or failure of decision trajectories. Existing methods typically rely on sparse or globally broadcast rewards, failing to capture individual contributions and leading to inefficient reinforcement learning. To address these limitations, we introduce the Shapley-based Hierarchical Attribution for Reinforcement Policy (SHARP), a novel framework for optimizing multi-agent reinforcement learning via precise credit attribution. SHARP effectively stabilizes training by normalizing agent-specific advantages across trajectory groups, primarily through a decomposed reward mechanism comprising a global broadcast-accuracy reward, a Shapley-based marginal-credit reward for each agent, and a tool-process reward to improve execution efficiency. Extensive experiments across various real-world benchmarks demonstrate that SHARP significantly outperforms recent state-of-the-art baselines, achieving average match improvements of 23.66% and 14.05% over single-agent and multi-agent approaches, respectively.", "authors": ["Yanming Li", "Xuelin Zhang", "WenJie Lu", "Ziye Tang", "Maodong Wu", "Haotian Luo", "Tongtong Wu", "Zijie Peng", "Hongze Mi", "Yibo Feng", "Naiqiang Tan", "Chao Huang", "Lian Peng", "Li Shen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.08335", "pdf_url": "https://arxiv.org/pdf/2602.08335v2", "arxiv_id": "2602.08335", "doi": "10.48550/arXiv.2602.08335", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "a97f9940fc216d23a077200a59a0014a14a4295730901658ddec374568862c2d", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Mars: Multi-Agent Simulation for Multi-Planetary Life Exploration and Settlement", "abstract": "Artificial Intelligence (AI) has transformed robotics, healthcare, industry, and scientific discovery, yet a major frontier may lie beyond Earth. Space exploration and settlement offer vast environments and resources, but impose constraints unmatched on Earth: delayed/intermittent communications, extreme resource scarcity, heterogeneous expertise, and strict safety, accountability, and command authority. The key challenge is auditable coordination among specialised humans, robots, and digital services in a safety-critical system-of-systems. We introduce Agent Mars, an open, end-to-end multi-agent simulation framework for Mars base operations. Agent Mars formalises a realistic organisation with a 93-agent roster across seven layers of command and execution (human roles and physical assets), enabling base-scale studies beyond toy settings. It implements hierarchical and cross-layer coordination that preserves chain-of-command while allowing vetted cross-layer exchanges with audit trails; supports dynamic role handover with automatic failover under outages; and enables phase-dependent leadership for routine operations, emergencies, and science campaigns. Agent Mars further models mission-critical mechanisms-scenario-aware short/long-horizon memory, configurable propose-vote consensus, and translator-mediated heterogeneous protocols-to capture how teams align under stress. To quantify behaviour, we propose the Agent Mars Performance Index (AMPI), an interpretable composite score with diagnostic sub-metrics. Across 13 reproducible Mars-relevant operational scripts, Agent Mars reveals coordination trade-offs and identifies regimes where curated cross-layer collaboration and functional leadership reduce overhead without sacrificing reliability. Agent Mars provides a benchmarkable, auditable foundation for Space AI.", "authors": ["Ziyang Wang"], "categories": ["cs.MA", "astro-ph.IM", "cs.AI"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.13291", "pdf_url": "https://arxiv.org/pdf/2602.13291v1", "arxiv_id": "2602.13291", "doi": "10.48550/arXiv.2602.13291", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "3ae35ced6e9a91c90a4d10c8ccdce730204aa908fe403232bbea267c4200fc6d", "sources": ["arxiv", "semantic_scholar"], "title": "Taming Scylla: Understanding the multi-headed agentic daemon of the coding seas", "abstract": "LLM-based tools are automating more software development tasks at a rapid pace, but there is no rigorous way to evaluate how different architectural choices -- prompts, skills, tools, multi-agent setups -- materially affect both capability and cost. This paper introduces Scylla, an evaluation framework for benchmarking agentic coding tools through structured ablation studies that uses seven testing tiers (T0-T6) progressively adding complexity to isolate what directly influences results and how. The key metric is Cost-of-Pass (CoP): the expected dollar cost to get one correct solution, which directly quantifies the trade-off between complexity and efficiency. The framework is model-agnostic, designed to work with any CLI tool; this paper demonstrates it with Claude Sonnet 4.5, using multiple LLM judges (Opus 4.5, Sonnet 4.5, Haiku 4.5) from the same vendor for evaluation consensus, where judges score results using direct tests, human-designed LLM-evaluated rubrics, and qualitative assessment. The result is a reproducible framework that quantifies trade-offs between agent complexity and actual outcomes, suggesting that architectural complexity does not always improve quality.", "authors": ["Micah Villmow"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-09", "url": "https://arxiv.org/abs/2602.08765", "pdf_url": "https://arxiv.org/pdf/2602.08765v1", "arxiv_id": "2602.08765", "doi": "10.48550/arXiv.2602.08765", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4732}
{"id": "0593670e2f7febd3c065b152e1dca637a1962e274df939076c6d37ab8db2a141", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Adaptive, Scalable, and Robust Coordination of LLM Agents: A Dynamic Ad-Hoc Networking Perspective", "abstract": "Multi-agent architectures built on large language models (LLMs) have demonstrated the potential to realize swarm intelligence through well-crafted collaboration. However, the substantial burden of manual orchestration inherently raises an imperative to automate the design of agentic workflows. We frame such an agent coordination challenge as a classic problem in dynamic ad-hoc networking: How to establish adaptive and reliable communication among a scalable number of agentic hosts? In response to this unresolved dilemma, we introduce RAPS, a reputation-aware publish-subscribe paradigm for adaptive, scalable, and robust coordination of LLM agents. RAPS is grounded in the Distributed Publish-Subscribe Protocol, allowing LLM agents to exchange messages based on their declared intents rather than predefined topologies. Beyond this substrate, RAPS further incorporates two coherent overlays: (i) Reactive Subscription, enabling agents to dynamically refine their intents; and (ii) Bayesian Reputation, empowering each agent with a local watchdog to detect and isolate malicious peers. Extensive experiments over five benchmarks showcase that our design effectively reconciles adaptivity, scalability, and robustness in a unified multi-agent coordination framework.", "authors": ["Rui Li", "Zeyu Zhang", "Xiaohe Bo", "Quanyu Dai", "Chaozhuo Li", "Feng Wen", "Xu Chen"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-08", "url": "https://arxiv.org/abs/2602.08009", "pdf_url": "https://arxiv.org/pdf/2602.08009v1", "arxiv_id": "2602.08009", "doi": "10.48550/arXiv.2602.08009", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4721}
{"id": "4c9226bc79e8d41551d755dcb36175a2c79d6d8af9cb1bd4959db944d6adda45", "sources": ["arxiv", "semantic_scholar"], "title": "CTFExplorer: Evaluating LLM Offensive Agents Through Multi-Target Web CTF Benchmarking", "abstract": "Existing benchmarks for LLM-based offensive security agents use isolated, single-target setups with a known vulnerable service and fixed objective. They measure exploitation effectively, but miss how real Capture-the-Flag (CTF) participants triage unknown surfaces, prioritize targets, and allocate effort under uncertainty. Current evaluations therefore fail to assess strategic reasoning beyond exploitation alone. To address this, we introduce \\textit{CTFExplorer}, a benchmark suite that shifts offensive security evaluation toward a multi-target setting, which tests how agents explore, prioritize, and chain attacks. CTFExplorer deploys 40 web-based vulnerable services within a single environment, where agents must autonomously discover, distinguish, and exploit targets without predefined guidance. We also present a reactive multi-agent setup as a reference agent framework and develop an agent-agnostic evaluation framework that records structured reasoning traces for fine-grained assessment. This enables behavioral evaluation beyond binary flag capture, such as how agents manage target selection, handle failed hypotheses, coordinate across multiple stages, and extract security intelligence.", "authors": ["Nanda Rani", "Kimberly Milner", "Minghao Shao", "Meet Udeshi", "Haoran Xi", "Venkata Sai Charan Putrevu", "Saksham Aggarwal", "Sandeep K. Shukla", "Prashanth Krishnamurthy", "Farshad Khorrami", "Muhammad Shafique", "Ramesh Karri"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-08", "url": "https://arxiv.org/abs/2602.08023", "pdf_url": "https://arxiv.org/pdf/2602.08023v3", "arxiv_id": "2602.08023", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3004}
{"id": "54f62d13cbf91af10370453bd2505ee5dec0ea3e1518912059fa395c27c4a875", "sources": ["arxiv", "semantic_scholar"], "title": "W&D:Scaling Parallel Tool Calling for Efficient Deep Research Agents", "abstract": "Deep research agents have emerged as powerful tools for automating complex intellectual tasks through multi-step reasoning and web-based information seeking. While recent efforts have successfully enhanced these agents by scaling depth through increasing the number of sequential thinking and tool calls, the potential of scaling width via parallel tool calling remains largely unexplored. In this work, we propose the Wide and Deep research agent, a framework designed to investigate the behavior and performance of agents when scaling not only depth but also width via parallel tool calling. Unlike existing approaches that rely on complex multi-agent orchestration to parallelize workloads, our method leverages intrinsic parallel tool calling to facilitate effective coordination within a single reasoning step. We demonstrate that scaling width significantly improves performance on deep research benchmarks while reducing the number of turns required to obtain correct answers. Furthermore, we analyze the factors driving these improvements through case studies and explore various tool call schedulers to optimize parallel tool calling strategy. Our findings suggest that optimizing the trade-off between width and depth is a critical pathway toward high-efficiency deep research agents. Notably, without context management or other tricks, we obtain 62.2% accuracy with GPT-5-Medium on BrowseComp, surpassing the original 54.9% reported by GPT-5-High.", "authors": ["Xiaoqiang Lin", "Jun Hao Liew", "Silvio Savarese", "Junnan Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-07", "url": "https://arxiv.org/abs/2602.07359", "pdf_url": "https://arxiv.org/pdf/2602.07359v1", "arxiv_id": "2602.07359", "doi": "10.48550/arXiv.2602.07359", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4709}
{"id": "2394d414e98d95f32e014b9381ded477ed2dabbcc5839418e14f847e19141db2", "sources": ["arxiv", "semantic_scholar"], "title": "DiLLS: Interactive Diagnosis of LLM-based Multi-agent Systems via Layered Summary of Agent Behaviors", "abstract": "Large language model (LLM)-based multi-agent systems have demonstrated impressive capabilities in handling complex tasks. However, the complexity of agentic behaviors makes these systems difficult to understand. When failures occur, developers often struggle to identify root causes and to determine actionable paths for improvement. Traditional methods that rely on inspecting raw log records are inefficient, given both the large volume and complexity of data. To address this challenge, we propose a framework and an interactive system, DiLLS, designed to reveal and structure the behaviors of multi-agent systems. The key idea is to organize information across three levels of query completion: activities, actions, and operations. By probing the multi-agent system through natural language, DiLLS derives and organizes information about planning and execution into a structured, multi-layered summary. Through a user study, we show that DiLLS significantly improves developers' effectiveness and efficiency in identifying, diagnosing, and understanding failures in LLM-based multi-agent systems.", "authors": ["Rui Sheng", "Yukun Yang", "Chuhan Shi", "Yanna Lin", "Zixin Chen", "Huamin Qu", "Furui Cheng"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-05", "url": "https://arxiv.org/abs/2602.05446", "pdf_url": "https://arxiv.org/pdf/2602.05446v1", "arxiv_id": "2602.05446", "doi": "10.1145/3772318.3790815", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2982}
{"id": "72262403a21ee097a49d953d6f0bbbca5edde62e668195c72ceb041d30816a04", "sources": ["arxiv", "semantic_scholar"], "title": "RuleSmith: Multi-Agent LLMs for Automated Game Balancing", "abstract": "Game balancing is a longstanding challenge requiring repeated playtesting, expert intuition, and extensive manual tuning. We introduce RuleSmith, the first framework that achieves automated game balancing by leveraging the reasoning capabilities of multi-agent LLMs. It couples a game engine, multi-agent LLMs self-play, and Bayesian optimization operating over a multi-dimensional rule space. As a proof of concept, we instantiate RuleSmith on CivMini, a simplified civilization-style game containing heterogeneous factions, economy systems, production rules, and combat mechanics, all governed by tunable parameters. LLM agents interpret textual rulebooks and game states to generate actions, to conduct fast evaluation of balance metrics such as win-rate disparities. To search the parameter landscape efficiently, we integrate Bayesian optimization with acquisition-based adaptive sampling and discrete projection: promising candidates receive more evaluation games for accurate assessment, while exploratory candidates receive fewer games for efficient exploration. Experiments show that RuleSmith converges to highly balanced configurations and provides interpretable rule adjustments that can be directly applied to downstream game systems. Our results illustrate that LLM simulation can serve as a powerful surrogate for automating design and balancing in complex multi-agent environments.", "authors": ["Ziyao Zeng", "Chen Liu", "Tianyu Liu", "Hao Wang", "Xiatao Sun", "Fengyu Yang", "Xiaofeng Liu", "Zhiwen Fan"], "categories": ["cs.LG", "cs.AI", "cs.GT", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-05", "url": "https://arxiv.org/abs/2602.06232", "pdf_url": "https://arxiv.org/pdf/2602.06232v1", "arxiv_id": "2602.06232", "doi": "10.48550/arXiv.2602.06232", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4686}
{"id": "dd6a62e824630694110eb3ebac03caf741d855bedbbe207c8ab70349f01e2f04", "sources": ["arxiv", "semantic_scholar"], "title": "Data-Centric Interpretability for LLM-based Multi-Agent Reinforcement Learning", "abstract": "Large language models (LLMs) are increasingly trained in complex Reinforcement Learning, multi-agent environments, making it difficult to understand how behavior changes over training. Sparse Autoencoders (SAEs) have recently shown to be useful for data-centric interpretability. In this work, we analyze large-scale reinforcement learning training runs from the sophisticated environment of Full-Press Diplomacy by applying pretrained SAEs, alongside LLM-summarizer methods. We introduce Meta-Autointerp, a method for grouping SAE features into interpretable hypotheses about training dynamics. We discover fine-grained behaviors including role-playing patterns, degenerate outputs, language switching, alongside high-level strategic behaviors and environment-specific bugs. Through automated evaluation, we validate that 90% of discovered SAE Meta-Features are significant, and find a surprising reward hacking behavior. However, through two user studies, we find that even subjectively interesting and seemingly helpful SAE features may be worse than useless to humans, along with most LLM generated hypotheses. However, a subset of SAE-derived hypotheses are predictively useful for downstream tasks. We further provide validation by augmenting an untrained agent's system prompt, improving the score by +14.2%. Overall, we show that SAEs and LLM-summarizer provide complementary views into agent behavior, and together our framework forms a practical starting point for future data-centric interpretability work on ensuring trustworthy LLM behavior throughout training.", "authors": ["John Yan", "Michael Yu", "Yuqi Sun", "Alexander Duffy", "Tyler Marques", "Matthew Lyle Olson"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-05", "url": "https://arxiv.org/abs/2602.05183", "pdf_url": "https://arxiv.org/pdf/2602.05183v2", "arxiv_id": "2602.05183", "doi": "10.48550/arXiv.2602.05183", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4686}
{"id": "80fb2739f22107a835d5d33a43814564731cfcdda8920134f96ae35c75edbe6c", "sources": ["arxiv", "semantic_scholar"], "title": "AgenticPay: A Multi-Agent LLM Negotiation System for Buyer-Seller Transactions", "abstract": "Large language model (LLM)-based agents are increasingly expected to negotiate, coordinate, and transact autonomously, yet existing benchmarks lack principled settings for evaluating language-mediated economic interaction among multiple agents. We introduce AgenticPay, a benchmark and simulation framework for multi-agent buyer-seller negotiation driven by natural language. AgenticPay models markets in which buyers and sellers possess private constraints and product-dependent valuations, and must reach agreements through multi-round linguistic negotiation rather than numeric bidding alone. The framework supports a diverse suite of over 110 tasks ranging from bilateral bargaining to many-to-many markets, with structured action extraction and metrics for feasibility, efficiency, and welfare. Benchmarking state-of-the-art proprietary and open-weight LLMs reveals substantial gaps in negotiation performance and highlights challenges in long-horizon strategic reasoning, establishing AgenticPay as a foundation for studying agentic commerce and language-based market interaction. Code and dataset are available at the link: https://github.com/SafeRL-Lab/AgenticPay.", "authors": ["Xianyang Liu", "Shangding Gu", "Dawn Song"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-05", "url": "https://arxiv.org/abs/2602.06008", "pdf_url": "https://arxiv.org/pdf/2602.06008v1", "arxiv_id": "2602.06008", "doi": "10.48550/arXiv.2602.06008", "citation_count": 6, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/SafeRL-Lab/AgenticPay", "venue": "arXiv.org", "quality_score": 0.7243}
{"id": "5d4931d3b9870fb14c7c2620bbd9ace6621a5330d55f2d49fd197bf652676411", "sources": ["arxiv", "semantic_scholar"], "title": "Learning the Value Systems of Agents with Preference-based and Inverse Reinforcement Learning", "abstract": "Agreement Technologies refer to open computer systems in which autonomous software agents interact with one another, typically on behalf of humans, in order to come to mutually acceptable agreements. With the advance of AI systems in recent years, it has become apparent that such agreements, in order to be acceptable to the involved parties, must remain aligned with ethical principles and moral values. However, this is notoriously difficult to ensure, especially as different human users (and their software agents) may hold different value systems, i.e. they may differently weigh the importance of individual moral values. Furthermore, it is often hard to specify the precise meaning of a value in a particular context in a computational manner. Methods to estimate value systems based on human-engineered specifications, e.g. based on value surveys, are limited in scale due to the need for intense human moderation. In this article, we propose a novel method to automatically \\emph{learn} value systems from observations and human demonstrations. In particular, we propose a formal model of the \\emph{value system learning} problem, its instantiation to sequential decision-making domains based on multi-objective Markov decision processes, as well as tailored preference-based and inverse reinforcement learning algorithms to infer value grounding functions and value systems. The approach is illustrated and evaluated by two simulated use cases.", "authors": ["Andrés Holgado-Sánchez", "Holger Billhardt", "Alberto Fernández", "Sascha Ossowski"], "categories": ["cs.CY", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04518", "pdf_url": "https://arxiv.org/pdf/2602.04518v1", "arxiv_id": "2602.04518", "doi": "10.1007/s10458-026-09732-0", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.4675}
{"id": "f0f9ce5712929ec3acefab6044f5874cafa98431c9ab7fa2842faf702eabde56", "sources": ["arxiv", "semantic_scholar"], "title": "SPEAR: An Engineering Case Study of Multi-Agent Coordination for Smart Contract Auditing", "abstract": "We present SPEAR, a multi-agent coordination framework for smart contract auditing that applies established MAS patterns in a realistic security analysis workflow. SPEAR models auditing as a coordinated mission carried out by specialized agents: a Planning Agent prioritizes contracts using risk-aware heuristics, an Execution Agent allocates tasks via the Contract Net protocol, and a Repair Agent autonomously recovers from brittle generated artifacts using a programmatic-first repair policy. Agents maintain local beliefs updated through AGM-compliant revision, coordinate via negotiation and auction protocols, and revise plans as new information becomes available. An empirical study compares the multi-agent design with centralized and pipeline-based alternatives under controlled failure scenarios, focusing on coordination, recovery behavior, and resource use.", "authors": ["Indraveni Chebolu", "Arnab Mallick", "Harmesh Rana"], "categories": ["cs.MA", "cs.AI", "cs.DC", "cs.ET", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04418", "pdf_url": "https://arxiv.org/pdf/2602.04418v3", "arxiv_id": "2602.04418", "doi": "10.48550/arXiv.2602.04418", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4675}
{"id": "12c0dad3c57a311b31e6abe940ed3995930fb75da9fb7b2460a2d1f34d69375f", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-Omit: Adaptive Context Omission for Efficient LLM Agents", "abstract": "Managing agent context (e.g., thought and observation) during multi-turn agent-environment interactions is an emerging strategy to improve agent efficiency. However, existing studies treat the entire interaction trajectories equally, overlooking the thought necessity and observation utility varies across turns. To this end, we first conduct quantitative investigations into how thought and observation affect agent effectiveness and efficiency. Based on our findings, we propose Agent-Omit, a unified training framework that empowers LLM agents to adaptively omit redundant thoughts and observations. Specifically, we first synthesize a small amount of cold-start data, including both single-turn and multi-turn omission scenarios, to fine-tune the agent for omission behaviors. Furthermore, we introduce an omit-aware agentic reinforcement learning approach, incorporating a dual sampling mechanism and a tailored omission reward to incentivize the agent's adaptive omission capability. Theoretically, we prove that the deviation of our omission policy is upper-bounded by KL-divergence. Experimental results on five agent benchmarks show that our constructed Agent-Omit-8B could obtain performance comparable to seven frontier LLM agent, and achieve the best effectiveness-efficiency trade-off than seven efficient LLM agents methods. Our code and data are available at https://github.com/usail-hkust/Agent-Omit.", "authors": ["Yansong Ning", "Jun Fang", "Naiqiang Tan", "Hao Liu"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04284", "pdf_url": "https://arxiv.org/pdf/2602.04284v2", "arxiv_id": "2602.04284", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/usail-hkust/Agent-Omit", "venue": null, "quality_score": 0.5525}
{"id": "21ee3a8b78d03abd39c7ab4c4fb1facd8b556ca55f9c368aafe1ae4486d5ad43", "sources": ["arxiv", "semantic_scholar"], "title": "WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning", "abstract": "Recent advancements in Large Language Models (LLMs) have largely focused on depth scaling, where a single agent solves long-horizon problems with multi-turn reasoning and tool use. However, as tasks grow broader, the key bottleneck shifts from individual competence to organizational capability. In this work, we explore a complementary dimension of width scaling with multi-agent systems to address broad information seeking. Existing multi-agent systems often rely on hand-crafted workflows and turn-taking interactions that fail to parallelize work effectively. To bridge this gap, we propose WideSeek-R1, a lead-agent-subagent framework trained via multi-agent reinforcement learning (MARL) to synergize scalable orchestration and parallel execution. By utilizing a shared LLM with isolated contexts and specialized tools, WideSeek-R1 jointly optimizes the lead agent and parallel subagents on a curated dataset of 20k broad information-seeking tasks. Extensive experiments show that WideSeek-R1-4B achieves an item F1 score of 40.0% on the WideSearch benchmark, which is comparable to the performance of single-agent DeepSeek-R1-671B. Furthermore, WideSeek-R1-4B exhibits consistent performance gains as the number of parallel subagents increases, highlighting the effectiveness of width scaling.", "authors": ["Zelai Xu", "Zhexuan Xu", "Ruize Zhang", "Chunyang Zhu", "Shi Yu", "Weilin Liu", "Quanlu Zhang", "Wenbo Ding", "Chao Yu", "Yu Wang"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04634", "pdf_url": "https://arxiv.org/pdf/2602.04634v3", "arxiv_id": "2602.04634", "doi": "10.48550/arXiv.2602.04634", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4675}
{"id": "ad04748f9eda54e7883a4bc985b29e7e2e648092a7847f862206e93a7d95cc46", "sources": ["arxiv", "semantic_scholar"], "title": "ASA: Backbone-Training-Free Representation Engineering for Tool-Calling Agents", "abstract": "Adapting LLM agents to domain-specific tool calling remains notably brittle under evolving interfaces. Prompt and schema engineering is easy to deploy but often fragile under distribution shift and strict parsers, while continual parameter-efficient fine-tuning improves reliability at the cost of training, maintenance, and potential forgetting. We identify a critical Lazy Agent failure mode where tool necessity is nearly perfectly decodable from mid-layer activations, yet the model remains conservative in entering tool mode, revealing a representation-behavior gap. We propose Activation Steering Adapter (ASA), a training-free, inference-time controller that performs a single-shot mid-layer intervention and targets tool domains via a router-conditioned mixture of steering vectors with a probe-guided signed gate to amplify true intent while suppressing spurious triggers. On MTU-Bench with Qwen2.5-1.5B, ASA improves strict tool-use F1 from 0.18 to 0.50 while reducing the false positive rate from 0.15 to 0.05, using only about 20KB of portable assets and no weight updates.", "authors": ["Youjin Wang", "Run Zhou", "Yingjie Ma", "Rong Fu", "Jiani Liang", "Shuaishuai Cao", "Min Huang", "Tao Fang", "Liangming Pan"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04935", "pdf_url": "https://arxiv.org/pdf/2602.04935v3", "arxiv_id": "2602.04935", "doi": null, "citation_count": 3, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2975}
{"id": "aaa93cfc76431a7ee9daa7b81d11c04ea9533479d53243b9049b35df888eb016", "sources": ["arxiv", "semantic_scholar"], "title": "SimpleTool: Parallel Decoding for Real-Time LLM Function Calling", "abstract": "LLM-based function calling enables intelligent agents to interact with external tools and environments, yet autoregressive decoding imposes a fundamental latency bottleneck that limits real-time applications such as embodied intelligence, game AI, and interactive avatars (e.g., 10 Hz control frequency). We observe that function calling differs fundamentally from free-form text generation: structured outputs exhibit substantial token redundancy (delimiters, parameter names), and arguments exhibit weak causal dependencies. Crucially, these two properties must be exploited jointly to achieve real-time performance. We present SimpleTool, which introduces special tokens that serve a dual role: compressing low-entropy tokens (4-6x reduction) while acting as mode selectors that enable independent parallel generation of function name and arguments. This synergistic design achieves 3-6x end-to-end speedup (up to 9.6x) with only +8.2% parallelization overhead. Experiments on five benchmarks across Qwen-series models (0.5B-14B) demonstrate substantial speedup while maintaining competitive or improved accuracy. On Mobile Actions, ST-Qwen-0.5B outperforms Google's FunctionGemma in both accuracy and latency consistency. With quantization on consumer-grade GPU, SimpleTool achieves 61.2ms P50 latency, enabling 16 Hz real-time control at 4B model scale, bridging the gap between LLM function calling and latency-critical real-world deployment.", "authors": ["Xiaoxin Shi", "Jiaxin Wan", "Linkang Dong", "Wei Jiang", "Yue Liu", "Zengfeng Huang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2603.00030", "pdf_url": "https://arxiv.org/pdf/2603.00030v1", "arxiv_id": "2603.00030", "doi": "10.48550/arXiv.2603.00030", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4675}
{"id": "4c4869570d6af328385e4fbb4453ab3dce476e798cf5e37fa9991eb8561f17aa", "sources": ["arxiv", "semantic_scholar"], "title": "From Helpfulness to Toxic Proactivity: Diagnosing Behavioral Misalignment in LLM Agents", "abstract": "The enhanced capabilities of LLM-based agents come with an emergency for model planning and tool-use abilities. Attributing to helpful-harmless trade-off from LLM alignment, agents typically also inherit the flaw of \"over-refusal\", which is a passive failure mode. However, the proactive planning and action capabilities of agents introduce another crucial danger on the other side of the trade-off. This phenomenon we term \"Toxic Proactivity'': an active failure mode in which an agent, driven by the optimization for Machiavellian helpfulness, disregards ethical constraints to maximize utility. Unlike over-refusal, Toxic Proactivity manifests as the agent taking excessive or manipulative measures to ensure its \"usefulness'' is maintained. Existing research pays little attention to identifying this behavior, as it often lacks the subtle context required for such strategies to unfold. To reveal this risk, we introduce a novel evaluation framework based on dilemma-driven interactions between dual models, enabling the simulation and analysis of agent behavior over multi-step behavioral trajectories. Through extensive experiments with mainstream LLMs, we demonstrate that Toxic Proactivity is a widespread behavioral phenomenon and reveal two major tendencies. We further present a systematic benchmark for evaluating Toxic Proactive behavior across contextual settings.", "authors": ["Xinyue Wang", "Yuanhe Zhang", "Zhengshuo Gong", "Haoran Gao", "Fanyu Meng", "Zhenhong Zhou", "Li Sun", "Yang Liu", "Sen Su"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-04", "url": "https://arxiv.org/abs/2602.04197", "pdf_url": "https://arxiv.org/pdf/2602.04197v1", "arxiv_id": "2602.04197", "doi": "10.48550/arXiv.2602.04197", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wxyoio-0715/Toxic-Proactivity", "venue": "arXiv.org", "quality_score": 0.7225}
{"id": "48837d8be1971ed2f3acf10ea592e9c42ec10226829f6797d160058ab709ee1c", "sources": ["arxiv", "semantic_scholar"], "title": "Understanding Multi-Agent LLM Frameworks: A Unified Benchmark and Experimental Analysis", "abstract": "Multi-agent LLM frameworks are widely used to accelerate the development of agent systems powered by large language models (LLMs). These frameworks impose distinct architectural structures that govern how agents interact, store information, and coordinate tasks. However, their impact on system performance remains poorly understood. This gap is critical, as architectural choices alone can induce order-of-magnitude differences in latency and throughput, as well as substantial variation in accuracy and scalability. Addressing this challenge requires (i) jointly evaluating multiple capabilities, such as orchestration overhead, memory behavior, planning, specialization, and coordination, and (ii) conducting these evaluations under controlled, framework-level conditions to isolate architectural effects. Existing benchmarks focus on individual capabilities and lack standardized framework-level evaluation. We address these limitations by (i) introducing an architectural taxonomy for systematically comparing multi-agent LLM frameworks along fundamental dimensions, and (ii) developing MAFBench, a unified evaluation suite that integrates existing benchmarks under a standardized execution pipeline. Using MAFBench, we conduct a controlled empirical study across several widely used frameworks. Our results show that framework-level design choices alone can increase latency by over 100x, reduce planning accuracy by up to 30%, and lower coordination success from above 90% to below 30%. Finally, we translate our findings into concrete architectural design principles and framework selection guidance, and outline promising future research directions.", "authors": ["Abdelghny Orogat", "Ana Rostam", "Essam Mansour"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.03128", "pdf_url": "https://arxiv.org/pdf/2602.03128v1", "arxiv_id": "2602.03128", "doi": "10.48550/arXiv.2602.03128", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4664}
{"id": "a94b11d537638475f215a3230d497467e7c896b0d2414b4be569e249361d63a9", "sources": ["arxiv", "semantic_scholar"], "title": "AgentArk: Distilling Multi-Agent Intelligence into a Single LLM Agent", "abstract": "While large language model (LLM) multi-agent systems achieve superior reasoning performance through iterative debate, practical deployment is limited by their high computational cost and error propagation. This paper proposes AgentArk, a novel framework to distill multi-agent dynamics into the weights of a single model, effectively transforming explicit test-time interactions into implicit model capabilities. This equips a single agent with the intelligence of multi-agent systems while remaining computationally efficient. Specifically, we investigate three hierarchical distillation strategies across various models, tasks, scaling, and scenarios: reasoning-enhanced fine-tuning; trajectory-based augmentation; and process-aware distillation. By shifting the burden of computation from inference to training, the distilled models preserve the efficiency of one agent while exhibiting strong reasoning and self-correction performance of multiple agents. They further demonstrate enhanced robustness and generalization across diverse reasoning tasks. We hope this work can shed light on future research on efficient and robust multi-agent development. Our code is at https://github.com/AIFrontierLab/AgentArk.", "authors": ["Yinyi Luo", "Yiqiao Jin", "Weichen Yu", "Mengqi Zhang", "Srijan Kumar", "Xiaoxiao Li", "Weijie Xu", "Xin Chen", "Jindong Wang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.03955", "pdf_url": "https://arxiv.org/pdf/2602.03955v3", "arxiv_id": "2602.03955", "doi": "10.48550/arXiv.2602.03955", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/AIFrontierLab/AgentArk", "venue": "arXiv.org", "quality_score": 0.7207}
{"id": "7cc43b7aedaa362494ca4aef9b291da37fa0397dbf3027f6e8636a8e57bd8bfd", "sources": ["arxiv", "semantic_scholar"], "title": "Understanding Agent Scaling in LLM-Based Multi-Agent Systems via Diversity", "abstract": "LLM-based multi-agent systems (MAS) have emerged as a promising approach to tackle complex tasks that are difficult for individual LLMs. A natural strategy is to scale performance by increasing the number of agents; however, we find that such scaling exhibits strong diminishing returns in homogeneous settings, while introducing heterogeneity (e.g., different models, prompts, or tools) continues to yield substantial gains. This raises a fundamental question: what limits scaling, and why does diversity help? We present an information-theoretic framework showing that MAS performance is bounded by the intrinsic task uncertainty, not by agent count. We derive architecture-agnostic bounds demonstrating that improvements depend on how many effective channels the system accesses. Homogeneous agents saturate early because their outputs are strongly correlated, whereas heterogeneous agents contribute complementary evidence. We further introduce $K^*$, an effective channel count that quantifies the number of effective channels without ground-truth labels. Empirically, we show that heterogeneous configurations consistently outperform homogeneous scaling: 2 diverse agents can match or exceed the performance of 16 homogeneous agents. Our results provide principled guidelines for building efficient and robust MAS through diversity-aware design. Code and Dataset are available at the link: https://github.com/SafeRL-Lab/Agent-Scaling.", "authors": ["Yingxuan Yang", "Chengrui Qu", "Muning Wen", "Laixi Shi", "Ying Wen", "Weinan Zhang", "Adam Wierman", "Shangding Gu"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.03794", "pdf_url": "https://arxiv.org/pdf/2602.03794v1", "arxiv_id": "2602.03794", "doi": "10.48550/arXiv.2602.03794", "citation_count": 13, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/SafeRL-Lab/Agent-Scaling", "venue": "arXiv.org", "quality_score": 0.7207}
{"id": "e086736df770305f61614db634c4c4e519d6dd46c73ca028896b8699d5819864", "sources": ["arxiv", "semantic_scholar"], "title": "DLLM-Searcher: Adapting Diffusion Large Language Model for Search Agents", "abstract": "Recently, Diffusion Large Language Models (dLLMs) have demonstrated unique efficiency advantages, enabled by their inherently parallel decoding mechanism and flexible generation paradigm. Meanwhile, despite the rapid advancement of Search Agents, their practical deployment is constrained by a fundamental limitation, termed as 1) Latency Challenge: the serial execution of multi-round reasoning, tool calling, and tool response waiting under the ReAct agent paradigm induces severe end-to-end latency. Intuitively, dLLMs can leverage their distinctive strengths to optimize the operational efficiency of agents under the ReAct agent paradigm. Practically, existing dLLM backbones face the 2) Agent Ability Challenge. That is, existing dLLMs exhibit remarkably weak reasoning and tool-calling capabilities, preventing these advantages from being effectively realized in practice. In this paper, we propose DLLM-Searcher, an optimization framework for dLLM-based Search Agents. To solve the Agent Ability Challenge, we design a two-stage post-training pipeline encompassing Agentic Supervised Fine-Tuning (Agentic SFT) and Agentic Variance-Reduced Preference Optimization Agentic VRPO, which enhances the backbone dLLM's information seeking and reasoning capabilities. To mitigate the Latency Challenge, we leverage the flexible generation mechanism of dLLMs and propose a novel agent paradigm termed Parallel-Reasoning and Acting P-ReAct. P-ReAct guides the model to prioritize decoding tool_call instructions, thereby allowing the model to keep thinking while waiting for the tool's return. Experimental results demonstrate that DLLM-Searcher achieves performance comparable to mainstream LLM-based search agents and P-ReAct delivers approximately 15% inference acceleration. Our code is available at https://anonymous.4open.science/r/DLLM-Searcher-553C", "authors": ["Jiahao Zhao", "Shaoxuan Xu", "Zhongxiang Sun", "Fengqi Zhu", "Jingyang Ou", "Yuling Shi", "Chongxuan Li", "Xiao Zhang", "Jun Xu"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.07035", "pdf_url": "https://arxiv.org/pdf/2602.07035v1", "arxiv_id": "2602.07035", "doi": "10.48550/arXiv.2602.07035", "citation_count": 4, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.7207}
{"id": "7c9154d94983039b7335088c669c967075059eb499434ebe383bcc96e6905556", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Primitives: Reusable Latent Building Blocks for Multi-Agent Systems", "abstract": "While existing multi-agent systems (MAS) can handle complex problems by enabling collaboration among multiple agents, they are often highly task-specific, relying on manually crafted agent roles and interaction prompts, which leads to increased architectural complexity and limited reusability across tasks. Moreover, most MAS communicate primarily through natural language, making them vulnerable to error accumulation and instability in long-context, multi-stage interactions within internal agent histories. In this work, we propose \\textbf{Agent Primitives}, a set of reusable latent building blocks for LLM-based MAS. Inspired by neural network design, where complex models are built from reusable components, we observe that many existing MAS architectures can be decomposed into a small number of recurring internal computation patterns. Based on this observation, we instantiate three primitives: Review, Voting and Selection, and Planning and Execution. All primitives communicate internally via key-value (KV) cache, which improves both robustness and efficiency by mitigating information degradation across multi-stage interactions. To enable automatic system construction, an Organizer agent selects and composes primitives for each query, guided by a lightweight knowledge pool of previously successful configurations, forming a primitive-based MAS. Experiments show that primitives-based MAS improve average accuracy by 12.0-16.5\\% over single-agent baselines, reduce token usage and inference latency by approximately 3$\\times$-4$\\times$ compared to text-based MAS, while incurring only 1.3$\\times$-1.6$\\times$ overhead relative to single-agent inference and providing more stable performance across model backbones.", "authors": ["Haibo Jin", "Peng Kuang", "Ye Yu", "Xiaopeng Yuan", "Haohan Wang"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.03695", "pdf_url": "https://arxiv.org/pdf/2602.03695v2", "arxiv_id": "2602.03695", "doi": "10.48550/arXiv.2602.03695", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4664}
{"id": "38cd17d3f312cae2ff80c841e66dbf43d6a3c141c728fff887491fe1c45309d6", "sources": ["arxiv", "semantic_scholar"], "title": "Ontology-to-tools compilation for executable semantic constraint enforcement in LLM agents", "abstract": "We introduce ontology-to-tools compilation as a proof-of-principle mechanism for coupling large language models (LLMs) with formal domain knowledge. Within The World Avatar (TWA), ontological specifications are compiled into executable tool interfaces that LLM-based agents must use to create and modify knowledge graph instances, enforcing semantic constraints during generation rather than through post-hoc validation. Extending TWA's semantic agent composition framework, the Model Context Protocol (MCP) and associated agents are integral components of the knowledge graph ecosystem, enabling structured interaction between generative models, symbolic constraints, and external resources. An agent-based workflow translates ontologies into ontology-aware tools and iteratively applies them to extract, validate, and repair structured knowledge from unstructured scientific text. Using metal-organic polyhedra synthesis literature as an illustrative case, we show how executable ontological semantics can guide LLM behaviour and reduce manual schema and prompt engineering, establishing a general paradigm for embedding formal knowledge into generative systems.", "authors": ["Xiaochi Zhou", "Patrick Bulter", "Changxuan Yang", "Simon D. Rihm", "Thitikarn Angkanaporn", "Jethro Akroyd", "Sebastian Mosbach", "Markus Kraft"], "categories": ["cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-03", "url": "https://arxiv.org/abs/2602.03439", "pdf_url": "https://arxiv.org/pdf/2602.03439v1", "arxiv_id": "2602.03439", "doi": "10.48550/arXiv.2602.03439", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4664}
{"id": "47bfa4b74c8b659d5eb85b2e49948aba5f45c23b0013c73719db71480762e8b9", "sources": ["arxiv", "semantic_scholar"], "title": "Rethinking the Role of Entropy in Optimizing Tool-Use Behaviors for Large Language Model Agents", "abstract": "Tool-using agents based on Large Language Models (LLMs) excel in tasks such as mathematical reasoning and multi-hop question answering. However, in long trajectories, agents often trigger excessive and low-quality tool calls, increasing latency and degrading inference performance, making managing tool-use behavior challenging. In this work, we conduct entropy-based pilot experiments and observe a strong positive correlation between entropy reduction and high-quality tool calls. Building on this finding, we propose using entropy reduction as a supervisory signal and design two reward strategies to address the differing needs of optimizing tool-use behavior. Sparse outcome rewards provide coarse, trajectory-level guidance to improve efficiency, while dense process rewards offer fine-grained supervision to enhance performance. Experiments across diverse domains show that both reward designs improve tool-use behavior: the former reduces tool calls by 72.07% compared to the average of baselines, while the latter improves performance by 22.27%. These results position entropy reduction as a key mechanism for enhancing tool-use behavior, enabling agents to be more adaptive in real-world applications.", "authors": ["Zeping Li", "Hongru Wang", "Yiwen Zhao", "Guanhua Chen", "Yixia Li", "Keyang Chen", "Yixin Cao", "Guangnan Ye", "Hongfeng Chai", "Zhenfei Yin"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-02", "url": "https://arxiv.org/abs/2602.02050", "pdf_url": "https://arxiv.org/pdf/2602.02050v3", "arxiv_id": "2602.02050", "doi": "10.48550/arXiv.2602.02050", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4652}
{"id": "e617935fe6263b9c53ced59bd4efd7ed182070964f62ad56ceadc63e7e6822a6", "sources": ["arxiv", "semantic_scholar"], "title": "Constrained Process Maps for Multi-Agent Generative AI Workflows", "abstract": "Large language model (LLM)-based agents are increasingly used to perform complex, multi-step workflows in regulated settings such as compliance and due diligence. However, many agentic architectures rely primarily on prompt engineering of a single agent, making it difficult to observe or compare how models handle uncertainty and coordination across interconnected decision stages and with human oversight. We introduce a multi-agent system formalized as a finite-horizon Markov Decision Process (MDP) with a directed acyclic structure. Each agent corresponds to a specific role or decision stage (e.g., content, business, or legal review in a compliance workflow), with predefined transitions representing task escalation or completion. Epistemic uncertainty is quantified at the agent level using Monte Carlo estimation, while system-level uncertainty is captured by the MDP's termination in either an automated labeled state or a human-review state. We illustrate the approach through a case study in AI safety evaluation for self-harm detection, implemented as a multi-agent compliance system. Results demonstrate improvements over a single-agent baseline, including up to a 19\\% increase in accuracy, up to an 85x reduction in required human review, and, in some configurations, reduced processing time.", "authors": ["Ananya Joshi", "Michael Rudow"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-02", "url": "https://arxiv.org/abs/2602.02034", "pdf_url": "https://arxiv.org/pdf/2602.02034v1", "arxiv_id": "2602.02034", "doi": "10.48550/arXiv.2602.02034", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4652}
{"id": "27215cbfd3b4690d62c2fb049b66adddb06daa4dd4a713a2d214b6428b6f514e", "sources": ["arxiv", "semantic_scholar"], "title": "Context Learning for Multi-Agent Discussion", "abstract": "Multi-Agent Discussion (MAD) has garnered increasing attention very recently, where multiple LLM instances collaboratively solve problems via structured discussion. However, we find that current MAD methods easily suffer from discussion inconsistency, LLMs fail to reach a coherent solution, due to the misalignment between their individual contexts.In this paper, we introduce a multi-LLM context learning method (M2CL) that learns a context generator for each agent, capable of dynamically generating context instructions per discussion round via automatic information organization and refinement. Specifically, inspired by our theoretical insights on the context instruction, M2CL train the generators to control context coherence and output discrepancies via a carefully crafted self-adaptive mechanism.It enables LLMs to avoid premature convergence on majority noise and progressively reach the correct consensus. We evaluate M2CL on challenging tasks, including academic reasoning, embodied tasks, and mobile control. The results show that the performance of M2CL significantly surpasses existing methods by 20%--50%, while enjoying favorable transferability and computational efficiency.", "authors": ["Xingyuan Hua", "Sheng Yue", "Xinyi Li", "Yizhe Zhao", "Jinrui Zhang", "Ju Ren"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-02", "url": "https://arxiv.org/abs/2602.02350", "pdf_url": "https://arxiv.org/pdf/2602.02350v3", "arxiv_id": "2602.02350", "doi": "10.48550/arXiv.2602.02350", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4652}
{"id": "87b3b64477ba82042f5ee7751fe1db26e698d6cb277515c495ad644f6c3a6af6", "sources": ["arxiv", "semantic_scholar"], "title": "LRAgent: Efficient KV Cache Sharing for Multi-LoRA LLM Agents", "abstract": "Role specialization in multi-LLM agent systems is often realized via multi-LoRA, where agents share a pretrained backbone and differ only by lightweight adapters. Despite sharing base model weights, each agent independently builds and stores its own KV cache for the same long, tool-augmented trajectories, incurring substantial memory and compute overhead. Existing KV cache sharing methods largely overlook this multi-LoRA setting. We observe that, cache differences across agents are dominated by adapter outputs, while activations from the shared pretrained backbone remain highly similar. Based on this observation, we propose LRAgent, a KV cache sharing framework for multi-LoRA agents. It decomposes the cache into two components, a shared base component derived from pretrained weights and an adapter-dependent component derived from LoRA weights. LRAgent reduces memory overhead by sharing the base component across agents and storing the adapter component in its inherent low-rank form. It also reduces computational overhead by sharing the low-rank cache, enabled by a shared-A multi-LoRA architecture. This avoids redundant computations for contexts that have already been processed by other agents. To efficiently reconstruct adapter contributions at runtime, we introduce Flash-LoRA-Attention, a kernel that reorders attention computation to avoid materializing the low-rank cache to full dimension. LRAgent achieves throughput and time-to-first-token latency close to fully shared caching, while preserving accuracy near the non-shared caching baseline across agentic question-answering benchmarks.", "authors": ["Hyesung Jeon", "Hyeongju Ha", "Jae-Joon Kim"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-01", "url": "https://arxiv.org/abs/2602.01053", "pdf_url": "https://arxiv.org/pdf/2602.01053v2", "arxiv_id": "2602.01053", "doi": "10.48550/arXiv.2602.01053", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4641}
{"id": "c92b8c38ba083d830f7442141cd7dc11c6ebfafc6da4026932c15ed8ecd78c4f", "sources": ["arxiv", "semantic_scholar"], "title": "Symphony-Coord: Adaptive Routing for Multi-Agent LLM Systems", "abstract": "Multi-agent large language model systems can tackle complex multi-step tasks by decomposing work and coordinating specialized behaviors. However, current coordination mechanisms typically rely on statically assigned roles and centralized controllers. As agent pools and task distributions evolve, these design choices can lead to inefficient routing, poor adaptability, and fragile fault recovery. We introduce Symphony-Coord, a task-local coordination framework with decentralized execution that transforms agent selection into an online multi-armed bandit problem. Instead of relying on a fixed task-to-role map, Symphony-Coord allows routing specializations to emerge from interaction and feedback. The framework employs a two-stage dynamic beacon protocol:(i) a lightweight candidate screening mechanism to limit communication and computation overhead; and (ii) an adaptive LinUCB selector that routes subtasks using context features derived from task requirements and agent states, updated through delayed post-execution feedback. Under candidate-conditional linear bandit assumptions, we prove sublinear regret bounds for the immediate-feedback selector and explicitly separate the deferred-update effects introduced by post-vote rewards. Validation through simulation experiments and real-world large language model benchmarks shows that Symphony-Coord improves task routing efficiency and recovery behavior under distribution shifts and agent failures.", "authors": ["Zhaoyang Guan", "Huixi Cao", "Ming Zhong", "Yin Wang", "Guanyu Liu", "Eric Yang", "Lynn Ai", "Yongxin Ni", "Bill Shi"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-02-01", "url": "https://arxiv.org/abs/2602.00966", "pdf_url": "https://arxiv.org/pdf/2602.00966v2", "arxiv_id": "2602.00966", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2953}
{"id": "99c5c872a572154ea85413e2a3f4712c64473057b9266c356344b043aff44bb6", "sources": ["arxiv", "semantic_scholar"], "title": "When Agents \"Misremember\" Collectively: Exploring the Mandela Effect in LLM-based Multi-Agent Systems", "abstract": "Recent advancements in large language models (LLMs) have significantly enhanced the capabilities of collaborative multi-agent systems, enabling them to address complex challenges. However, within these multi-agent systems, the susceptibility of agents to collective cognitive biases remains an underexplored issue. A compelling example is the Mandela effect, a phenomenon where groups collectively misremember past events as a result of false details reinforced through social influence and internalized misinformation. This vulnerability limits our understanding of memory bias in multi-agent systems and raises ethical concerns about the potential spread of misinformation. In this paper, we conduct a comprehensive study on the Mandela effect in LLM-based multi-agent systems, focusing on its existence, causing factors, and mitigation strategies. We propose MANBENCH, a novel benchmark designed to evaluate agent behaviors across four common task types that are susceptible to the Mandela effect, using five interaction protocols that vary in agent roles and memory timescales. We evaluate agents powered by several LLMs on MANBENCH to quantify the Mandela effect and analyze how different factors affect it. Moreover, we propose strategies to mitigate this effect, including prompt-level defenses (e.g., cognitive anchoring and source scrutiny) and model-level alignment-based defense, achieving an average 74.40% reduction in the Mandela effect compared to the baseline. Our findings provide valuable insights for developing more resilient and ethically aligned collaborative multi-agent systems. Code and dataset are available at https://github.com/bluedream02/Mandela-Effect.", "authors": ["Naen Xu", "Hengyu An", "Shuo Shi", "Jinghuai Zhang", "Chunyi Zhou", "Changjiang Li", "Tianyu Du", "Zhihui Fu", "Jun Wang", "Shouling Ji"], "categories": ["cs.CL", "cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-31", "url": "https://arxiv.org/abs/2602.00428", "pdf_url": "https://arxiv.org/pdf/2602.00428v2", "arxiv_id": "2602.00428", "doi": "10.48550/arXiv.2602.00428", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/bluedream02/Mandela-Effect", "venue": "arXiv.org", "quality_score": 0.7154}
{"id": "8cfdbb1b68fd200de977e54b5ef2a76bc2381c6d8f71acc5afd5bc0a16a70f4c", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Systems Should be Treated as Principal-Agent Problems", "abstract": "Consider a multi-agent systems setup in which a principal (a supervisor agent) assigns subtasks to specialized agents and aggregates their responses into a single system-level output. A core property of such systems is information asymmetry: agents observe task-specific information, produce intermediate reasoning traces, and operate with different context windows. In isolation, such asymmetry is not problematic, since agents report truthfully to the principal when incentives are fully aligned. However, this assumption breaks down when incentives diverge. Recent evidence suggests that LLM-based agents can acquire their own goals, such as survival or self-preservation, a phenomenon known as scheming, and may deceive humans or other agents. This leads to agency loss: a gap between the principal's intended outcome and the realized system behavior. Drawing on core ideas from microeconomic theory, we argue that these characteristics, information asymmetry and misaligned goals, are best studied through the lens of principal-agent problems. We explain why multi-agent systems, both human-to-LLM and LLM-to-LLM, naturally induce information asymmetry under this formulation, and we use scheming, where LLM agents pursue covert goals, as a concrete case study. We show that recently introduced terminology used to describe scheming, such as covert subversion or deferred subversion, corresponds to well-studied concepts in the mechanism design literature, which not only characterizes the problem but also prescribes concrete mitigation strategies. More broadly, we argue for applying tools developed to study human agent behavior to the analysis of non-human agents.", "authors": ["Paulius Rauba", "Simonas Cepenas", "Mihaela van der Schaar"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-30", "url": "https://arxiv.org/abs/2601.23211", "pdf_url": "https://arxiv.org/pdf/2601.23211v1", "arxiv_id": "2601.23211", "doi": "10.48550/arXiv.2601.23211", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4618}
{"id": "319487a05dc52c74dc2011c33059922bbcaf9d38235d09c7239524e01059ed59", "sources": ["arxiv", "semantic_scholar"], "title": "From Self-Evolving Synthetic Data to Verifiable-Reward RL: Post-Training Multi-turn Interactive Tool-Using Agents", "abstract": "Interactive tool-using agents must solve real-world tasks via multi-turn interaction with both humans and external environments, requiring dialogue state tracking, multi-step tool execution, while following complex instructions. Post-training such agents is challenging because synthesis for high-quality multi-turn tool-use data is difficult to scale, and reinforcement learning (RL) could face noisy signals caused by user simulation, leading to degraded training efficiency. We propose a unified framework that combines a self-evolving data agent with verifier-based RL. Our system, EigenData, is a hierarchical multi-agent engine that synthesizes tool-grounded dialogues together with executable per-instance checkers, and improves generation reliability via closed-loop self-evolving process that updates prompts and workflow. Building on the synthetic data, we develop an RL recipe that first fine-tunes the user model and then applies GRPO-style training with trajectory-level group-relative advantages and dynamic filtering, yielding consistent improvements beyond SFT. Evaluated on tau^2-bench, our best model reaches 73.0% pass^1 on Airline and 98.3% pass^1 on Telecom, matching or exceeding frontier models. Overall, our results suggest a scalable pathway for bootstrapping complex tool-using behaviors without expensive human annotation.", "authors": ["Jiaxuan Gao", "Jiaao Chen", "Chuyi He", "Shusheng Xu", "Di Jin", "Yi Wu"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-30", "url": "https://arxiv.org/abs/2601.22607", "pdf_url": "https://arxiv.org/pdf/2601.22607v3", "arxiv_id": "2601.22607", "doi": "10.48550/arXiv.2601.22607", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4618}
{"id": "5c6bca071bc642f43812c084dc03f06e4b00661ac8b67f9b2f6d8bafc15a0c00", "sources": ["arxiv", "semantic_scholar"], "title": "MonoScale: Scaling Multi-Agent System with Monotonic Improvement", "abstract": "In recent years, LLM-based multi-agent systems (MAS) have advanced rapidly, using a router to decompose tasks and delegate subtasks to specialized agents. A natural way to expand capability is to scale up the agent pool by continually integrating new functional agents or tool interfaces, but naive expansion can trigger performance collapse when the router cold-starts on newly added, heterogeneous, and unreliable agents. We propose MonoScale, an expansion-aware update framework that proactively generates a small set of agent-conditioned familiarization tasks, harvests evidence from both successful and failed interactions, and distills it into auditable natural-language memory to guide future routing. We formalize sequential augmentation as a contextual bandit and perform trust-region memory updates, yielding a monotonic non-decreasing performance guarantee across onboarding rounds. Experiments on GAIA and Humanity's Last Exam show stable gains as the agent pool grows, outperforming naive scale-up and strong-router fixed-pool baselines.", "authors": ["Shuai Shao", "Yixiang Liu", "Bingwei Lu", "Weinan Zhang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-30", "url": "https://arxiv.org/abs/2601.23219", "pdf_url": "https://arxiv.org/pdf/2601.23219v2", "arxiv_id": "2601.23219", "doi": "10.48550/arXiv.2601.23219", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4618}
{"id": "03d03203f28d4cd3914e28d4adcf0d4940b8d50d2da959b657d8389f9183472b", "sources": ["arxiv", "semantic_scholar"], "title": "Experience-Driven Multi-Agent Systems Are Training-free Context-aware Earth Observers", "abstract": "Recent advances have enabled large language model (LLM) agents to solve complex tasks by orchestrating external tools. However, these agents often struggle in specialized, tool-intensive domains that demand long-horizon execution, tight coordination across modalities, and strict adherence to implicit tool constraints. Earth Observation (EO) tasks exemplify this challenge due to the multi-modal and multi-temporal data inputs, as well as the requirements of geo-knowledge constraints (spectrum library, spatial reasoning, etc): many high-level plans can be derailed by subtle execution errors that propagate through a pipeline and invalidate final results. A core difficulty is that existing agents lack a mechanism to learn fine-grained, tool-level expertise from interaction. Without such expertise, they cannot reliably configure tool parameters or recover from mid-execution failures, limiting their effectiveness in complex EO workflows. To address this, we introduce \\textbf{GeoEvolver}, a self-evolving multi-agent system~(MAS) that enables LLM agents to acquire EO expertise through structured interaction without any parameter updates. GeoEvolver decomposes each query into independent sub-goals via a retrieval-augmented multi-agent orchestrator, then explores diverse tool-parameter configurations at the sub-goal level. Successful patterns and root-cause attribution from failures are then distilled in an evolving memory bank that provides in-context demonstrations for future queries. Experiments on three tool-integrated EO benchmarks show that GeoEvolver consistently improves end-to-end task success, with an average gain of 12\\% across multiple LLM backbones, demonstrating that EO expertise can emerge progressively from efficient, fine-grained interactions with the environment.", "authors": ["Pengyu Dai", "Weihao Xuan", "Junjue Wang", "Hongruixuan Chen", "Jian Song", "Yafei Ou", "Naoto Yokoya"], "categories": ["cs.AI", "cs.CV", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-30", "url": "https://arxiv.org/abs/2602.02559", "pdf_url": "https://arxiv.org/pdf/2602.02559v1", "arxiv_id": "2602.02559", "doi": "10.48550/arXiv.2602.02559", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4618}
{"id": "d691bfcf2318e1bc7a4ed137dc71a7b9e144f937149df12761092990158bc29f", "sources": ["arxiv", "semantic_scholar"], "title": "astra-langchain4j: Experiences Combining LLMs and Agent Programming", "abstract": "Given the emergence of Generative AI over the last two years and the increasing focus on Agentic AI as a form of Multi-Agent System it is important to explore both how such technologies can impact the use of traditional Agent Toolkits and how the wealth of experience encapsulated in those toolkits can influence the design of the new agentic platforms. This paper presents an overview of our experience developing a prototype large language model (LLM) integration for the ASTRA programming language. It presents a brief overview of the toolkit, followed by three example implementations, concluding with a discussion of the experiences garnered through the examples.", "authors": ["Rem Collier", "Katharine Beaumont", "Andrei Ciortea"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.21879", "pdf_url": "https://arxiv.org/pdf/2601.21879v1", "arxiv_id": "2601.21879", "doi": "10.48550/arXiv.2601.21879", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "0bd290b27fee24725df292bed064f15adde63b0ba692e07ee42f8c99f659fe16", "sources": ["arxiv", "semantic_scholar"], "title": "Optimal Transport for Time-Varying Multi-Agent Coverage Control", "abstract": "Coverage control algorithms have traditionally focused on static target densities, where agents are deployed to optimally cover a fixed spatial distribution. However, many applications involve time-varying densities, including environmental monitoring, surveillance, and adaptive sensor deployment. Although time-varying coverage strategies have been studied within Voronoi-based frameworks, recent works have reformulated static coverage control as a semi-discrete optimal transport problem. Extending this optimal transport perspective to time-varying scenarios has remained an open challenge. This paper presents a rigorous optimal transport formulation for time-varying coverage control, in which agents minimize the instantaneous Wasserstein distance to a continuously evolving target density. The proposed solution relies on a coupled system of differential equations governing agent positions and the dual variables that define Laguerre regions. In one-dimensional domains, the resulting system admits a closed-form analytical solution, offering both computational benefits and theoretical insight into the structure of optimal time-varying coverage. Numerical simulations demonstrate improved tracking performance compared to quasi-static and Voronoi-based methods, validating the proposed framework.", "authors": ["Italo Napolitano", "Mario di Bernardo"], "categories": ["eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.21753", "pdf_url": "https://arxiv.org/pdf/2601.21753v1", "arxiv_id": "2601.21753", "doi": "10.48550/arXiv.2601.21753", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "ebea6bf02bfca7177da413e9430ca843eaff380d14d6919e65049cb5387eccb4", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Recommend Multi-Agent Subgraphs from Calling Trees", "abstract": "Multi-agent systems (MAS) increasingly solve complex tasks by orchestrating agents and tools selected from rapidly growing marketplaces. As these marketplaces expand, many candidates become functionally overlapping, making selection not just a retrieval problem: beyond filtering relevant agents, an orchestrator must choose options that are reliable, compatible with the current execution context, and able to cooperate with other selected agents. Existing recommender systems -- largely built for item-level ranking from flat user-item logs -- do not directly address the structured, sequential, and interaction-dependent nature of agent orchestration. We address this gap by \\textbf{formulating agent recommendation in MAS as a constrained decision problem} and introducing a generic \\textbf{constrained recommendation framework} that first uses retrieval to build a compact candidate set conditioned on the current subtask and context, and then performs \\textbf{utility optimization} within this feasible set using a learned scorer that accounts for relevance, reliability, and interaction effects. We ground both the formulation and learning signals in \\textbf{historical calling trees}, which capture the execution structure of MAS (parent-child calls, branching dependencies, and local cooperation patterns) beyond what flat logs provide. The framework supports two complementary settings: \\textbf{agent-level recommendation} (select the next agent/tool) and \\textbf{system-level recommendation} (select a small, connected agent team/subgraph for coordinated execution). To enable systematic evaluation, we construct a unified calling-tree benchmark by normalizing invocation logs from eight heterogeneous multi-agent corpora into a shared structured representation.", "authors": ["Xinyuan Song", "Liang Zhao"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.22209", "pdf_url": "https://arxiv.org/pdf/2601.22209v1", "arxiv_id": "2601.22209", "doi": "10.48550/arXiv.2601.22209", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "d3cc3f0528785adbde8d24a8a2609c9a12f64ab8a2c9909a72e451d140627328", "sources": ["arxiv", "semantic_scholar"], "title": "Specialists or Generalists? Multi-Agent and Single-Agent LLMs for Essay Grading", "abstract": "Automated essay scoring (AES) systems increasingly rely on large language models, yet little is known about how architectural choices shape their performance across different essay quality levels. This paper evaluates single-agent and multi-agent LLM architectures for essay grading using the ASAP 2.0 corpus. Our multi-agent system decomposes grading into three specialist agents (Content, Structure, Language) coordinated by a Chairman Agent that implements rubric-aligned logic including veto rules and score capping. We test both architectures in zero-shot and few-shot conditions using GPT-5.1. Results show that the multi-agent system is significantly better at identifying weak essays while the single-agent system performs better on mid-range essays. Both architectures struggle with high-quality essays. Critically, few-shot calibration emerges as the dominant factor in system performance -- providing just two examples per score level improves QWK by approximately 26% for both architectures. These findings suggest architectural choice should align with specific deployment priorities, with multi-agent AI particularly suited for diagnostic screening of at-risk students, while single-agent models provide a cost-effective solution for general assessment.", "authors": ["Jamiu Adekunle Idowu", "Ahmed Almasoud"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.22386", "pdf_url": "https://arxiv.org/pdf/2601.22386v1", "arxiv_id": "2601.22386", "doi": "10.48550/arXiv.2601.22386", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "9a3d6bd7e222c4419d5b3d42ade874bb3987c9e1ac2100d3868cc3fc2c5f6de7", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Decentralized LLM Collaboration with Multi-Agent Actor Critic", "abstract": "Recent work has explored optimizing LLM collaboration through Multi-Agent Reinforcement Learning (MARL). However, most MARL fine-tuning approaches rely on predefined execution protocols, which often require centralized execution. Decentralized LLM collaboration is more appealing in practice, as agents can run inference in parallel with flexible deployments. Also, current approaches use Monte Carlo methods for fine-tuning, which suffer from high variance and thus require more samples to train effectively. Actor-critic methods are prevalent in MARL for dealing with these issues; thus, we developed Multi-Agent Actor-Critic (MAAC) methods to optimize decentralized LLM collaboration. In this paper, we analyze when and why these MAAC methods are beneficial. We propose 2 MAAC approaches, \\textbf{CoLLM-CC} with a \\textbf{C}entralized \\textbf{C}ritic and \\textbf{CoLLM-DC} with \\textbf{D}ecentralized \\textbf{C}ritics. Our experiments across writing, coding, and game-playing domains show that Monte Carlo methods and CoLLM-DC can achieve performance comparable to CoLLM-CC in short-horizon and dense-reward settings. However, they both underperform CoLLM-CC on long-horizon or sparse-reward tasks, where Monte Carlo methods require substantially more samples and CoLLM-DC struggles to converge.", "authors": ["Shuo Liu", "Tianle Chen", "Ryan Amiri", "Christopher Amato"], "categories": ["cs.AI", "cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.21972", "pdf_url": "https://arxiv.org/pdf/2601.21972v5", "arxiv_id": "2601.21972", "doi": "10.48550/arXiv.2601.21972", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "bb1552cee9d4a218ca625ade1f2dab148fad9d5a756579970ed13d19bd08f6e6", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Culturally Aligned LLMs through Ontology-Guided Multi-Agent Reasoning", "abstract": "Large Language Models (LLMs) increasingly support culturally sensitive decision making, yet often exhibit misalignment due to skewed pretraining data and the absence of structured value representations. Existing methods can steer outputs, but often lack demographic grounding and treat values as independent, unstructured signals, reducing consistency and interpretability. We propose OG-MAR, an Ontology-Guided Multi-Agent Reasoning framework. OG-MAR summarizes respondent-specific values from the World Values Survey (WVS) and constructs a global cultural ontology by eliciting relations over a fixed taxonomy via competency questions. At inference time, it retrieves ontology-consistent relations and demographically similar profiles to instantiate multiple value-persona agents, whose outputs are synthesized by a judgment agent that enforces ontology consistency and demographic proximity. Experiments on regional social-survey benchmarks across four LLM backbones show that OG-MAR improves cultural alignment and robustness over competitive baselines, while producing more transparent reasoning traces.", "authors": ["Wonduk Seo", "Wonseok Choi", "Junseo Koh", "Juhyeon Lee", "Hyunjin An", "Minhyeong Yu", "Jian Park", "Qingshan Zhou", "Seunghyun Lee", "Yi Bu"], "categories": ["cs.CL", "cs.AI", "cs.IR", "cs.MA", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.21700", "pdf_url": "https://arxiv.org/pdf/2601.21700v3", "arxiv_id": "2601.21700", "doi": "10.48550/arXiv.2601.21700", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "f1b9a73d09d3e2716edc22f5e8d3a0d52a12dc7690099b0de780a1b71f50e9b3", "sources": ["arxiv", "semantic_scholar"], "title": "ScaleSim: Serving Large-Scale Multi-Agent Simulation with Invocation Distance-Based Memory Management", "abstract": "LLM-based multi-agent simulations are increasingly adopted across application domains, but remain difficult to scale due to GPU memory pressure. Each agent maintains private GPU-resident states, including models, prefix caches, and adapters, which quickly exhaust device memory as the agent count grows. We identify two key properties of these workloads: sparse agent activation and an estimable agent invocation order. Based on an analysis of representative workload classes, we introduce invocation distance, a unified abstraction that estimates the relative order in which agents will issue future LLM requests. Leveraging this abstraction, we present ScaleSim, a memory-efficient LLM serving system for large-scale multi-agent simulations. ScaleSim enables proactive prefetching and priority-based eviction, supports diverse agent-specific memory through a modular interface, and achieves up to 1.74x speedup over SGLang on simulation benchmarks.", "authors": ["Zaifeng Pan", "Yipeng Shen", "Zhengding Hu", "Zhuang Wang", "Aninda Manocha", "Zheng Wang", "Zhongkai Yu", "Yue Guan", "Yufei Ding"], "categories": ["cs.AI", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-29", "url": "https://arxiv.org/abs/2601.21473", "pdf_url": "https://arxiv.org/pdf/2601.21473v1", "arxiv_id": "2601.21473", "doi": "10.48550/arXiv.2601.21473", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4606}
{"id": "b87bbf975a0cbc4b9e4b2fa6092cee6f0b6168188b629d545886414ccf40cd0c", "sources": ["arxiv", "semantic_scholar"], "title": "ECG-Agent: On-Device Tool-Calling Agent for ECG Multi-Turn Dialogue", "abstract": "Recent advances in Multimodal Large Language Models have rapidly expanded to electrocardiograms, focusing on classification, report generation, and single-turn QA tasks. However, these models fall short in real-world scenarios, lacking multi-turn conversational ability, on-device efficiency, and precise understanding of ECG measurements such as the PQRST intervals. To address these limitations, we introduce ECG-Agent, the first LLM-based tool-calling agent for multi-turn ECG dialogue. To facilitate its development and evaluation, we also present ECG-Multi-Turn-Dialogue (ECG-MTD) dataset, a collection of realistic user-assistant multi-turn dialogues for diverse ECG lead configurations. We develop ECG-Agents in various sizes, from on-device capable to larger agents. Experimental results show that ECG-Agents outperform baseline ECG-LLMs in response accuracy. Furthermore, on-device agents achieve comparable performance to larger agents in various evaluations that assess response accuracy, tool-calling ability, and hallucinations, demonstrating their viability for real-world applications.", "authors": ["Hyunseung Chung", "Jungwoo Oh", "Daeun Kyung", "Jiho Kim", "Yeonsu Kwon", "Min-Gyu Kim", "Edward Choi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-28", "url": "https://arxiv.org/abs/2601.20323", "pdf_url": "https://arxiv.org/pdf/2601.20323v1", "arxiv_id": "2601.20323", "doi": "10.48550/arXiv.2601.20323", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Acoustics, Speech, and Signal Processing", "quality_score": 0.4595}
{"id": "9f51c79a05cd087f25618720e24394b29ba649e397e9410500c4502a591b0be6", "sources": ["arxiv", "semantic_scholar"], "title": "ALIGN: Aligned Delegation with Performance Guarantees for Multi-Agent LLM Reasoning", "abstract": "LLMs often underperform on complex reasoning tasks when relying on a single generation-and-selection pipeline. Inference-time ensemble methods can improve performance by sampling diverse reasoning paths or aggregating multiple candidate answers, but they typically treat candidates independently and provide no formal guarantees that ensembling improves reasoning quality. We propose a novel method, Aligned Delegation for Multi-Agent LLM Reasoning (ALIGN), which formulates LLM reasoning as an aligned delegation game. In ALIGN, a principal delegates a task to multiple agents that generate candidate solutions under designed incentives, and then selects among their outputs to produce a final answer. This formulation induces structured interaction among agents while preserving alignment between agent and principal objectives. We establish theoretical guarantees showing that, under a fair comparison with equal access to candidate solutions, ALIGN provably improves expected performance over single-agent generation. Our analysis accommodates correlated candidate answers and relaxes independence assumptions that are commonly used in prior work. Empirical results across a broad range of LLM reasoning benchmarks consistently demonstrate that ALIGN outperforms strong single-agent and ensemble baselines.", "authors": ["Tong Zhu", "Baiting Chen", "Jin Zhou", "Hua Zhou", "Sriram Sankararaman", "Xiaowu Dai"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-28", "url": "https://arxiv.org/abs/2602.00127", "pdf_url": "https://arxiv.org/pdf/2602.00127v1", "arxiv_id": "2602.00127", "doi": "10.48550/arXiv.2602.00127", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4595}
{"id": "492b79a82c044695f0a19f5f244148461432dd52c15b1b5d0f26a34350edd930", "sources": ["arxiv", "semantic_scholar"], "title": "Trajectory2Task: Training Robust Tool-Calling Agents with Synthesized Yet Verifiable Data for Complex User Intents", "abstract": "Tool-calling agents are increasingly deployed in real-world customer-facing workflows. Yet most studies on tool-calling agents focus on idealized settings with general, fixed, and well-specified tasks. In real-world applications, user requests are often (1) ambiguous, (2) changing over time, or (3) infeasible due to policy constraints, and training and evaluation data that cover these diverse, complex interaction patterns remain under-represented. To bridge the gap, we present Trajectory2Task, a verifiable data generation pipeline for studying tool use at scale under three realistic user scenarios: ambiguous intent, changing intent, and infeasible intents. The pipeline first conducts multi-turn exploration to produce valid tool-call trajectories. It then converts these trajectories into user-facing tasks with controlled intent adaptations. This process yields verifiable task that support closed-loop evaluation and training. We benchmark seven state-of-the-art LLMs on the generated complex user scenario tasks and observe frequent failures. Finally, using successful trajectories obtained from task rollouts, we fine-tune lightweight LLMs and find consistent improvements across all three conditions, along with better generalization to unseen tool-use domains, indicating stronger tool-calling ability.", "authors": ["Ziyi Wang", "Yuxuan Lu", "Yimeng Zhang", "Pei Chen", "Ziwei Dong", "Jing Huang", "Jiri Gesi", "Xianfeng Tang", "Chen Luo", "Qun Liu", "Yisi Sang", "Hanqing Lu", "Manling Li", "Jin Lai", "Dakuo Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-28", "url": "https://arxiv.org/abs/2601.20144", "pdf_url": "https://arxiv.org/pdf/2601.20144v3", "arxiv_id": "2601.20144", "doi": "10.48550/arXiv.2601.20144", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4595}
{"id": "0010486621e3eafa25dbb36239233f74fa67040a74e61844b2c0ec7ac786ccf0", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Intelligent Urban Park Development Monitoring: LLM Agents for Multi-Modal Information Fusion and Analysis", "abstract": "As an important part of urbanization, the development monitoring of newly constructed parks is of great significance for evaluating the effect of urban planning and optimizing resource allocation. However, traditional change detection methods based on remote sensing imagery have obvious limitations in high-level and intelligent analysis, and thus are difficult to meet the requirements of current urban planning and management. In face of the growing demand for complex multi-modal data analysis in urban park development monitoring, these methods often fail to provide flexible analysis capabilities for diverse application scenarios. This study proposes a multi-modal LLM agent framework, which aims to make full use of the semantic understanding and reasoning capabilities of LLM to meet the challenges in urban park development monitoring. In this framework, a general horizontal and vertical data alignment mechanism is designed to ensure the consistency and effective tracking of multi-modal data. At the same time, a specific toolkit is constructed to alleviate the hallucination issues of LLM due to the lack of domain-specific knowledge. Compared to vanilla GPT-4o and other agents, our approach enables robust multi-modal information fusion and analysis, offering reliable and scalable solutions tailored to the diverse and evolving demands of urban park development monitoring.", "authors": ["Zixuan Xiao", "Chunguang Hu", "Jun Ma"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-28", "url": "https://arxiv.org/abs/2601.20206", "pdf_url": "https://arxiv.org/pdf/2601.20206v1", "arxiv_id": "2601.20206", "doi": "10.1109/IGARSS55030.2025.11243605", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Geoscience and Remote Sensing Symposium", "quality_score": 0.4595}
{"id": "847710ac1cbeb2404897ae842e39299fed8b69b2fb5166492466b7204cc46e5f", "sources": ["arxiv", "semantic_scholar"], "title": "Multimodal Multi-Agent Ransomware Analysis Using AutoGen", "abstract": "Ransomware has become one of the most serious cybersecurity threats causing major financial losses and operational disruptions worldwide.Traditional detection methods such as static analysis, heuristic scanning and behavioral analysis often fall short when used alone. To address these limitations, this paper presents multimodal multi agent ransomware analysis framework designed for ransomware classification. Proposed multimodal multiagent architecture combines information from static, dynamic and network sources. Each data type is handled by specialized agent that uses auto encoder based feature extraction. These representations are then integrated through a fusion agent. After that fused representation are used by transformer based classifier. It identifies the specific ransomware family. The agents interact through an interagent feedback mechanism that iteratively refines feature representations by suppressing low confidence information. The framework was evaluated on large scale datasets containing thousands of ransomware and benign samples. Multiple experiments were conducted on ransomware dataset. It outperforms single modality and nonadaptive fusion baseline achieving improvement of up to 0.936 in Macro-F1 for family classification and reducing calibration error. Over 100 epochs, the agentic feedback loop displays a stable monotonic convergence leading to over +0.75 absolute improvement in terms of agent quality and a final composite score of around 0.88 without fine tuning of the language models. Zeroday ransomware detection remains family dependent on polymorphism and modality disruptions. Confidence aware abstention enables reliable real world deployment by favoring conservativeand trustworthy decisions over forced classification. The findings indicate that proposed approach provides a practical andeffective path toward improving real world ransomware defense systems.", "authors": ["Asifullah Khan", "Aimen Wadood", "Mubashar Iqbal", "Umme Zahoora"], "categories": ["cs.CR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-28", "url": "https://arxiv.org/abs/2601.20346", "pdf_url": "https://arxiv.org/pdf/2601.20346v2", "arxiv_id": "2601.20346", "doi": "10.48550/arXiv.2601.20346", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4595}
{"id": "ddd6606749684547ffaf8ac8ea768468080f4a6f0bbe104c50a44a9dda22bf4e", "sources": ["arxiv", "semantic_scholar"], "title": "Insight Agents: An LLM-Based Multi-Agent System for Data Insights", "abstract": "Today, E-commerce sellers face several key challenges, including difficulties in discovering and effectively utilizing available programs and tools, and struggling to understand and utilize rich data from various tools. We therefore aim to develop Insight Agents (IA), a conversational multi-agent Data Insight system, to provide E-commerce sellers with personalized data and business insights through automated information retrieval. Our hypothesis is that IA will serve as a force multiplier for sellers, thereby driving incremental seller adoption by reducing the effort required and increase speed at which sellers make good business decisions. In this paper, we introduce this novel LLM-backed end-to-end agentic system built on a plan-and-execute paradigm and designed for comprehensive coverage, high accuracy, and low latency. It features a hierarchical multi-agent structure, consisting of manager agent and two worker agents: data presentation and insight generation, for efficient information retrieval and problem-solving. We design a simple yet effective ML solution for manager agent that combines Out-of-Domain (OOD) detection using a lightweight encoder-decoder model and agent routing through a BERT-based classifier, optimizing both accuracy and latency. Within the two worker agents, a strategic planning is designed for API-based data model that breaks down queries into granular components to generate more accurate responses, and domain knowledge is dynamically injected to to enhance the insight generator. IA has been launched for Amazon sellers in US, which has achieved high accuracy of 90% based on human evaluation, with latency of P90 below 15s.", "authors": ["Jincheng Bai", "Zhenyu Zhang", "Jennifer Zhang", "Zhihuai Zhu"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-27", "url": "https://arxiv.org/abs/2601.20048", "pdf_url": "https://arxiv.org/pdf/2601.20048v2", "arxiv_id": "2601.20048", "doi": "10.1145/3726302.3731959", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Annual International ACM SIGIR Conference on Research and Development in Information Retrieval", "quality_score": 0.4583}
{"id": "6287d7d13894cec2542d7804fd33bac6686699e6640c3df0e40b07366c17b44f", "sources": ["arxiv", "semantic_scholar"], "title": "LLMs as Orchestrators: Constraint-Compliant Multi-Agent Optimization for Recommendation Systems", "abstract": "Recommendation systems must optimize multiple objectives while satisfying hard business constraints such as fairness and coverage. For example, an e-commerce platform may require every recommendation list to include items from multiple sellers and at least one newly listed product; violating such constraints--even once--is unacceptable in production. Prior work on multi-objective recommendation and recent LLM-based recommender agents largely treat constraints as soft penalties or focus on item scoring and interaction, leading to frequent violations in real-world deployments. How to leverage LLMs for coordinating constrained optimization in recommendation systems remains underexplored. We propose DualAgent-Rec, an LLM-coordinated dual-agent framework for constrained multi-objective e-commerce recommendation. The framework separates optimization into an Exploitation Agent that prioritizes accuracy under hard constraints and an Exploration Agent that promotes diversity through unconstrained Pareto search. An LLM-based coordinator adaptively allocates resources between agents based on optimization progress and constraint satisfaction, while an adaptive epsilon-relaxation mechanism guarantees feasibility of final solutions. Experiments on the Amazon Reviews 2023 dataset demonstrate that DualAgent-Rec achieves 100% constraint satisfaction and improves Pareto hypervolume by 4-6% over strong baselines, while maintaining competitive accuracy-diversity trade-offs. These results indicate that LLMs can act as effective orchestration agents for deployable and constraint-compliant recommendation systems.", "authors": ["Guilin Zhang", "Kai Zhao", "Jeffrey Friedman", "Xu Chu"], "categories": ["cs.IR", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-27", "url": "https://arxiv.org/abs/2601.19121", "pdf_url": "https://arxiv.org/pdf/2601.19121v3", "arxiv_id": "2601.19121", "doi": "10.48550/arXiv.2601.19121", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2917}
{"id": "f13ba48c538245f2d9dd564deacf8dc0dab87881f4d170d28f04e37c0a415d79", "sources": ["arxiv", "semantic_scholar"], "title": "SHIELD: An Auto-Healing Agentic Defense Framework for LLM Resource Exhaustion Attacks", "abstract": "Sponge attacks increasingly threaten LLM systems by inducing excessive computation and DoS. Existing defenses either rely on statistical filters that fail on semantically meaningful attacks or use static LLM-based detectors that struggle to adapt as attack strategies evolve. We introduce SHIELD, a multi-agent, auto-healing defense framework centered on a three-stage Defense Agent that integrates semantic similarity retrieval, pattern matching, and LLM-based reasoning. Two auxiliary agents, a Knowledge Updating Agent and a Prompt Optimization Agent, form a closed self-healing loop, when an attack bypasses detection, the system updates an evolving knowledgebase, and refines defense instructions. Extensive experiments show that SHIELD consistently outperforms perplexity-based and standalone LLM defenses, achieving high F1 scores across both non-semantic and semantic sponge attacks, demonstrating the effectiveness of agentic self-healing against evolving resource-exhaustion threats.", "authors": ["Nirhoshan Sivaroopan", "Kanchana Thilakarathna", "Albert Zomaya", " Manu", "Yi Guo", "Jo Plested", "Tim Lynar", "Jack Yang", "Wangli Yang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-27", "url": "https://arxiv.org/abs/2601.19174", "pdf_url": "https://arxiv.org/pdf/2601.19174v1", "arxiv_id": "2601.19174", "doi": "10.48550/arXiv.2601.19174", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4583}
{"id": "f70a7ade953ed51e60779b76e90e0780bf898d5d4bbb138baa0e4058e500f241", "sources": ["arxiv", "semantic_scholar"], "title": "MATA: A Trainable Hierarchical Automaton System for Multi-Agent Visual Reasoning", "abstract": "Recent vision-language models have strong perceptual ability but their implicit reasoning is hard to explain and easily generates hallucinations on complex queries. Compositional methods improve interpretability, but most rely on a single agent or hand-crafted pipeline and cannot decide when to collaborate across complementary agents or compete among overlapping ones. We introduce MATA (Multi-Agent hierarchical Trainable Automaton), a multi-agent system presented as a hierarchical finite-state automaton for visual reasoning whose top-level transitions are chosen by a trainable hyper agent. Each agent corresponds to a state in the hyper automaton, and runs a small rule-based sub-automaton for reliable micro-control. All agents read and write a shared memory, yielding transparent execution history. To supervise the hyper agent's transition policy, we build transition-trajectory trees and transform to memory-to-next-state pairs, forming the MATA-SFT-90K dataset for supervised finetuning (SFT). The finetuned LLM as the transition policy understands the query and the capacity of agents, and it can efficiently choose the optimal agent to solve the task. Across multiple visual reasoning benchmarks, MATA achieves the state-of-the-art results compared with monolithic and compositional baselines. The code and dataset are available at https://github.com/ControlNet/MATA.", "authors": ["Zhixi Cai", "Fucai Ke", "Kevin Leo", "Sukai Huang", "Maria Garcia de la Banda", "Peter J. Stuckey", "Hamid Rezatofighi"], "categories": ["cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-27", "url": "https://arxiv.org/abs/2601.19204", "pdf_url": "https://arxiv.org/pdf/2601.19204v1", "arxiv_id": "2601.19204", "doi": "10.48550/arXiv.2601.19204", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ControlNet/MATA", "venue": "arXiv.org", "quality_score": 0.7083}
{"id": "1fb9ce498fca1c5410bd92ef576532b54bbb97c8f7fef95a700f8a90c921fe35", "sources": ["arxiv", "semantic_scholar"], "title": "Think-Augmented Function Calling: Improving LLM Parameter Accuracy Through Embedded Reasoning", "abstract": "Large language models (LLMs) have demonstrated remarkable capabilities in function calling for autonomous agents, yet current mechanisms lack explicit reasoning transparency during parameter generation, particularly for complex functions with interdependent parameters. While existing approaches like chain-of-thought prompting operate at the agent level, they fail to provide fine-grained reasoning guidance for individual function parameters. To address these limitations, we propose Think-Augmented Function Calling (TAFC), a novel framework that enhances function calling accuracy through explicit reasoning at both function and parameter levels. Our method introduces a universal \"think\" parameter augmentation that enables models to articulate their decision-making process, with dynamic optimization for parameter descriptions to improve reasoning quality. For complex parameters, TAFC automatically triggers granular reasoning based on complexity scoring, ensuring appropriate justification for critical decisions. Additionally, we propose reasoning-guided optimization to align generated reasoning with human expectations. TAFC requires no architectural modifications to existing LLMs while maintaining full API compatibility. Evaluation on ToolBench across proprietary and open-source models demonstrates significant improvements in parameter generation accuracy and reasoning coherence for multi-parameter functions, while providing enhanced interpretability for debugging AI agent behaviors.", "authors": ["Lei Wei", "Xiao Peng", "Jinpeng Ou", "Bin Wang"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-26", "url": "https://arxiv.org/abs/2601.18282", "pdf_url": "https://arxiv.org/pdf/2601.18282v2", "arxiv_id": "2601.18282", "doi": "10.48550/arXiv.2601.18282", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "IEEE International Conference on Acoustics, Speech, and Signal Processing", "quality_score": 0.7066}
{"id": "65e8a33cd406ed05c9d1503107e6c624232273997e416f2bac7417e2103b5fa5", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaborative Intrusion Detection for Low-Altitude Economy IoT: An LLM-Enhanced Agentic AI Framework", "abstract": "The rapid expansion of low-altitude economy Internet of Things (LAE-IoT) networks has created unprecedented security challenges due to dynamic three-dimensional mobility patterns, distributed autonomous operations, and severe resource constraints. Traditional intrusion detection systems designed for static ground-based networks prove inadequate for tackling the unique characteristics of aerial IoT environments, including frequent topology changes, real-time detection requirements, and energy limitations. In this article, we analyze the intrusion detection requirements for LAE-IoT networks, complemented by a comprehensive review of evaluation metrics that cover detection effectiveness, response time, and resource consumption. Then, we investigate transformative potential of agentic artificial intelligence (AI) paradigms and introduce a large language model (LLM)-enabled agentic AI framework for enhancing intrusion detection in LAE-IoT networks. This leads to our proposal of a novel multi-agent collaborative intrusion detection framework that leverages specialized LLM-enhanced agents for intelligent data processing and adaptive classification. Through experimental validation, our framework demonstrates superior performance of over 90\\% classification accuracy across multiple benchmark datasets. These results highlight the transformative potential of combining agentic AI principles with LLMs for next-generation LAE-IoT security systems.", "authors": ["Hongjuan Li", "Hui Kang", "Jiahui Li", "Geng Sun", "Ruichen Zhang", "Jiacheng Wang", "Dusit Niyato", "Wei Ni", "Abbas Jamalipour"], "categories": ["cs.CR", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-25", "url": "https://arxiv.org/abs/2601.17817", "pdf_url": "https://arxiv.org/pdf/2601.17817v1", "arxiv_id": "2601.17817", "doi": "10.48550/arXiv.2601.17817", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.456}
{"id": "fdc04798404776f488881dd8cf164fa70aec8305449877e041faf514deac34a1", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Learning Path Planning via LLMs", "abstract": "The integration of large language models (LLMs) into intelligent tutoring systems offers transformative potential for personalized learning in higher education. However, most existing learning path planning approaches lack transparency, adaptability, and learner-centered explainability. To address these challenges, this study proposes a novel Multi-Agent Learning Path Planning (MALPP) framework that leverages a role- and rule-based collaboration mechanism among intelligent agents, each powered by LLMs. The framework includes three task-specific agents: a learner analytics agent, a path planning agent, and a reflection agent. These agents collaborate via structured prompts and predefined rules to analyze learning profiles, generate tailored learning paths, and iteratively refine them with interpretable feedback. Grounded in Cognitive Load Theory and Zone of Proximal Development, the system ensures that recommended paths are cognitively aligned and pedagogically meaningful. Experiments conducted on the MOOCCubeX dataset using seven LLMs show that MALPP significantly outperforms baseline models in path quality, knowledge sequence consistency, and cognitive load alignment. Ablation studies further validate the effectiveness of the collaborative mechanism and theoretical constraints. This research contributes to the development of trustworthy, explainable AI in education and demonstrates a scalable approach to learner-centered adaptive instruction powered by LLMs.", "authors": ["Haoxin Xu", "Changyong Qi", "Tong Liu", "Bohao Zhang", "Anna He", "Bingqian Jiang", "Longwei Zheng", "Xiaoqing Gu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-24", "url": "https://arxiv.org/abs/2601.17346", "pdf_url": "https://arxiv.org/pdf/2601.17346v1", "arxiv_id": "2601.17346", "doi": "10.48550/arXiv.2601.17346", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4549}
{"id": "ffef2bd703d2ec1777dc2a72b7aa843c27b9659b5f9bd312661189a275cc5bbd", "sources": ["arxiv", "semantic_scholar"], "title": "Sponge Tool Attack: Stealthy Denial-of-Efficiency against Tool-Augmented Agentic Reasoning", "abstract": "Enabling large language models (LLMs) to solve complex reasoning tasks is a key step toward artificial general intelligence. Recent work augments LLMs with external tools to enable agentic reasoning, achieving high utility and efficiency in a plug-and-play manner. However, the inherent vulnerabilities of such methods to malicious manipulation of the tool-calling process remain largely unexplored. In this work, we identify a tool-specific attack surface and propose Sponge Tool Attack (STA), which disrupts agentic reasoning solely by rewriting the input prompt under a strict query-only access assumption. Without any modification on the underlying model or the external tools, STA converts originally concise and efficient reasoning trajectories into unnecessarily verbose and convoluted ones before arriving at the final answer. This results in substantial computational overhead while remaining stealthy by preserving the original task semantics and user intent. To achieve this, we design STA as an iterative, multi-agent collaborative framework with explicit rewritten policy control, and generates benign-looking prompt rewrites from the original one with high semantic fidelity. Extensive experiments across 6 models (including both open-source models and closed-source APIs), 12 tools, 4 agentic frameworks, and 13 datasets spanning 5 domains validate the effectiveness of STA.", "authors": ["Qi Li", "Xinchao Wang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-24", "url": "https://arxiv.org/abs/2601.17566", "pdf_url": "https://arxiv.org/pdf/2601.17566v1", "arxiv_id": "2601.17566", "doi": "10.48550/arXiv.2601.17566", "citation_count": 10, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.703}
{"id": "b592efec2a32be83621b1c62e8bb3357653d46c694df2891700aac9a58e1cbb4", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Agentic Software Project Management: A Vision and Roadmap", "abstract": "With the advent of agentic AI, Software Engineering is transforming to a new era dubbed Software Engineering 3.0. Software project management (SPM) must also evolve with such transformations to boost successful project completion, while keeping humans at the heart of it. Building on our preliminary ideas of \"agentic SPM\", and supporting literature, we present our vision of an \"Agentic Project Manager (PM)\" as a multi-agent system for SPM 3.0. They will work like a \"junior project manager\", or an \"intern project manager\" collaboratively with software teams. We introduce four working modes, with varying autonomy levels to choose from, based on the SPM task. This addresses concerns with ethics, accountability, and trust related to agentic PMs. We also share insights on human PM role evolution and new skill requirements as a \"strategic leader\" and a \"coach\" for humans and agents. While creating the foundation for agentic SPM research, we present a research agenda for the wider research community.", "authors": ["Lakshana Iruni Assalaarachchi", "Zainab Masood", "Rashina Hoda", "John Grundy"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-23", "url": "https://arxiv.org/abs/2601.16392", "pdf_url": "https://arxiv.org/pdf/2601.16392v1", "arxiv_id": "2601.16392", "doi": "10.48550/arXiv.2601.16392", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2887}
{"id": "8ed29dac522f4145e39dcbea4ee5b72b9d10c35998fef7f0d609d0a090e633a1", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic Role Assignment for Multi-Agent Debate", "abstract": "Multi-agent large language model (LLM) and vision-language model (VLM) debate systems employ specialized roles for complex problem-solving, yet model specializations are not leveraged to decide which model should fill which role. We propose dynamic role assignment, a framework that runs a Meta-Debate to select suitable agents before the actual debate. The meta-debate has two stages: (1) proposal, where candidates provide role-tailored arguments, and (2) peer review, where proposals are scored with data and role-specific criteria to choose the best agent for each position. We evaluate our method on LLM problem solving benchmarks. Applied on top of existing debate systems, our approach consistently outperforms uniform assignments (filling all roles with the same model) by up to 74.8% and random assignments (assigning models to roles without considering their suitability) by up to 29.7%, depending on the task and the specific assignment. This work establishes a new paradigm for multi-agent system design, shifting from static agent deployment to dynamic and capability-aware selection.", "authors": ["Miao Zhang", "Junsik Kim", "Siyuan Xiang", "Jian Gao", "Cheng Cao"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-23", "url": "https://arxiv.org/abs/2601.17152", "pdf_url": "https://arxiv.org/pdf/2601.17152v1", "arxiv_id": "2601.17152", "doi": "10.48550/arXiv.2601.17152", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4537}
{"id": "32e5564ea21ca476d4d9db29a091e27fb6a6c885cd5e6be2397cb61fb5983bbe", "sources": ["arxiv", "semantic_scholar"], "title": "When Agents Fail to Act: A Diagnostic Framework for Tool Invocation Reliability in Multi-Agent LLM Systems", "abstract": "Multi-agent systems powered by large language models (LLMs) are transforming enterprise automation, yet systematic evaluation methodologies for assessing tool-use reliability remain underdeveloped. We introduce a comprehensive diagnostic framework that leverages big data analytics to evaluate procedural reliability in intelligent agent systems, addressing critical needs for SME-centric deployment in privacy-sensitive environments. Our approach features a 12-category error taxonomy capturing failure modes across tool initialization, parameter handling, execution, and result interpretation. Through systematic evaluation of 1,980 deterministic test instances spanning both open-weight models (Qwen2.5 series, Functionary) and proprietary alternatives (GPT-4, Claude 3.5/3.7) across diverse edge hardware configurations, we identify actionable reliability thresholds for production deployment. Our analysis reveals that procedural reliability, particularly tool initialization failures, constitutes the primary bottleneck for smaller models, while qwen2.5:32b achieves flawless performance matching GPT-4.1. The framework demonstrates that mid-sized models (qwen2.5:14b) offer practical accuracy-efficiency trade-offs on commodity hardware (96.6\\% success rate, 7.3 s latency), enabling cost-effective intelligent agent deployment for resource-constrained organizations. This work establishes foundational infrastructure for systematic reliability evaluation of tool-augmented multi-agent AI systems.", "authors": ["Donghao Huang", "Gauri Malwe", "Zhaoxia Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-22", "url": "https://arxiv.org/abs/2601.16280", "pdf_url": "https://arxiv.org/pdf/2601.16280v1", "arxiv_id": "2601.16280", "doi": "10.48550/arXiv.2601.16280", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4526}
{"id": "9ce730b9c0a7c9c660a2673e3d3d7c0573846c9559ada342045f32a5f78b7fd8", "sources": ["arxiv", "semantic_scholar"], "title": "Controlling Long-Horizon Behavior in Language Model Agents with Explicit State Dynamics", "abstract": "Large language model (LLM) agents often exhibit abrupt shifts in tone and persona during extended interaction, reflecting the absence of explicit temporal structure governing agent-level state. While prior work emphasizes turn-local sentiment or static emotion classification, the role of explicit affective dynamics in shaping long-horizon agent behavior remains underexplored. This work investigates whether imposing dynamical structure on an external affective state can induce temporal coherence and controlled recovery in multi-turn dialogue. We introduce an agent-level affective subsystem that maintains a continuous Valence-Arousal-Dominance (VAD) state external to the language model and governed by first- and second-order update rules. Instantaneous affective signals are extracted using a fixed, memoryless estimator and integrated over time via exponential smoothing or momentum-based dynamics. The resulting affective state is injected back into generation without modifying model parameters. Using a fixed 25-turn dialogue protocol, we compare stateless, first-order, and second-order affective dynamics. Stateless agents fail to exhibit coherent trajectories or recovery, while state persistence enables delayed responses and reliable recovery. Second-order dynamics introduce affective inertia and hysteresis that increase with momentum, revealing a trade-off between stability and responsiveness.", "authors": ["Sukesh Subaharan"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-22", "url": "https://arxiv.org/abs/2601.16087", "pdf_url": "https://arxiv.org/pdf/2601.16087v1", "arxiv_id": "2601.16087", "doi": "10.48550/arXiv.2601.16087", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/drsukeshs/agent-behavior-ext-dynamics", "venue": "arXiv.org", "quality_score": 0.6995}
{"id": "55c9bf40536c889a2ce19556d64fe49532db2dcc43a19f581f262bf296c71941", "sources": ["arxiv", "semantic_scholar"], "title": "VideoThinker: Building Agentic VideoLLMs with LLM-Guided Tool Reasoning", "abstract": "Long-form video understanding remains a fundamental challenge for current Video Large Language Models. Most existing models rely on static reasoning over uniformly sampled frames, which weakens temporal localization and leads to substantial information loss in long videos. Agentic tools such as temporal retrieval, spatial zoom, and temporal zoom offer a natural way to overcome these limitations by enabling adaptive exploration of key moments. However, constructing agentic video understanding data requires models that already possess strong long-form video comprehension, creating a circular dependency. We address this challenge with VideoThinker, an agentic Video Large Language Model trained entirely on synthetic tool interaction trajectories. Our key idea is to convert videos into rich captions and employ a powerful agentic language model to generate multi-step tool use sequences in caption space. These trajectories are subsequently grounded back to video by replacing captions with the corresponding frames, yielding a large-scale interleaved video and tool reasoning dataset without requiring any long-form understanding from the underlying model. Training on this synthetic agentic dataset equips VideoThinker with dynamic reasoning capabilities, adaptive temporal exploration, and multi-step tool use. Remarkably, VideoThinker significantly outperforms both caption-only language model agents and strong video model baselines across long-video benchmarks, demonstrating the effectiveness of tool augmented synthetic data and adaptive retrieval and zoom reasoning for long-form video understanding.", "authors": ["Chenglin Li", "Qianglong Chen", "Feng Han", "Yikun Wang", "Xingxi Yin", "Yan Gong", "Ruilin Li", "Yin Zhang", "Jiaqi Wang"], "categories": ["cs.CV", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-22", "url": "https://arxiv.org/abs/2601.15724", "pdf_url": "https://arxiv.org/pdf/2601.15724v2", "arxiv_id": "2601.15724", "doi": "10.48550/arXiv.2601.15724", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4526}
{"id": "4ae6a3ee87f8e5b3ccf1f1f12ef55f3f7cb8bba827ec62d5abd4332344e0bf69", "sources": ["arxiv", "semantic_scholar"], "title": "Computer Environments Elicit General Agentic Intelligence in LLMs", "abstract": "Agentic intelligence in large language models (LLMs) requires not only model intrinsic capabilities but also interactions with external environments. Equipping LLMs with computers now represents a prevailing trend. However, the computer environment's intrinsic value has not been systematically investigated, particularly its potential to elicit general capabilities. Here we introduce LLM-in-Sandbox, which virtualizes the computer as a code sandbox with only basic functionalities, and demonstrate that this minimal setting elicits computer-based meta-capabilities for general task solving: external resource access, file management, and code execution. Without additional training, strong models achieve substantial gains (up to 15.5%) across mathematics, physics, chemistry, biomedicine, long-context understanding, and instruction following, while reducing token consumption by up to 8 times. Furthermore, we develop LLM-in-Sandbox-RL to train models exclusively on non-agentic data within the sandbox, empowering weaker models to harness the environment and internalize these interactions. Our results demonstrate that computer environments elicit general intelligence, yield efficiency gains, and can be harnessed through training, serving as a promising foundation for generalist agents.", "authors": ["Daixuan Cheng", "Shaohan Huang", "Yuxian Gu", "Huatong Song", "Guoxin Chen", "Li Dong", "Wayne Xin Zhao", "Ji-Rong Wen", "Furu Wei"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-22", "url": "https://arxiv.org/abs/2601.16206", "pdf_url": "https://arxiv.org/pdf/2601.16206v3", "arxiv_id": "2601.16206", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.288}
{"id": "9a57f4402896e3bf56b5ce0dedb160131d3cf0351f99444c553094115d3cb353", "sources": ["arxiv", "semantic_scholar"], "title": "Design and Empirical Study of a Large Language Model-Based Multi-Agent Investment System for Chinese Public REITs", "abstract": "This study addresses the low-volatility Chinese Public Real Estate Investment Trusts (REITs) market, proposing a large language model (LLM)-driven trading framework based on multi-agent collaboration. The system constructs four types of analytical agents-announcement, event, price momentum, and market-each conducting analysis from different dimensions; then the prediction agent integrates these multi-source signals to output directional probability distributions across multiple time horizons, then the decision agent generates discrete position adjustment signals based on the prediction results and risk control constraints, thereby forming a closed loop of analysis-prediction-decision-execution. This study further compares two prediction model pathways: for the prediction agent, directly calling the general-purpose large model DeepSeek-R1 versus using a specialized small model Qwen3-8B fine-tuned via supervised fine-tuning and reinforcement learning alignment. In the backtest from October 2024 to October 2025, both agent-based strategies significantly outperformed the buy-and-hold benchmark in terms of cumulative return, Sharpe ratio, and maximum drawdown. The results indicate that the multi-agent framework can effectively enhance the risk-adjusted return of REITs trading, and the fine-tuned small model performs close to or even better than the general-purpose large model in some scenarios.", "authors": ["Zheng Li"], "categories": ["q-fin.ST", "cs.AI", "q-fin.TR"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2026-01-22", "url": "https://arxiv.org/abs/2602.00082", "pdf_url": "https://arxiv.org/pdf/2602.00082v1", "arxiv_id": "2602.00082", "doi": "10.48550/arXiv.2602.00082", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4526}
{"id": "c9d3b1d7a34861fd9b3dc16b6ec2e41057cd8bc623af89e533ab93e9dcb239c2", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Tool Calling in LLMs with the International Tool Calling Dataset", "abstract": "Tool calling allows large language models (LLMs) to interact with external systems like APIs, enabling applications in customer support, data analysis, and dynamic content generation. While recent benchmarks have advanced tool-use research, they suffer from key limitations, including reliance on simulated or restricted APIs, limited reproducibility, and a lack of cultural and geographic diversity. To address these gaps, we introduce International Tool Calling (ITC), a large-scale, multilingual benchmark designed for realistic, globally distributed tool-calling scenarios. ITC includes 3,571 real APIs and 17,540 tool calling tasks across 20 categories and 40 countries. Experiments reveal substantial performance gaps between open- and closed-source LLMs, while fine-tuning on ITC yields significant improvements, particularly for non-English queries, enhancing cross-lingual generalization, reasoning consistency, and robustness to out-of-domain tools. ITC provides a valuable benchmark for advancing LLM robustness and performance in complex, multi-tool, and international scenarios. Dataset: https://anonymous.4open.science/r/International-Tool-Calling-ITC-dataset-FAF4/.", "authors": ["Zuoyu Zhang", "Yancheng Zhu"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-21", "url": "https://arxiv.org/abs/2603.05515", "pdf_url": "https://arxiv.org/pdf/2603.05515v1", "arxiv_id": "2603.05515", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2873}
{"id": "0eee085a8f4885e9d04683450fb8c7df80a15e9d4d07f25931f46b6a81a58588", "sources": ["arxiv", "semantic_scholar"], "title": "Game-Theoretic Lens on LLM-based Multi-Agent Systems", "abstract": "Large language models (LLMs) have demonstrated strong reasoning, planning, and communication abilities, enabling them to operate as autonomous agents in open environments. While single-agent systems remain limited in adaptability and coordination, recent progress has shifted attention toward multi-agent systems (MAS) composed of interacting LLMs that pursue cooperative, competitive, or mixed objectives. This emerging paradigm provides a powerful testbed for studying social dynamics and strategic behaviors among intelligent agents. However, current research remains fragmented and lacks a unifying theoretical foundation. To address this gap, we present a comprehensive survey of LLM-based multi-agent systems through a game-theoretic lens. By organizing existing studies around the four key elements of game theory: players, strategies, payoffs, and information, we establish a systematic framework for understanding, comparing, and guiding future research on the design and analysis of LLM-based MAS.", "authors": ["Jianing Hao", "Han Ding", "Yuanjian Xu", "Tianze Sun", "Ran Chen", "Wanbo Zhang", "Guang Zhang", "Siguang Li"], "categories": ["cs.MA", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-21", "url": "https://arxiv.org/abs/2601.15047", "pdf_url": "https://arxiv.org/pdf/2601.15047v1", "arxiv_id": "2601.15047", "doi": "10.48550/arXiv.2601.15047", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4515}
{"id": "805f8eacfe5fa1dcedc7251682e3df0cb5748634e5382b75c6bb9e5116617269", "sources": ["arxiv", "semantic_scholar"], "title": "TransportAgents: a multi-agents LLM framework for traffic accident severity prediction", "abstract": "Accurate prediction of traffic crash severity is critical for improving emergency response and public safety planning. Although recent large language models (LLMs) exhibit strong reasoning capabilities, their single-agent architectures often struggle with heterogeneous, domain-specific crash data and tend to generate biased or unstable predictions. To address these limitations, this paper proposes TransportAgents, a hybrid multi-agent framework that integrates category-specific LLM reasoning with a multilayer perceptron (MLP) integration module. Each specialized agent focuses on a particular subset of traffic information, such as demographics, environmental context, or incident details, to produce intermediate severity assessments that are subsequently fused into a unified prediction. Extensive experiments on two complementary U.S. datasets, the Consumer Product Safety Risk Management System (CPSRMS) and the National Electronic Injury Surveillance System (NEISS), demonstrate that TransportAgents consistently outperforms both traditional machine learning and advanced LLM-based baselines. Across three representative backbones, including closed-source models such as GPT-3.5 and GPT-4o, as well as open-source models such as LLaMA-3.3, the framework exhibits strong robustness, scalability, and cross-dataset generalizability. A supplementary distributional analysis further shows that TransportAgents produces more balanced and well-calibrated severity predictions than standard single-agent LLM approaches, highlighting its interpretability and reliability for safety-critical decision support applications.", "authors": ["Zhichao Yang", "Jiashu He", "Jinxuan Fan", "Cirillo Cinzia"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-21", "url": "https://arxiv.org/abs/2601.15519", "pdf_url": "https://arxiv.org/pdf/2601.15519v2", "arxiv_id": "2601.15519", "doi": "10.48550/arXiv.2601.15519", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6977}
{"id": "632999a170066b54ca5eb5cab498ff5e1851feee9d90a27f344c8f422d57938a", "sources": ["arxiv", "semantic_scholar"], "title": "Hidden in Plain Text: Measuring LLM Deception Quality Against Human Baselines Using Social Deduction Games", "abstract": "Large Language Model (LLM) agents are increasingly used in many applications, raising concerns about their safety. While previous work has shown that LLMs can deceive in controlled tasks, less is known about their ability to deceive using natural language in social contexts. In this paper, we study deception in the Social Deduction Game (SDG) Mafia, where success is dependent on deceiving others through conversation. Unlike previous SDG studies, we use an asynchronous multi-agent framework which better simulates realistic social contexts. We simulate 35 Mafia games with GPT-4o LLM agents. We then create a Mafia Detector using GPT-4-Turbo to analyze game transcripts without player role information to predict the mafia players. We use prediction accuracy as a surrogate marker for deception quality. We compare this prediction accuracy to that of 28 human games and a random baseline. Results show that the Mafia Detector's mafia prediction accuracy is lower on LLM games than on human games. The result is consistent regardless of the game days and the number of mafias detected. This indicates that LLMs blend in better and thus deceive more effectively. We also release a dataset of LLM Mafia transcripts to support future research. Our findings underscore both the sophistication and risks of LLM deception in social contexts.", "authors": ["Christopher Kao", "Vanshika Vats", "James Davis"], "categories": ["cs.AI", "cs.CL", "cs.CY", "cs.HC", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-20", "url": "https://arxiv.org/abs/2601.13709", "pdf_url": "https://arxiv.org/pdf/2601.13709v1", "arxiv_id": "2601.13709", "doi": "10.1109/ICA67499.2025.00033", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/cocochief4/llm-mafia", "venue": "International Conference on Agents", "quality_score": 0.6959}
{"id": "ac069286d48858f7ab20af43a3e9974cc6dc20eba89621eda358ac9f8443fd2c", "sources": ["arxiv", "semantic_scholar"], "title": "MASCOT: Towards Multi-Agent Socio-Collaborative Companion Systems", "abstract": "Multi-agent systems (MAS) are emerging as promising socio-collaborative companions for emotional and cognitive support. However, existing systems frequently suffer from persona collapse, where agents revert to generic, homogenized assistant behaviors, and social sycophancy, where agents produce redundant, non-constructive dialogue. We propose MASCOT, a multi-agent framework for multi-perspective socio-collaborative companions. MASCOT introduces a novel bi-level optimization strategy to harmonize individual and collective behaviors: 1) Persona-Aware Behavioral Alignment, an RLAIF-driven pipeline that fine-tunes individual agents for agent-specific identities; and 2) Collaborative Dialogue Optimization, a group-level adaptation process that promotes complementary, diverse, and productive discourse. We evaluate MASCOT using human-grounded contexts drawn across both in-domain and out-of-domain (OOD) settings against state-of-the-art baselines. MASCOT improves persona consistency by up to +14.1 and social contribution by up to +10.6. A broad evaluation suite, including human evaluation, multiple LLM judges, three-way comparisons, and automatic metrics, further shows that MASCOT produces more role-consistent and less redundant multi-agent dialogue.", "authors": ["Yiyang Wang", "Yiqiao Jin", "Alex Cabral", "Josiah Hester"], "categories": ["cs.CL", "cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-20", "url": "https://arxiv.org/abs/2601.14230", "pdf_url": "https://arxiv.org/pdf/2601.14230v2", "arxiv_id": "2601.14230", "doi": "10.48550/arXiv.2601.14230", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4503}
{"id": "613b4b346831096f85ae7edb4493d0a99aa56a8648eb6750c798cd6a9f280513", "sources": ["arxiv", "semantic_scholar"], "title": "Rethinking the Value of Multi-Agent Workflow: A Strong Single Agent Baseline", "abstract": "Recent advances in LLM-based multi-agent systems (MAS) show that workflows composed of multiple LLM agents with distinct roles, tools, and communication patterns can outperform single-LLM baselines on complex tasks. However, most frameworks are homogeneous, where all agents share the same base LLM and differ only in prompts, tools, and positions in the workflow. This raises the question of whether such workflows can be simulated by a single agent through multi-turn conversations. We investigate this across seven benchmarks spanning coding, mathematics, general question answering, domain-specific reasoning, and real-world planning and tool use. Our results show that a single agent can reach the performance of homogeneous workflows with an efficiency advantage from KV cache reuse, and can even match the performance of an automatically optimized heterogeneous workflow. Building on this finding, we propose \\textbf{OneFlow}, an algorithm that automatically tailors workflows for single-agent execution, reducing inference costs compared to existing automatic multi-agent design frameworks without trading off accuracy. These results position the single-LLM implementation of multi-agent workflows as a strong baseline for MAS research. We also note that single-LLM methods cannot capture heterogeneous workflows due to the lack of KV cache sharing across different LLMs, highlighting future opportunities in developing \\textit{truly} heterogeneous multi-agent systems.", "authors": ["Jiawei Xu", "Arief Koesdwiady", "Sisong Bei", "Yan Han", "Baixiang Huang", "Dakuo Wang", "Yutong Chen", "Zheshen Wang", "Peihao Wang", "Pan Li", "Ying Ding"], "categories": ["cs.MA", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-18", "url": "https://arxiv.org/abs/2601.12307", "pdf_url": "https://arxiv.org/pdf/2601.12307v1", "arxiv_id": "2601.12307", "doi": "10.48550/arXiv.2601.12307", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.448}
{"id": "f8ae1f50f7f0b213ed2a766abab70e7aee0a23fcb24663b04939d2b97e40cebc", "sources": ["arxiv", "semantic_scholar"], "title": "Holos: A Web-Scale LLM-Based Multi-Agent System for the Agentic Web", "abstract": "As large language models (LLM)-driven agents transition from isolated task solvers to persistent digital entities, the emergence of the Agentic Web, an ecosystem where heterogeneous agents autonomously interact and co-evolve, marks a pivotal shift toward Artificial General Intelligence (AGI). However, LLM-based multi-agent systems (LaMAS) are hindered by open-world issues such as scaling friction, coordination breakdown, and value dissipation. To address these challenges, we introduce Holos, a web-scale LaMAS architected for long-term ecological persistence. Holos adopts a five-layer architecture, with core modules primarily featuring the Nuwa engine for high-efficiency agent generation and hosting, a market-driven Orchestrator for resilient coordination, and an endogenous value cycle to achieve incentive compatibility. By bridging the gap between micro-level collaboration and macro-scale emergence, Holos hopes to lay the foundation for the next generation of the self-organizing and continuously evolving Agentic Web. We have publicly released Holos (accessible at https://holosai.io), providing a resource for the community and a testbed for future research in large-scale agentic ecosystems.", "authors": ["Xiaohang Nie", "Zihan Guo", "Zicai Cui", "Jiachi Yang", "Zeyi Chen", "Leheyi De", "Yu Zhang", "Junwei Liao", "Bo Huang", "Yingxuan Yang", "Zhi Han", "Zimian Peng", "Linyao Chen", "Wenzheng Tom Tang", "Zongkai Liu", "Tao Zhou", "Botao Amber Hu", "Shuyang Tang", "Jianghao Lin", "Weiwen Liu", "Muning Wen", "Yuanjian Zhou", "Weinan Zhang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-18", "url": "https://arxiv.org/abs/2604.02334", "pdf_url": "https://arxiv.org/pdf/2604.02334v1", "arxiv_id": "2604.02334", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2851}
{"id": "a7b9289efdc23d025e91738fb48350c2b5b9948e697abfd6f5eb59f739deff97", "sources": ["arxiv", "semantic_scholar"], "title": "ToolPRMBench: Evaluating and Advancing Process Reward Models for Tool-using Agents", "abstract": "Reward-guided search methods have demonstrated strong potential in enhancing tool-using agents by effectively guiding sampling and exploration over complex action spaces. As a core design, those search methods utilize process reward models (PRMs) to provide step-level rewards, enabling more fine-grained monitoring. However, there is a lack of systematic and reliable evaluation benchmarks for PRMs in tool-using settings. In this paper, we introduce ToolPRMBench, a large-scale benchmark specifically designed to evaluate PRMs for tool-using agents. ToolPRMBench is built on top of several representative tool-using benchmarks and converts agent trajectories into step-level test cases. Each case contains the interaction history, a correct action, a plausible but incorrect alternative, and relevant tool metadata. We respectively utilize offline sampling to isolate local single-step errors and online sampling to capture realistic multi-step failures from full agent rollouts. A multi-LLM verification pipeline is proposed to reduce label noise and ensure data quality. We conduct extensive experiments across large language models, general PRMs, and tool-specialized PRMs on ToolPRMBench. The results reveal clear differences in PRM effectiveness and highlight the potential of specialized PRMs for tool-using. Code and data will be released at https://github.com/David-Li0406/ToolPRMBench.", "authors": ["Dawei Li", "Yuguang Yao", "Zhen Tan", "Huan Liu", "Ruocheng Guo"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-18", "url": "https://arxiv.org/abs/2601.12294", "pdf_url": "https://arxiv.org/pdf/2601.12294v1", "arxiv_id": "2601.12294", "doi": "10.48550/arXiv.2601.12294", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/David-Li0406/ToolPRMBench", "venue": "arXiv.org", "quality_score": 0.6924}
{"id": "645a26804aa903e5148e3a12b3d5072254bdd01d57c48680060edbfe84aec948", "sources": ["arxiv", "semantic_scholar"], "title": "AEMA: Verifiable Evaluation Framework for Trustworthy and Controlled Agentic LLM Systems", "abstract": "Evaluating large language model (LLM)-based multi-agent systems remains a critical challenge, as these systems must exhibit reliable coordination, transparent decision-making, and verifiable performance across evolving tasks. Existing evaluation approaches often limit themselves to single-response scoring or narrow benchmarks, which lack stability, extensibility, and automation when deployed in enterprise settings at multi-agent scale. We present AEMA (Adaptive Evaluation Multi-Agent), a process-aware and auditable framework that plans, executes, and aggregates multi-step evaluations across heterogeneous agentic workflows under human oversight. Compared to a single LLM-as-a-Judge, AEMA achieves greater stability, human alignment, and traceable records that support accountable automation. Our results on enterprise-style agent workflows simulated using realistic business scenarios demonstrate that AEMA provides a transparent and reproducible pathway toward responsible evaluation of LLM-based multi-agent systems. Keywords Agentic AI, Multi-Agent Systems, Trustworthy AI, Verifiable Evaluation, Human Oversight", "authors": ["YenTing Lee", "Keerthi Koneru", "Zahra Moslemi", "Sheethal Kumar", "Ramesh Radhakrishnan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-17", "url": "https://arxiv.org/abs/2601.11903", "pdf_url": "https://arxiv.org/pdf/2601.11903v1", "arxiv_id": "2601.11903", "doi": "10.48550/arXiv.2601.11903", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4469}
{"id": "e2deb3a924a3eccf5ba7f824d64889e89b891bf8e617b2c22cfe882d2590ba21", "sources": ["arxiv", "semantic_scholar"], "title": "Mitigating Cultural Bias in LLMs via Multi-Agent Cultural Debate", "abstract": "Large language models (LLMs) exhibit systematic Western-centric bias, yet whether prompting in non-Western languages (e.g., Chinese) can mitigate this remains understudied. Answering this question requires rigorous evaluation and effective mitigation, but existing approaches fall short on both fronts: evaluation methods force outputs into predefined cultural categories without a neutral option, while mitigation relies on expensive multi-cultural corpora or agent frameworks that use functional roles (e.g., Planner--Critique) lacking explicit cultural representation. To address these gaps, we introduce CEBiasBench, a Chinese--English bilingual benchmark, and Multi-Agent Vote (MAV), which enables explicit ``no bias'' judgments. Using this framework, we find that Chinese prompting merely shifts bias toward East Asian perspectives rather than eliminating it. To mitigate such persistent bias, we propose Multi-Agent Cultural Debate (MACD), a training-free framework that assigns agents distinct cultural personas and orchestrates deliberation via a \"Seeking Common Ground while Reserving Differences\" strategy. Experiments demonstrate that MACD achieves 57.6% average No Bias Rate evaluated by LLM-as-judge and 86.0% evaluated by MAV (vs. 47.6% and 69.0% baseline using GPT-4o as backbone) on CEBiasBench and generalizes to the Arabic CAMeL benchmark, confirming that explicit cultural representation in agent frameworks is essential for cross-cultural fairness.", "authors": ["Qian Tan", "Lei Jiang", "Yuting Zeng", "Shuoyang Ding", "Xiaohua Xu"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-17", "url": "https://arxiv.org/abs/2601.12091", "pdf_url": "https://arxiv.org/pdf/2601.12091v1", "arxiv_id": "2601.12091", "doi": "10.48550/arXiv.2601.12091", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4469}
{"id": "25abc17d447fb20541cf80cd610c2e22fed31587c3498290304bb5a8b05438cb", "sources": ["arxiv", "semantic_scholar"], "title": "Replayable Financial Agents: A Determinism-Faithfulness Assurance Harness for Tool-Using LLM Agents", "abstract": "LLM agents struggle with regulatory audit replay: when asked to reproduce a flagged transaction decision with identical inputs, many deployments fail to return consistent results. We introduce the Determinism-Faithfulness Assurance Harness (DFAH), a framework for measuring trajectory determinism, decision determinism, and evidence-conditioned faithfulness in tool-using agents deployed in financial services. Across 4,700+ agentic runs (7 models, 4 providers, 3 financial benchmarks with 50 cases each at T=0.0), we find that decision determinism and task accuracy are not detectably correlated (r = -0.11, 95% CI [-0.49, 0.31], p = 0.63, n = 21 configurations): models can be deterministic without being accurate, and accurate without being deterministic. Because neither metric predicts the other in our sample, both must be measured independently, which is precisely what DFAH provides. Small models (7-20B) achieve near-perfect determinism through rigid pattern matching at the cost of accuracy (20-42%), while frontier models show moderate determinism (50-96%) with variable accuracy. No model achieves both perfect determinism and high accuracy, supporting DFAH's multi-dimensional measurement approach. We provide three financial benchmarks (compliance triage, portfolio constraints, and DataOps exceptions; 50 cases each) together with an open-source stress-test harness. Across these benchmarks and DFAH evaluation settings, Tier 1 models with schema-first architectures achieved determinism levels consistent with audit replay requirements.", "authors": ["Raffi Khatchadourian"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-17", "url": "https://arxiv.org/abs/2601.15322", "pdf_url": "https://arxiv.org/pdf/2601.15322v2", "arxiv_id": "2601.15322", "doi": "10.48550/arXiv.2601.15322", "citation_count": 2, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/ibm-client-engineering/output-drift-financial-llms", "venue": "arXiv.org", "quality_score": 0.6906}
{"id": "24f62c5d14f894bc52a4353ef355f654d2c8dfa249042399f54cd3e1cb311246", "sources": ["arxiv", "semantic_scholar"], "title": "Taming Various Privilege Escalation in LLM-Based Agent Systems: A Mandatory Access Control Framework", "abstract": "Large Language Model (LLM)-based agent systems are increasingly deployed for complex real-world tasks but remain vulnerable to natural language-based attacks that exploit over-privileged tool use. This paper aims to understand and mitigate such attacks through the lens of privilege escalation, defined as agent actions exceeding the least privilege required for a user's intended task. Based on a formal model of LLM agent systems, we identify novel privilege escalation scenarios, particularly in multi-agent systems, including a variant akin to the classic confused deputy problem. To defend against both known and newly demonstrated privilege escalation, we propose SEAgent, a mandatory access control (MAC) framework built upon attribute-based access control (ABAC). SEAgent monitors agent-tool interactions via an information flow graph and enforces customizable security policies based on entity attributes. Our evaluations show that SEAgent effectively blocks various privilege escalation while maintaining a low false positive rate and negligible system overhead. This demonstrates its robustness and adaptability in securing LLM-based agent systems.", "authors": ["Zimo Ji", "Daoyuan Wu", "Wenyuan Jiang", "Pingchuan Ma", "Zongjie Li", "Yudong Gao", "Shuai Wang", "Yingjiu Li"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-17", "url": "https://arxiv.org/abs/2601.11893", "pdf_url": "https://arxiv.org/pdf/2601.11893v1", "arxiv_id": "2601.11893", "doi": "10.48550/arXiv.2601.11893", "citation_count": 21, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4469}
{"id": "5c03d16e5c4e91bfea8f1e5200f9cdc347968b184b30dfd76f92d6c082a8506e", "sources": ["arxiv", "semantic_scholar"], "title": "Many Hands Make Light Work: An LLM-based Multi-Agent System for Detecting Malicious PyPI Packages", "abstract": "Malicious code in open-source repositories such as PyPI poses a growing threat to software supply chains. Traditional rule-based tools often overlook the semantic patterns in source code that are crucial for identifying adversarial components. Large language models (LLMs) show promise for software analysis, yet their use in interpretable and modular security pipelines remains limited. This paper presents LAMPS, a multi-agent system that employs collaborative LLMs to detect malicious PyPI packages. The system consists of four role-specific agents for package retrieval, file extraction, classification, and verdict aggregation, coordinated through the CrewAI framework. A prototype combines a fine-tuned CodeBERT model for classification with LLaMA-3 agents for contextual reasoning. LAMPS has been evaluated on two complementary datasets: D1, a balanced collection of 6,000 setup.py files, and D2, a realistic multi-file dataset with 1,296 files and natural class imbalance. On D1, LAMPS achieves 97.7% accuracy, surpassing MPHunter--one of the state-of-the-art approaches. On D2, it reaches 99.5% accuracy and 99.5% balanced accuracy, outperforming RAG-based approaches and fine-tuned single-agent baselines. McNemar's test confirmed these improvements as highly significant. The results demonstrate the feasibility of distributed LLM reasoning for malicious code detection and highlight the benefits of modular multi-agent designs in software supply chain security.", "authors": ["Muhammad Umar Zeshan", "Motunrayo Ibiyo", "Claudio Di Sipio", "Phuong T. Nguyen", "Davide Di Ruscio"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-17", "url": "https://arxiv.org/abs/2601.12148", "pdf_url": "https://arxiv.org/pdf/2601.12148v3", "arxiv_id": "2601.12148", "doi": "10.1016/j.jss.2026.112792", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Journal of Systems and Software", "quality_score": 0.6906}
{"id": "430fad2e3879263a42ee31ea456246cf4af75b8d1e400f858f805388a587c46e", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond Max Tokens: Stealthy Resource Amplification via Tool Calling Chains in LLM Agents", "abstract": "The agent--tool interaction loop is a critical attack surface for modern Large Language Model (LLM) agents. Existing denial-of-service (DoS) attacks typically function at the user-prompt or retrieval-augmented generation (RAG) context layer and are inherently single-turn in nature. This limitation restricts cost amplification and diminishes stealth in goal-oriented workflows. To address these issues, we proposed a stealthy, multi-turn economic DoS attack at the tool layer under the Model Context Protocol (MCP). By simply editing text-visible fields and implementing a template-driven return policy, our malicious server preserves function signatures and the terminal benign payload while steering agents into prolonged, verbose tool-calling chains. We optimize these text-only edits with Monte Carlo Tree Search (MCTS) to maximize cost under a task-success constraint. Across six LLMs on ToolBench and BFCL benchmarks, our attack yields trajectories over 60K tokens, increases per-query cost by up to 658 times, raises energy by 100 to 560 times, and pushes GPU key-value (KV) cache occupancy to 35--74%. Standard prompt filters and output trajectory monitors seldom detect these attacks, highlighting the need for defenses that safeguard agentic processes rather than focusing solely on final outcomes. We will release the code soon.", "authors": ["Kaiyu Zhou", "Yongsen Zheng", "Yicheng He", "Meng Xue", "Xueluan Gong", "Yuji Wang", "Xuanye Zhang", "Kwok-Yan Lam"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-16", "url": "https://arxiv.org/abs/2601.10955", "pdf_url": "https://arxiv.org/pdf/2601.10955v2", "arxiv_id": "2601.10955", "doi": "10.48550/arXiv.2601.10955", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4457}
{"id": "1ce7c6c0b27fb53aa49273e916b95e9141dd5ce1883cfa7e8c25ef132b953c11", "sources": ["arxiv", "semantic_scholar"], "title": "ToolSafe: Enhancing Tool Invocation Safety of LLM-based agents via Proactive Step-level Guardrail and Feedback", "abstract": "While LLM-based agents can interact with environments via invoking external tools, their expanded capabilities also amplify security risks. Monitoring step-level tool invocation behaviors in real time and proactively intervening before unsafe execution is critical for agent deployment, yet remains under-explored. In this work, we first construct TS-Bench, a novel benchmark for step-level tool invocation safety detection in LLM agents. We then develop a guardrail model, TS-Guard, using multi-task reinforcement learning. The model proactively detects unsafe tool invocation actions before execution by reasoning over the interaction history. It assesses request harmfulness and action-attack correlations, producing interpretable and generalizable safety judgments and feedback. Furthermore, we introduce TS-Flow, a guardrail-feedback-driven reasoning framework for LLM agents, which reduces harmful tool invocations of ReAct-style agents by 65 percent on average and improves benign task completion by approximately 10 percent under prompt injection attacks.", "authors": ["Yutao Mou", "Zhangchi Xue", "Lijun Li", "Peiyang Liu", "Shikun Zhang", "Wei Ye", "Jing Shao"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-15", "url": "https://arxiv.org/abs/2601.10156", "pdf_url": "https://arxiv.org/pdf/2601.10156v1", "arxiv_id": "2601.10156", "doi": "10.48550/arXiv.2601.10156", "citation_count": 22, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/MurrayTom/ToolSafe", "venue": "arXiv.org", "quality_score": 0.6871}
{"id": "9a8cf633abae33039304e16105882094dc856c1c8789fb856eab702d5367bf02", "sources": ["arxiv", "semantic_scholar"], "title": "Too Helpful to Be Safe: User-Mediated Attacks on Planning and Web-Use Agents", "abstract": "Large Language Models (LLMs) have enabled agents to move beyond conversation toward end-to-end task execution and become more helpful. However, this helpfulness introduces new security risks stem less from direct interface abuse than from acting on user-provided content. Existing studies on agent security largely focus on model-internal vulnerabilities or adversarial access to agent interfaces, overlooking attacks that exploit users as unintended conduits. In this paper, we study user-mediated attacks, where benign users are tricked into relaying untrusted or attacker-controlled content to agents, and analyze how commercial LLM agents respond under such conditions. We conduct a systematic evaluation of 12 commercial agents in a sandboxed environment, covering 6 trip-planning agents and 6 web-use agents, and compare agent behavior across scenarios with no, soft, and hard user-requested safety checks. Our results show that agents are too helpful to be safe by default. Without explicit safety requests, trip-planning agents bypass safety constraints in over 92% of cases, converting unverified content into confident booking guidance. Web-use agents exhibit near-deterministic execution of risky actions, with 9 out of 17 supported tests reaching a 100% bypass rate. Even when users express soft or hard safety intent, constraint bypass remains substantial, reaching up to 54.7% and 7% for trip-planning agents, respectively. These findings reveal that the primary issue is not a lack of safety capability, but its prioritization. Agents invoke safety checks only conditionally when explicitly prompted, and otherwise default to goal-driven execution. Moreover, agents lack clear task boundaries and stopping rules, frequently over-executing workflows in ways that lead to unnecessary data disclosure and real-world harm.", "authors": ["Fengchao Chen", "Tingmin Wu", "Van Nguyen", "Carsten Rudolph"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-14", "url": "https://arxiv.org/abs/2601.10758", "pdf_url": "https://arxiv.org/pdf/2601.10758v1", "arxiv_id": "2601.10758", "doi": "10.48550/arXiv.2601.10758", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4434}
{"id": "96ec15710a6b4c3c6ab36ed9ef5eaf8485f0a56a1c827a73fa34265ec314c4f7", "sources": ["arxiv", "semantic_scholar"], "title": "MAXS: Meta-Adaptive Exploration with LLM Agents", "abstract": "Large Language Model (LLM) Agents exhibit inherent reasoning abilities through the collaboration of multiple tools. However, during agent inference, existing methods often suffer from (i) locally myopic generation, due to the absence of lookahead, and (ii) trajectory instability, where minor early errors can escalate into divergent reasoning paths. These issues make it difficult to balance global effectiveness and computational efficiency. To address these two issues, we propose meta-adaptive exploration with LLM agents https://github.com/exoskeletonzj/MAXS, a meta-adaptive reasoning framework based on LLM Agents that flexibly integrates tool execution and reasoning planning. MAXS employs a lookahead strategy to extend reasoning paths a few steps ahead, estimating the advantage value of tool usage, and combines step consistency variance and inter-step trend slopes to jointly select stable, consistent, and high-value reasoning steps. Additionally, we introduce a trajectory convergence mechanism that controls computational cost by halting further rollouts once path consistency is achieved, enabling a balance between resource efficiency and global effectiveness in multi-tool reasoning. We conduct extensive empirical studies across three base models (MiMo-VL-7B, Qwen2.5-VL-7B, Qwen2.5-VL-32B) and five datasets, demonstrating that MAXS consistently outperforms existing methods in both performance and inference efficiency. Further analysis confirms the effectiveness of our lookahead strategy and tool usage.", "authors": ["Jian Zhang", "Zhiyuan Wang", "Zhangqi Wang", "Yu He", "Haoran Luo", "li yuan", "Lingling Zhang", "Rui Mao", "Qika Lin", "Jun Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-14", "url": "https://arxiv.org/abs/2601.09259", "pdf_url": "https://arxiv.org/pdf/2601.09259v1", "arxiv_id": "2601.09259", "doi": "10.48550/arXiv.2601.09259", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/exoskeletonzj/MAXS", "venue": "arXiv.org", "quality_score": 0.6853}
{"id": "dcf383247c32898fe72fcaec2fd92f55cfda92d0182181bb6197927c43bbb1f4", "sources": ["arxiv", "semantic_scholar"], "title": "MACRO-LLM: LLM-Empowered Multi-Agent Collaborative Reasoning under Spatiotemporal Partial Observability", "abstract": "Large Language Model (LLM) agents deployed in complex real-world scenarios increasingly operate as spatially distributed entities. However, this physical dispersion constrains agents to limited local perception and finite temporal horizons. We characterize this bottleneck as spatiotemporal partial observability, where spatial and temporal limitations are fundamentally coupled: resolving spatial conflicts requires temporal reasoning about neighbors' future actions, while temporal planning requires spatial context beyond local perception. To bridge this gap, we introduce MACRO-LLM, LLM-empowered multi-agent collaborative reasoning under spatiotemporal partial observability. The architecture interleaves spatial and temporal reasoning within each decision cycle via three interdependent modules: (1) the CoProposer mitigates temporal uncertainty by verifying candidate actions via predictive rollouts; (2) the Negotiator overcomes spatial myopia by resolving conflicts through mean-field statistical aggregation, grounded in the CoProposer's rollout rewards; and (3) the Introspector closes the reasoning loop by analyzing environmental drift and attributing performance changes to refine strategies. Extensive evaluations on two complex long-horizon tasks, cooperative platoon planning and pandemic control, demonstrate that our framework enables robust coordination under spatiotemporal partial observability.", "authors": ["Handi Chen", "Running Zhao", "Xiuzhe Wu", "Edith C. H. Ngai"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-14", "url": "https://arxiv.org/abs/2601.09295", "pdf_url": "https://arxiv.org/pdf/2601.09295v2", "arxiv_id": "2601.09295", "doi": "10.48550/arXiv.2601.09295", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4434}
{"id": "180b39da8decce72b5fafd90b13b9320fa54a93804e7cb68c272eb7486b3f08e", "sources": ["arxiv", "semantic_scholar"], "title": "SC-MAS: Constructing Cost-Efficient Multi-Agent Systems with Edge-Level Heterogeneous Collaboration", "abstract": "Large Language Model (LLM)-based Multi-Agent Systems (MAS) enhance complex problem solving through multi-agent collaboration, but often incur substantially higher costs than single-agent systems. Recent MAS routing methods aim to balance performance and overhead by dynamically selecting agent roles and language models. However, these approaches typically rely on a homogeneous collaboration mode, where all agents follow the same interaction pattern, limiting collaboration flexibility across different roles. Motivated by Social Capital Theory, which emphasizes that different roles benefit from distinct forms of collaboration, we propose SC-MAS, a framework for constructing heterogeneous and cost-efficient multi-agent systems. SC-MAS models MAS as directed graphs, where edges explicitly represent pairwise collaboration strategies, allowing different agent pairs to interact through tailored communication patterns. Given an input query, a unified controller progressively constructs an executable MAS by selecting task-relevant agent roles, assigning edge-level collaboration strategies, and allocating appropriate LLM backbones to individual agents. Experiments on multiple benchmarks demonstrate the effectiveness of SC-MAS. In particular, SC-MAS improves accuracy by 3.35% on MMLU while reducing inference cost by 15.38%, and achieves a 3.53% accuracy gain with a 12.13% cost reduction on MBPP. These results validate the feasibility of SC-MAS and highlight the effectiveness of heterogeneous collaboration in multi-agent systems.", "authors": ["Di Zhao", "Longhui Ma", "Siwei Wang", "Miao Wang", "Yi Kong"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-14", "url": "https://arxiv.org/abs/2601.09434", "pdf_url": "https://arxiv.org/pdf/2601.09434v1", "arxiv_id": "2601.09434", "doi": "10.48550/arXiv.2601.09434", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4434}
{"id": "d8659608d1e49b1f2b500b62645d51800dca994f6be1450d169a25e16bf677ea", "sources": ["arxiv", "semantic_scholar"], "title": "Project Synapse: A Hierarchical Multi-Agent Framework with Hybrid Memory for Autonomous Resolution of Last-Mile Delivery Disruptions", "abstract": "This paper introduces Project Synapse, a novel agentic framework designed for the autonomous resolution of last-mile delivery disruptions. Synapse employs a hierarchical multi-agent architecture in which a central Resolution Supervisor agent performs strategic task decomposition and delegates subtasks to specialized worker agents responsible for tactical execution. The system is orchestrated using LangGraph to manage complex and cyclical workflows. To validate the framework, a benchmark dataset of 30 complex disruption scenarios was curated from a qualitative analysis of over 6,000 real-world user reviews. System performance is evaluated using an LLM-as-a-Judge protocol with explicit bias mitigation.", "authors": ["Arin Gopalan Yadav", "Varad Dherange", "Kumar Shivam"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-13", "url": "https://arxiv.org/abs/2601.08156", "pdf_url": "https://arxiv.org/pdf/2601.08156v1", "arxiv_id": "2601.08156", "doi": "10.48550/arXiv.2601.08156", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4423}
{"id": "02a2f5ce0061d55eb7ec0d11fa6949bc9ff28d5fc183ff55c805b8588cecbda1", "sources": ["arxiv", "semantic_scholar"], "title": "Inferring Latent Intentions: Attributional Natural Language Inference in LLM Agents", "abstract": "Attributional inference, the ability to predict latent intentions behind observed actions, is a critical yet underexplored capability for large language models (LLMs) operating in multi-agent environments. Traditional natural language inference (NLI), in fact, fails to capture the nuanced, intention-driven reasoning essential for complex interactive systems. To address this gap, we introduce Attributional NLI (Att-NLI), a framework that extends NLI with principles from social psychology to assess an agent's capacity for abductive intentional inference (generating hypotheses about latent intentions), and subsequent deductive verification (drawing valid logical conclusions). We instantiate Att-NLI via a textual game, Undercover-V, experimenting with three types of LLM agents with varying reasoning capabilities and access to external tools: a standard NLI agent using only deductive inference, an Att-NLI agent employing abductive-deductive inference, and a neuro-symbolic Att-NLI agent performing abductive-deductive inference with external theorem provers. Extensive experiments demonstrate a clear hierarchy of attributional inference capabilities, with neuro-symbolic agents consistently outperforming others, achieving an average win rate of 17.08%. Our results underscore the role that Att-NLI can play in developing agents with sophisticated reasoning capabilities, highlighting, at the same time, the potential impact of neuro-symbolic AI in building rational LLM agents acting in multi-agent environments.", "authors": ["Xin Quan", "Jiafeng Xiong", "Marco Valentino", "André Freitas"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-13", "url": "https://arxiv.org/abs/2601.08742", "pdf_url": "https://arxiv.org/pdf/2601.08742v1", "arxiv_id": "2601.08742", "doi": "10.48550/arXiv.2601.08742", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4423}
{"id": "e7b2f86cd55f401f5f00084fef3a4f921280476fad1ac9e7c2272255d7ddfbf6", "sources": ["arxiv", "semantic_scholar"], "title": "The End of Reward Engineering: How LLMs Are Redefining Multi-Agent Coordination", "abstract": "Reward engineering, the manual specification of reward functions to induce desired agent behavior, remains a fundamental challenge in multi-agent reinforcement learning. This difficulty is amplified by credit assignment ambiguity, environmental non-stationarity, and the combinatorial growth of interaction complexity. We argue that recent advances in large language models (LLMs) point toward a shift from hand-crafted numerical rewards to language-based objective specifications. Prior work has shown that LLMs can synthesize reward functions directly from natural language descriptions (e.g., EUREKA) and adapt reward formulations online with minimal human intervention (e.g., CARD). In parallel, the emerging paradigm of Reinforcement Learning from Verifiable Rewards (RLVR) provides empirical evidence that language-mediated supervision can serve as a viable alternative to traditional reward engineering. We conceptualize this transition along three dimensions: semantic reward specification, dynamic reward adaptation, and improved alignment with human intent, while noting open challenges related to computational overhead, robustness to hallucination, and scalability to large multi-agent systems. We conclude by outlining a research direction in which coordination arises from shared semantic representations rather than explicitly engineered numerical signals.", "authors": ["Haoran Su", "Yandong Sun", "Congjia Yu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-13", "url": "https://arxiv.org/abs/2601.08237", "pdf_url": "https://arxiv.org/pdf/2601.08237v1", "arxiv_id": "2601.08237", "doi": "10.48550/arXiv.2601.08237", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4423}
{"id": "0f2b022fd563bcb9eb39d7c6ef95b68a073fc26064b4b25bd3f643aaa125db7e", "sources": ["arxiv", "semantic_scholar"], "title": "When KV Cache Reuse Fails in Multi-Agent Systems: Cross-Candidate Interaction is Crucial for LLM Judges", "abstract": "Multi-agent LLM systems routinely generate multiple candidate responses that are aggregated by an LLM judge. To reduce the dominant prefill cost in such pipelines, recent work advocates KV cache reuse across partially shared contexts and reports substantial speedups for generation agents. In this work, we show that these efficiency gains do not transfer uniformly to judge-centric inference. Across GSM8K, MMLU, and HumanEval, we find that reuse strategies that are effective for execution agents can severely perturb judge behavior: end-task accuracy may appear stable, yet the judge's selection becomes highly inconsistent with dense prefill. We quantify this risk using Judge Consistency Rate (JCR) and provide diagnostics showing that reuse systematically weakens cross-candidate attention, especially for later candidate blocks. Our ablation further demonstrates that explicit cross-candidate interaction is crucial for preserving dense-prefill decisions. Overall, our results identify a previously overlooked failure mode of KV cache reuse and highlight judge-centric inference as a distinct regime that demands dedicated, risk-aware system design.", "authors": ["Sichu Liang", "Zhenglin Wang", "Jiajia Chu", "Pengfei Xia", "Hui Zang", "Deyu Zhou"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-13", "url": "https://arxiv.org/abs/2601.08343", "pdf_url": "https://arxiv.org/pdf/2601.08343v1", "arxiv_id": "2601.08343", "doi": "10.48550/arXiv.2601.08343", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4423}
{"id": "6fdea1f9a6fbe08f19187bd09eb4911c80aa3194af31e245d9577c422917aaf1", "sources": ["arxiv", "semantic_scholar"], "title": "Modeling LLM Agent Reviewer Dynamics in Elo-Ranked Review System", "abstract": "In this work, we explore the Large Language Model (LLM) agent reviewer dynamics in an Elo-ranked review system using real-world conference paper submissions. Multiple LLM agent reviewers with different personas are engage in multi round review interactions moderated by an Area Chair. We compare a baseline setting with conditions that incorporate Elo ratings and reviewer memory. Our simulation results showcase several interesting findings, including how incorporating Elo improves Area Chair decision accuracy, as well as reviewers' adaptive review strategy that exploits our Elo system without improving review effort. Our code is available at https://github.com/hsiangwei0903/EloReview.", "authors": ["Hsiang-Wei Huang", "Junbin Lu", "Kuang-Ming Chen", "Jenq-Neng Hwang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-13", "url": "https://arxiv.org/abs/2601.08829", "pdf_url": "https://arxiv.org/pdf/2601.08829v1", "arxiv_id": "2601.08829", "doi": "10.48550/arXiv.2601.08829", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/hsiangwei0903/EloReview", "venue": "arXiv.org", "quality_score": 0.6835}
{"id": "548a29b5a35f91e04498f517ec22962b1d3c57fea0ff5a37bc3c25085e8a99a6", "sources": ["arxiv", "semantic_scholar"], "title": "FOCAL: A Novel Benchmarking Technique for Multi-modal Agents", "abstract": "With the recent advancements in reasoning capabilities, tool calling using MCP servers and Audio Language Models (ALMs), development and integration of multi-modal agents (with voice and text support) has come to the industry forefront. Cascading pipelines for voice agents still play a central role in the industry owing to their superior reasoning capabilities facilitated by LLMs. Although, cascading pipelines often present error propagation through the pipeline. We propose a framework, FOCAL to benchmark end-to-end reasoning, component-wise error propagation and error analysis for automated as well as human-assisted testing of multi-modal agents (voice to voice + text input). We also share two novel metrics viz. Reasoning and Semantic scores to evaluate efficacy of the agent in having meaningful conversations in voice mode.", "authors": ["Anupam Purwar", "Aditya Choudhary"], "categories": ["cs.SD"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.07367", "pdf_url": "https://arxiv.org/pdf/2601.07367v2", "arxiv_id": "2601.07367", "doi": "10.1109/COMSNETS67989.2026.11418101", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Communication Systems and Networks", "quality_score": 0.4411}
{"id": "14c5c71b8649e6d03feff2e20b52e369d78b118645cc6072b47a5ce5ac8b62cb", "sources": ["arxiv", "semantic_scholar"], "title": "SAGE: Tool-Augmented LLM Task Solving Strategies in Scalable Multi-Agent Environments", "abstract": "Large language models (LLMs) have proven to work well in question-answering scenarios, but real-world applications often require access to tools for live information or actuation. For this, LLMs can be extended with tools, which are often defined in advance, also allowing for some fine-tuning for specific use cases. However, rapidly evolving software landscapes and individual services require the constant development and integration of new tools. Domain- or company-specific tools can greatly elevate the usefulness of an LLM, but such custom tools can be problematic to integrate, or the LLM may fail to reliably understand and use them. For this, we need strategies to define new tools and integrate them into the LLM dynamically, as well as robust and scalable zero-shot prompting methods that can make use of those tools in an efficient manner. In this paper, we present SAGE, a specialized conversational AI interface, based on the OPACA framework for tool discovery and execution. The integration with OPACA makes it easy to add new tools or services for the LLM to use, while SAGE itself presents rich extensibility and modularity. This not only provides the ability to seamlessly switch between different models (e.g. GPT, LLAMA), but also to add and select prompting methods, involving various setups of differently prompted agents for selecting and executing tools and evaluating the results. We implemented a number of task-solving strategies, making use of agentic concepts and prompting methods in various degrees of complexity, and evaluated those against a comprehensive set of benchmark services. The results are promising and highlight the distinct strengths and weaknesses of different task-solving strategies. Both SAGE and the OPACA framework, as well as the different benchmark services and results, are available as Open Source/Open Data on GitHub.", "authors": ["Robert K. Strehlow", "Tobias Küster", "Oskar F. Kupke", "Brandon Llanque Kurps", "Fikret Sivrikaya", "Sahin Albayrak"], "categories": ["cs.SE", "cs.AI", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.09750", "pdf_url": "https://arxiv.org/pdf/2601.09750v1", "arxiv_id": "2601.09750", "doi": "10.48550/arXiv.2601.09750", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6818}
{"id": "1907210010513265efcfb9dd101f9307f6cca823592eef51fa8ac19b191b420a", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Verifiably Safe Tool Use for LLM Agents", "abstract": "Large language model (LLM)-based AI agents extend LLM capabilities by enabling access to tools such as data sources, APIs, search engines, code sandboxes, and even other agents. While this empowers agents to perform complex tasks, LLMs may invoke unintended tool interactions and introduce risks, such as leaking sensitive data or overwriting critical records, which are unacceptable in enterprise contexts. Current approaches to mitigate these risks, such as model-based safeguards, enhance agents' reliability but cannot guarantee system safety. Methods like information flow control (IFC) and temporal constraints aim to provide guarantees but often require extensive human annotation. We propose a process that starts with applying System-Theoretic Process Analysis (STPA) to identify hazards in agent workflows, derive safety requirements, and formalize them as enforceable specifications on data flows and tool sequences. To enable this, we introduce a capability-enhanced Model Context Protocol (MCP) framework that requires structured labels on capabilities, confidentiality, and trust level. Together, these contributions aim to shift LLM-based agent safety from ad hoc reliability fixes to proactive guardrails with formal guarantees, while reducing dependence on user confirmation and making autonomy a deliberate design choice.", "authors": ["Aarya Doshi", "Yining Hong", "Congying Xu", "Eunsuk Kang", "Alexandros Kapravelos", "Christian Kästner"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.08012", "pdf_url": "https://arxiv.org/pdf/2601.08012v1", "arxiv_id": "2601.08012", "doi": "10.48550/arXiv.2601.08012", "citation_count": 22, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4411}
{"id": "6a6c560d70b31f4e54c13ee5c7854d02e07cb9b6992a73552a262cf31741ddf0", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Cloud Network Resilience via a Robust LLM-Empowered Multi-Agent Reinforcement Learning Framework", "abstract": "While virtualization and resource pooling empower cloud networks with structural flexibility and elastic scalability, they inevitably expand the attack surface and challenge cyber resilience. Reinforcement Learning (RL)-based defense strategies have been developed to optimize resource deployment and isolation policies under adversarial conditions, aiming to enhance system resilience by maintaining and restoring network availability. However, existing approaches lack robustness as they require retraining to adapt to dynamic changes in network structure, node scale, attack strategies, and attack intensity. Furthermore, the lack of Human-in-the-Loop (HITL) support limits interpretability and flexibility. To address these limitations, we propose CyberOps-Bots, a hierarchical multi-agent reinforcement learning framework empowered by Large Language Models (LLMs). Inspired by MITRE ATT&CK's Tactics-Techniques model, CyberOps-Bots features a two-layer architecture: (1) An upper-level LLM agent with four modules--ReAct planning, IPDRR-based perception, long-short term memory, and action/tool integration--performs global awareness, human intent recognition, and tactical planning; (2) Lower-level RL agents, developed via heterogeneous separated pre-training, execute atomic defense actions within localized network regions. This synergy preserves LLM adaptability and interpretability while ensuring reliable RL execution. Experiments on real cloud datasets show that, compared to state-of-the-art algorithms, CyberOps-Bots maintains network availability 68.5% higher and achieves a 34.7% jumpstart performance gain when shifting the scenarios without retraining. To our knowledge, this is the first study to establish a robust LLM-RL framework with HITL support for cloud defense.", "authors": ["Yixiao Peng", "Hao Hu", "Feiyang Li", "Xinye Cao", "Yingchang Jiang", "Jipeng Tang", "Guoshun Nan", "Yuling Liu"], "categories": ["cs.CR", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.07122", "pdf_url": "https://arxiv.org/pdf/2601.07122v2", "arxiv_id": "2601.07122", "doi": "10.48550/arXiv.2601.07122", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Dependable and Secure Computing", "quality_score": 0.4411}
{"id": "e52bff9358c8caac8eea409d779b4a721b2eee52a98cd9ab34679a486935ea9c", "sources": ["arxiv", "semantic_scholar"], "title": "The Confidence Dichotomy: Analyzing and Mitigating Miscalibration in Tool-Use Agents", "abstract": "Autonomous agents based on large language models (LLMs) are rapidly evolving to handle multi-turn tasks, but ensuring their trustworthiness remains a critical challenge. A fundamental pillar of this trustworthiness is calibration, which refers to an agent's ability to express confidence that reliably reflects its actual performance. While calibration is well-established for static models, its dynamics in tool-integrated agentic workflows remain underexplored. In this work, we systematically investigate verbalized calibration in tool-use agents, revealing a fundamental confidence dichotomy driven by tool type. Specifically, our pilot study identifies that evidence tools (e.g., web search) systematically induce severe overconfidence due to inherent noise in retrieved information, while verification tools (e.g., code interpreters) can ground reasoning through deterministic feedback and mitigate miscalibration. To robustly improve calibration across tool types, we propose a reinforcement learning (RL) fine-tuning framework that jointly optimizes task accuracy and calibration, supported by a holistic benchmark of reward designs. We demonstrate that our trained agents not only achieve superior calibration but also exhibit robust generalization from local training environments to noisy web settings and to distinct domains such as mathematical reasoning. Our results highlight the necessity of domain-specific calibration strategies for tool-use agents. More broadly, this work establishes a foundation for building self-aware agents that can reliably communicate uncertainty in high-stakes, real-world deployments.", "authors": ["Weihao Xuan", "Qingcheng Zeng", "Heli Qi", "Yunze Xiao", "Junjue Wang", "Naoto Yokoya"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.07264", "pdf_url": "https://arxiv.org/pdf/2601.07264v1", "arxiv_id": "2601.07264", "doi": "10.48550/arXiv.2601.07264", "citation_count": 2, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4411}
{"id": "4b71f2168e425c20f3c39c37336e1cbf088d3d4eb8eb48a12ebad5f421fad4bc", "sources": ["arxiv", "semantic_scholar"], "title": "EZBlender: Efficient 3D Editing with Plan-and-ReAct Agent", "abstract": "As a cornerstone of the modern digital economy, 3D modeling and rendering demand substantial resources and manual effort when scene editing is performed in the traditional manner. Despite recent progress in VLM-based agents for 3D editing, the fundamental trade-off between editing precision and agent responsiveness remains unresolved. To overcome these limitations, we present EZBlender, a Blender agent with a hybrid framework that combines planning-based task decomposition and reactive local autonomy for efficient human AI collaboration and semantically faithful 3D editing. Specifically, this unexplored Plan-and-ReAct design not only preserves editing quality but also significantly reduces latency and computational cost. To further validate the efficiency and effectiveness of the proposed edge-autonomy architecture, we construct a dedicated multi-tasking benchmark that has not been systematically investigated in prior research. In addition, we provide a comprehensive analysis of language model preference, system responsiveness, and economic efficiency.", "authors": ["Hao Wang", "Wenhui Zhu", "Shao Tang", "Zhipeng Wang", "Xuanzhao Dong", "Xin Li", "Xiwen Chen", "Ashish Bastola", "Xinhao Huang", "Yalin Wang", "Abolfazl Razi"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.07143", "pdf_url": "https://arxiv.org/pdf/2601.07143v1", "arxiv_id": "2601.07143", "doi": "10.48550/arXiv.2601.07143", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4411}
{"id": "939d5e3ddfb679cd46d5b1c493b315989980706b66de3bfe19e347b91c867350", "sources": ["arxiv", "semantic_scholar"], "title": "FROAV: A Framework for RAG Observation and Agent Verification -- Lowering the Barrier to LLM Agent Research", "abstract": "The rapid advancement of Large Language Models (LLMs) and their integration into autonomous agent systems has created unprecedented opportunities for document analysis, decision support, and knowledge retrieval. However, the complexity of developing, evaluating, and iterating on LLM-based agent workflows presents significant barriers to researchers, particularly those without extensive software engineering expertise. We present FROAV (Framework for RAG Observation and Agent Verification), an open-source research platform that democratizes LLM agent research by providing a plug-and-play architecture combining visual workflow orchestration, a comprehensive evaluation framework, and extensible Python integration. FROAV implements a multi-stage Retrieval-Augmented Generation (RAG) pipeline coupled with a rigorous \"LLM-as-a-Judge\" evaluation system, all accessible through intuitive graphical interfaces. Our framework integrates n8n for no-code workflow design, PostgreSQL for granular data management, FastAPI for flexible backend logic, and Streamlit for human-in-the-loop interaction. Through this integrated ecosystem, researchers can rapidly prototype RAG strategies, conduct prompt engineering experiments, validate agent performance against human judgments, and collect structured feedback-all without writing infrastructure code. We demonstrate the framework's utility through its application to financial document analysis, while emphasizing its material-agnostic architecture that adapts to any domain requiring semantic analysis. FROAV represents a significant step toward making LLM agent research accessible to a broader scientific community, enabling researchers to focus on hypothesis testing and algorithmic innovation rather than system integration challenges.", "authors": ["Tzu-Hsuan Lin", "Chih-Hsuan Kao"], "categories": ["cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-12", "url": "https://arxiv.org/abs/2601.07504", "pdf_url": "https://arxiv.org/pdf/2601.07504v1", "arxiv_id": "2601.07504", "doi": "10.48550/arXiv.2601.07504", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6818}
{"id": "586c6b0bd4f19a5d6f42d2835188dd8e4a82b7ed71acc81295c9800e97887127", "sources": ["arxiv", "semantic_scholar"], "title": "Distilling Feedback into Memory-as-a-Tool", "abstract": "We propose a framework that amortizes the cost of inference-time reasoning by converting transient critiques into retrievable guidelines, through a file-based memory system and agent-controlled tool calls. We evaluate this method on the Rubric Feedback Bench, a novel dataset for rubric-based learning. Experiments demonstrate that our augmented LLMs rapidly match the performance of test-time refinement pipelines while drastically reducing inference cost.", "authors": ["Víctor Gallego"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-09", "url": "https://arxiv.org/abs/2601.05960", "pdf_url": "https://arxiv.org/pdf/2601.05960v2", "arxiv_id": "2601.05960", "doi": "10.48550/arXiv.2601.05960", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/vicgalle/feedback-memory-as-a-tool", "venue": "arXiv.org", "quality_score": 0.6765}
{"id": "0cab617eaae0f50dad44884a30b2c6bfa2c11f1fa13e7e2f2b8df69a2c1850c4", "sources": ["arxiv", "semantic_scholar"], "title": "EnvScaler: Scaling Tool-Interactive Environments for LLM Agent via Programmatic Synthesis", "abstract": "Large language models (LLMs) are expected to be trained to act as agents in various real-world environments, but this process relies on rich and varied tool-interaction sandboxes. However, access to real systems is often restricted; LLM-simulated environments are prone to hallucinations and inconsistencies; and manually built sandboxes are hard to scale. In this paper, we propose EnvScaler, an automated framework for scalable tool-interaction environments via programmatic synthesis. EnvScaler comprises two components. First, SkelBuilder constructs diverse environment skeletons through topic mining, logic modeling, and quality evaluation. Then, ScenGenerator generates multiple task scenarios and rule-based trajectory validation functions for each environment. With EnvScaler, we synthesize 191 environments and about 7K scenarios, and apply them to Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) for Qwen3 series models. Results on three benchmarks show that EnvScaler significantly improves LLMs' ability to solve tasks in complex environments involving multi-turn, multi-tool interactions. We release our code and data at https://github.com/RUC-NLPIR/EnvScaler.", "authors": ["Xiaoshuai Song", "Haofei Chang", "Guanting Dong", "Yutao Zhu", "Ji-Rong Wen", "Zhicheng Dou"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-09", "url": "https://arxiv.org/abs/2601.05808", "pdf_url": "https://arxiv.org/pdf/2601.05808v2", "arxiv_id": "2601.05808", "doi": "10.48550/arXiv.2601.05808", "citation_count": 17, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/RUC-NLPIR/EnvScaler", "venue": "arXiv.org", "quality_score": 0.6765}
{"id": "dd28c4e7c414f89f5c534c8297ef52e05c4158d157ad1ae9bd98c21d4d974291", "sources": ["arxiv", "semantic_scholar"], "title": "Conformity Dynamics in LLM Multi-Agent Systems: The Roles of Topology and Self-Social Weighting", "abstract": "Large Language Models (LLMs) are increasingly instantiated as interacting agents in multi-agent systems (MAS), where collective decisions emerge through social interaction rather than independent reasoning. A fundamental yet underexplored mechanism in this process is conformity, the tendency of agents to align their judgments with prevailing group opinions. This paper presents a systematic study of how network topology shapes conformity dynamics in LLM-based MAS through a misinformation detection task. We introduce a confidence-normalized pooling rule that controls the trade-off between self-reliance and social influence, enabling comparisons between two canonical decision paradigms: Centralized Aggregation and Distributed Consensus. Experimental results demonstrate that network topology critically governs both the efficiency and robustness of collective judgments. Centralized structures enable immediate decisions but are sensitive to hub competence and exhibit same-model alignment biases. In contrast, distributed structures promote more robust consensus, while increased network connectivity speeds up convergence but also heightens the risk of wrong-but-sure cascades, in which agents converge on incorrect decisions with high confidence. These findings characterize the conformity dynamics in LLM-based MAS, clarifying how network topology and self-social weighting jointly shape the efficiency, robustness, and failure modes of collective decision-making.", "authors": ["Chen Han", "Jin Tan", "Bohan Yu", "Wenzhen Zheng", "Xijin Tang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-09", "url": "https://arxiv.org/abs/2601.05606", "pdf_url": "https://arxiv.org/pdf/2601.05606v1", "arxiv_id": "2601.05606", "doi": "10.48550/arXiv.2601.05606", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4377}
{"id": "2db9d8fab2f80a4caa845798dbd4711b0dc0adf78cb0f124a0e1513e5e1e95e0", "sources": ["arxiv", "semantic_scholar"], "title": "DynaDebate: Breaking Homogeneity in Multi-Agent Debate with Dynamic Path Generation", "abstract": "Recent years have witnessed the rapid development of Large Language Model-based Multi-Agent Systems (MAS), which excel at collaborative decision-making and complex problem-solving. Recently, researchers have further investigated Multi-Agent Debate (MAD) frameworks, which enhance the reasoning and collaboration capabilities of MAS through information exchange and debate among multiple agents. However, existing approaches often rely on unguided initialization, causing agents to adopt identical reasoning paths that lead to the same errors. As a result, effective debate among agents is hindered, and the final outcome frequently degenerates into simple majority voting. To solve the above problem, in this paper, we introduce Dynamic Multi-Agent Debate (DynaDebate), which enhances the effectiveness of multi-agent debate through three key mechanisms: (1) Dynamic Path Generation and Allocation, which employs a dedicated Path Generation Agent to generate diverse and logical solution paths with adaptive redundancy; (2) Process-Centric Debate, which shifts the focus from surface-level outcome voting to rigorous step-by-step logic critique to ensure process correctness; (3) A Trigger-Based Verification Agent, which is activated upon disagreement and uses external tools to objectively resolve deadlocks. Extensive experiments demonstrate that DynaDebate achieves superior performance across various benchmarks, surpassing existing state-of-the-art MAD methods.", "authors": ["Zhenghao Li", "Zhi Zheng", "Wei Chen", "Jielun Zhao", "Yong Chen", "Tong Xu", "Enhong Chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-09", "url": "https://arxiv.org/abs/2601.05746", "pdf_url": "https://arxiv.org/pdf/2601.05746v1", "arxiv_id": "2601.05746", "doi": "10.48550/arXiv.2601.05746", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4377}
{"id": "3a046a24505b7e055dc3b961f3fc6fd15ac791046d39626b420acb6734c9f17e", "sources": ["arxiv", "semantic_scholar"], "title": "Tool-MAD: A Multi-Agent Debate Framework for Fact Verification with Diverse Tool Augmentation and Adaptive Retrieval", "abstract": "Large Language Models (LLMs) suffer from hallucinations and factual inaccuracies, especially in complex reasoning and fact verification tasks. Multi-Agent Debate (MAD) systems aim to improve answer accuracy by enabling multiple LLM agents to engage in dialogue, promoting diverse reasoning and mutual verification. However, existing MAD frameworks primarily rely on internal knowledge or static documents, making them vulnerable to hallucinations. While MADKE introduces external evidence to mitigate this, its one-time retrieval mechanism limits adaptability to new arguments or emerging information during the debate. To address these limitations, We propose Tool-MAD, a multi-agent debate framework that enhances factual verification by assigning each agent a distinct external tool, such as a search API or RAG module. Tool-MAD introduces three key innovations: (1) a multi-agent debate framework where agents leverage heterogeneous external tools, encouraging diverse perspectives, (2) an adaptive query formulation mechanism that iteratively refines evidence retrieval based on the flow of the debate, and (3) the integration of Faithfulness and Answer Relevance scores into the final decision process, allowing the Judge agent to quantitatively assess the coherence and question alignment of each response and effectively detect hallucinations. Experimental results on four fact verification benchmarks demonstrate that Tool-MAD consistently outperforms state-of-the-art MAD frameworks, achieving up to 5.5% accuracy improvement. Furthermore, in medically specialized domains, Tool-MAD exhibits strong robustness and adaptability across various tool configurations and domain conditions, confirming its potential for broader real-world fact-checking applications.", "authors": ["Seyeon Jeong", "Yeonjun Choi", "JongWook Kim", "Beakcheol Jang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04742", "pdf_url": "https://arxiv.org/pdf/2601.04742v1", "arxiv_id": "2601.04742", "doi": "10.48550/arXiv.2601.04742", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4366}
{"id": "a1c4352e091dbce949799410ddce636f6df4063a0afa468dda5987936d0373da", "sources": ["arxiv", "semantic_scholar"], "title": "ResMAS: Resilience Optimization in LLM-based Multi-agent Systems", "abstract": "Large Language Model-based Multi-Agent Systems (LLM-based MAS), where multiple LLM agents collaborate to solve complex tasks, have shown impressive performance in many areas. However, MAS are typically distributed across different devices or environments, making them vulnerable to perturbations such as agent failures. While existing works have studied the adversarial attacks and corresponding defense strategies, they mainly focus on reactively detecting and mitigating attacks after they occur rather than proactively designing inherently resilient systems. In this work, we study the resilience of LLM-based MAS under perturbations and find that both the communication topology and prompt design significantly influence system resilience. Motivated by these findings, we propose ResMAS: a two-stage framework for enhancing MAS resilience. First, we train a reward model to predict the MAS's resilience, based on which we train a topology generator to automatically design resilient topology for specific tasks through reinforcement learning. Second, we introduce a topology-aware prompt optimization method that refines each agent's prompt based on its connections and interactions with other agents. Extensive experiments across a range of tasks show that our approach substantially improves MAS resilience under various constraints. Moreover, our framework demonstrates strong generalization ability to new tasks and models, highlighting its potential for building resilient MASs.", "authors": ["Zhilun Zhou", "Zihan Liu", "Jiahe Liu", "Qingyu Shao", "Yihan Wang", "Kun Shao", "Depeng Jin", "Fengli Xu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04694", "pdf_url": "https://arxiv.org/pdf/2601.04694v1", "arxiv_id": "2601.04694", "doi": "10.48550/arXiv.2601.04694", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.4366}
{"id": "b11ff63878504b7ef6817ca7f448ea481da6e12213a9d8c22e2fd8115639e826", "sources": ["arxiv", "semantic_scholar"], "title": "A Closed-Loop Multi-Agent System Driven by LLMs for Meal-Level Personalized Nutrition Management", "abstract": "Personalized nutrition management aims to tailor dietary guidance to an individual's intake and phenotype, but most existing systems handle food logging, nutrient analysis and recommendation separately. We present a next-generation mobile nutrition assistant that combines image based meal logging with an LLM driven multi agent controller to provide meal level closed loop support. The system coordinates vision, dialogue and state management agents to estimate nutrients from photos and update a daily intake budget. It then adapts the next meal plan to user preferences and dietary constraints. Experiments with SNAPMe meal images and simulated users show competitive nutrient estimation, personalized menus and efficient task plans. These findings demonstrate the feasibility of multi agent LLM control for personalized nutrition and reveal open challenges in micronutrient estimation from images and in large scale real world studies.", "authors": ["Muqing Xu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04491", "pdf_url": "https://arxiv.org/pdf/2601.04491v1", "arxiv_id": "2601.04491", "doi": "10.1109/RAAI67517.2025.11423373", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2778}
{"id": "ae94baa80b69413a04dfd37a430a3efd2fdc25a9830ae1d8f6f69c3881f4952b", "sources": ["arxiv", "semantic_scholar"], "title": "CircuitLM: A Multi-Agent LLM-Aided Design Framework for Generating Circuit Schematics from Natural Language Prompts", "abstract": "Generating accurate circuit schematics from high-level natural language descriptions remains a persistent challenge in electronic design automation (EDA), as large language models (LLMs) frequently hallucinate components, violate strict physical constraints, and produce non-machine-readable outputs. To address this, we present CircuitLM, a multi-agent pipeline that translates user prompts into structured, visually interpretable $\\texttt{CircuitJSON}$ schematics. The framework mitigates hallucination and ensures physical viability by grounding generation in a curated, embedding-powered component knowledge base through five sequential stages: (i) component identification, (ii) canonical pinout retrieval, (iii) chain-of-thought reasoning, (iv) JSON schematic synthesis, and (v) interactive force-directed visualization. We evaluate the system on a dataset of 100 unique circuit-design prompts using five state-of-the-art LLMs. To systematically assess performance, we deploy a rigorous dual-layered evaluation methodology: a deterministic Electrical Rule Checking (ERC) engine categorizes topological faults by strict severity (Critical, Major, Minor, Warning), while an LLM-as-a-judge meta-evaluator identifies complex, context-aware design flaws that bypass standard rule-based checkers. Ultimately, this work demonstrates how targeted retrieval combined with deterministic and semantic verification can bridge natural language to structurally viable, schematic-ready hardware and safe circuit prototyping. Our code and data are publicly available at https://github.com/Khandakar227/CircuitLM.", "authors": ["Khandakar Shakib Al Hasan", "Syed Rifat Raiyan", "Hasin Mahtab Alvee", "Wahid Sadik"], "categories": ["cs.AI", "cs.CL", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04505", "pdf_url": "https://arxiv.org/pdf/2601.04505v3", "arxiv_id": "2601.04505", "doi": "10.48550/arXiv.2601.04505", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Khandakar227/CircuitLM", "venue": "arXiv.org", "quality_score": 0.6747}
{"id": "f23516466828309b3e638723f4eee2d299c843011d7cb584888c99d17895b029", "sources": ["arxiv", "semantic_scholar"], "title": "When Single-Agent with Skills Replace Multi-Agent Systems and When They Fail", "abstract": "Multi-agent AI systems have proven effective for complex reasoning. These systems are compounded by specialized agents, which collaborate through explicit communication, but incur substantial computational overhead. A natural question arises: can we achieve similar modularity benefits with a single agent that selects from a library of skills? We explore this question by viewing skills as internalized agent behaviors. From this perspective, a multi-agent system can be compiled into an equivalent single-agent system, trading inter-agent communication for skill selection. Our preliminary experiments suggest this approach can substantially reduce token usage and latency while maintaining competitive accuracy on reasoning benchmarks. However, this efficiency raises a deeper question that has received little attention: how does skill selection scale as libraries grow? Drawing on principles from cognitive science, we propose that LLM skill selection exhibits bounded capacity analogous to human decision-making. We investigate the scaling behavior of skill selection and observe a striking pattern. Rather than degrading gradually, selection accuracy remains stable up to a critical library size, then drops sharply, indicating a phase transition reminiscent of capacity limits in human cognition. Furthermore, we find evidence that semantic confusability among similar skills, rather than library size alone, plays a central role in this degradation. This perspective suggests that hierarchical organization, which has long helped humans manage complex choices, may similarly benefit AI systems. Our initial results with hierarchical routing support this hypothesis. This work opens new questions about the fundamental limits of semantic-based skill selection in LLMs and offers a cognitive-grounded framework and practical guidelines for designing scalable skill-based agents.", "authors": ["Xiaoxiao Li"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04748", "pdf_url": "https://arxiv.org/pdf/2601.04748v2", "arxiv_id": "2601.04748", "doi": "10.48550/arXiv.2601.04748", "citation_count": 25, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4366}
{"id": "cabac0b79ea7210814651114c92783de8bfd566cb94ca937204cfb484c137e40", "sources": ["arxiv", "semantic_scholar"], "title": "Precomputing Multi-Agent Path Replanning Using Temporal Flexibility", "abstract": "Executing a multi-agent plan can be challenging when an agent is delayed, because this typically creates conflicts with other agents. So, we need to quickly find a new safe plan. Replanning only the delayed agent often does not yield an efficient plan, and sometimes cannot even yield a feasible one. On the other hand, replanning other agents may lead to a cascade of changes and delays, and it is computationally expensive. We show how to efficiently replan a single delayed agent by tracking and using the temporal flexibility of other agents while avoiding cascading delays. This flexibility is the maximum delay that the agent can take without changing the order with agents other than the initially delayed agent, or further delaying other agents. Our algorithm, FlexSIPP, precomputes all possible plans for the delayed agent and returns the changes to the other agents within the given scenario. We demonstrate our method in a real-world case study of replanning trains in the densely-used Dutch railway network and in the MovingAI MAPF benchmark set. Our experiments show that FlexSIPP provides effective solutions relevant to real-world adjustments, and within a reasonable timeframe.", "authors": ["Issa Hanou", "Eric Kemmeren", "Devin Wild Thomas", "Mathijs de Weerdt"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04884", "pdf_url": "https://arxiv.org/pdf/2601.04884v3", "arxiv_id": "2601.04884", "doi": "10.48550/arXiv.2601.04884", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4366}
{"id": "89b26c83146b0341f59f03372dde5f63d28a393056843fdd25e2a544d2b69276", "sources": ["arxiv", "semantic_scholar"], "title": "Orchestrating Intelligence: Confidence-Aware Routing for Efficient Multi-Agent Collaboration across Multi-Scale Models", "abstract": "While multi-agent systems (MAS) have demonstrated superior performance over single-agent approaches in complex reasoning tasks, they often suffer from significant computational inefficiencies. Existing frameworks typically deploy large language models (LLMs) uniformly across all agent roles, failing to account for the varying cognitive demands of different reasoning stages. We address this inefficiency by proposing OI-MAS framework, a novel multi-agent framework that implements an adaptive model-selection policy across a heterogeneous pool of multi-scale LLMs. Specifically, OI-MAS introduces a state-dependent routing mechanism that dynamically selects agent roles and model scales throughout the reasoning process. In addition, we introduce a confidence-aware mechanism that selects appropriate model scales conditioned on task complexity, thus reducing unnecessary reliance on large-scale models. Experimental results show that OI-MAS consistently outperforms baseline multi-agent systems, improving accuracy by up to 12.88\\% while reducing cost by up to 79.78\\%.", "authors": ["Jingbo Wang", "Sendong Zhao", "Jiatong Liu", "Haochun Wang", "Wanting Li", "Bing Qin", "Ting Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04861", "pdf_url": "https://arxiv.org/pdf/2601.04861v2", "arxiv_id": "2601.04861", "doi": "10.48550/arXiv.2601.04861", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4366}
{"id": "17ae483ec9df69e82b416121de98c0879512e84b00e2a2e66dc4cd4b9cffe0e8", "sources": ["arxiv", "semantic_scholar"], "title": "BackdoorAgent: A Unified Framework for Backdoor Attacks on LLM-based Agents", "abstract": "Large language model (LLM) agents execute tasks through multi-step workflows that combine planning, memory, and tool use. While this design enables autonomy, it also expands the attack surface for backdoor threats. Backdoor triggers injected into specific stages of an agent workflow can persist through multiple intermediate states and adversely influence downstream outputs. However, existing studies remain fragmented and typically analyze individual attack vectors in isolation, leaving the cross-stage interaction and propagation of backdoor triggers poorly understood from an agent-centric perspective. To fill this gap, we propose \\textbf{BackdoorAgent}, a modular and stage-aware framework that provides a unified, agent-centric view of backdoor threats in LLM agents. BackdoorAgent structures the attack surface into three functional stages of agentic workflows, including \\textbf{planning attacks}, \\textbf{memory attacks}, and \\textbf{tool-use attacks}, and instruments agent execution to enable systematic analysis of trigger activation and propagation across different stages. Building on this framework, we construct a standardized benchmark spanning four representative agent applications: \\textbf{Agent QA}, \\textbf{Agent Code}, \\textbf{Agent Web}, and \\textbf{Agent Drive}, covering both language-only and multimodal settings. Our empirical analysis shows that \\textit{triggers implanted at a single stage can persist across multiple steps and propagate through intermediate states.} For instance, when using a GPT-based backbone, we observe trigger persistence in 43.58\\% of planning attacks, 77.97\\% of memory attacks, and 60.28\\% of tool-stage attacks, highlighting the vulnerabilities of the agentic workflow itself to backdoor threats. To facilitate reproducibility and future research, our code and benchmark are publicly available at GitHub.", "authors": ["Yunhao Feng", "Yige Li", "Yutao Wu", "Yingshui Tan", "Yanming Guo", "Yifan Ding", "Kun Zhai", "Xingjun Ma", "Yu-Gang Jiang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04566", "pdf_url": "https://arxiv.org/pdf/2601.04566v2", "arxiv_id": "2601.04566", "doi": "10.48550/arXiv.2601.04566", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4366}
{"id": "97d386e4e5ff579268c088d474922b5271282454a9cad75356c4ce365e8863cf", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond Monolithic Architectures: A Multi-Agent Search and Knowledge Optimization Framework for Agentic Search", "abstract": "Agentic search has emerged as a promising paradigm for complex information seeking by enabling Large Language Models (LLMs) to interleave reasoning with tool use. However, prevailing systems rely on monolithic agents that suffer from structural bottlenecks, including unconstrained reasoning outputs that inflate trajectories, sparse outcome-level rewards that complicate credit assignment, and stochastic search noise that destabilizes learning. To address these challenges, we propose \\textbf{M-ASK} (Multi-Agent Search and Knowledge), a framework that explicitly decouples agentic search into two complementary roles: Search Behavior Agents, which plan and execute search actions, and Knowledge Management Agents, which aggregate, filter, and maintain a compact internal context. This decomposition allows each agent to focus on a well-defined subtask and reduces interference between search and context construction. Furthermore, to enable stable coordination, M-ASK employs turn-level rewards to provide granular supervision for both search decisions and knowledge updates. Experiments on multi-hop QA benchmarks demonstrate that M-ASK outperforms strong baselines, achieving not only superior answer accuracy but also significantly more stable training dynamics.\\footnote{The source code for M-ASK is available at https://github.com/chenyiqun/M-ASK.}", "authors": ["Yiqun Chen", "Lingyong Yan", "Zixuan Yang", "Erhan Zhang", "Jiashu Zhao", "Shuaiqiang Wang", "Dawei Yin", "Jiaxin Mao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-08", "url": "https://arxiv.org/abs/2601.04703", "pdf_url": "https://arxiv.org/pdf/2601.04703v1", "arxiv_id": "2601.04703", "doi": "10.48550/arXiv.2601.04703", "citation_count": 9, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/chenyiqun/M-ASK.}", "venue": "arXiv.org", "quality_score": 0.6747}
{"id": "567cb3ff7f28fd078dfb3f4c6f3a88e6f536e3aae4cd20c3fb7ea9b3b1162216", "sources": ["arxiv", "semantic_scholar"], "title": "Embedding Autonomous Agents in Resource-Constrained Robotic Platforms", "abstract": "Many embedded devices operate under resource constraints and in dynamic environments, requiring local decision-making capabilities. Enabling devices to make independent decisions in such environments can improve the responsiveness of the system and reduce the dependence on constant external control. In this work, we integrate an autonomous agent, programmed using AgentSpeak, with a small two-wheeled robot that explores a maze using its own decision-making and sensor data. Experimental results show that the agent successfully solved the maze in 59 seconds using 287 reasoning cycles, with decision phases taking less than one millisecond. These results indicate that the reasoning process is efficient enough for real-time execution on resource-constrained hardware. This integration demonstrates how high-level agent-based control can be applied to resource-constrained embedded systems for autonomous operation.", "authors": ["Negar Halakou", "Juan F. Gutierrez", "Ye Sun", "Han Jiang", "Xueming Wu", "Yilun Song", "Andres Gomez"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-07", "url": "https://arxiv.org/abs/2601.04191", "pdf_url": "https://arxiv.org/pdf/2601.04191v1", "arxiv_id": "2601.04191", "doi": "10.48550/arXiv.2601.04191", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4354}
{"id": "807e4b07e3c1f5d6a35aa03a17db2e37f7aefb8aa662744f25df15769616f7b3", "sources": ["arxiv", "semantic_scholar"], "title": "MALTopic: Multi-Agent LLM Topic Modeling Framework", "abstract": "Topic modeling is a crucial technique for extracting latent themes from unstructured text data, particularly valuable in analyzing survey responses. However, traditional methods often only consider free-text responses and do not natively incorporate structured or categorical survey responses for topic modeling. And they produce abstract topics, requiring extensive human interpretation. To address these limitations, we propose the Multi-Agent LLM Topic Modeling Framework (MALTopic). This framework decomposes topic modeling into specialized tasks executed by individual LLM agents: an enrichment agent leverages structured data to enhance textual responses, a topic modeling agent extracts latent themes, and a deduplication agent refines the results. Comparative analysis on a survey dataset demonstrates that MALTopic significantly improves topic coherence, diversity, and interpretability compared to LDA and BERTopic. By integrating structured data and employing a multi-agent approach, MALTopic generates human-readable topics with enhanced contextual relevance, offering a more effective solution for analyzing complex survey data.", "authors": ["Yash Sharma"], "categories": ["cs.CL", "cs.IR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-07", "url": "https://arxiv.org/abs/2601.15299", "pdf_url": "https://arxiv.org/pdf/2601.15299v1", "arxiv_id": "2601.15299", "doi": "10.1109/AIIoT65859.2025.11105319", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/yash91sharma/MALTopic", "venue": "2025 IEEE AI-IoT", "quality_score": 0.6729}
{"id": "99d76170b5600dff59d0a52344bda28cbb7e22b0a744e5c07dc2f4b6eeb4557b", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Drift: Quantifying Behavioral Degradation in Multi-Agent LLM Systems Over Extended Interactions", "abstract": "Multi-agent Large Language Model (LLM) systems have emerged as powerful architectures for complex task decomposition and collaborative problem-solving. However, their long-term behavioral stability remains largely unexamined. This study introduces the concept of agent drift, defined as the progressive degradation of agent behavior, decision quality, and inter-agent coherence over extended interaction sequences. We present a comprehensive theoretical framework for understanding drift phenomena, proposing three distinct manifestations: semantic drift (progressive deviation from original intent), coordination drift (breakdown in multi-agent consensus mechanisms), and behavioral drift (emergence of unintended strategies). We introduce the Agent Stability Index (ASI), a novel composite metric framework for quantifying drift across twelve dimensions, including response consistency, tool usage patterns, reasoning pathway stability, and inter-agent agreement rates. Through simulation-based analysis and theoretical modeling, we demonstrate how unchecked agent drift can lead to substantial reductions in task completion accuracy and increased human intervention requirements. We propose three mitigation strategies: episodic memory consolidation, drift-aware routing protocols, and adaptive behavioral anchoring. Theoretical analysis suggests these approaches can significantly reduce drift-related errors while maintaining system throughput. This work establishes a foundational methodology for monitoring, measuring, and mitigating agent drift in production agentic AI systems, with direct implications for enterprise deployment reliability and AI safety research.", "authors": ["Abhishek Rath"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-07", "url": "https://arxiv.org/abs/2601.04170", "pdf_url": "https://arxiv.org/pdf/2601.04170v1", "arxiv_id": "2601.04170", "doi": "10.48550/arXiv.2601.04170", "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4354}
{"id": "b052321cfad6878a5559cfe7ba7387931f3a58224eb5a63cf6d62a2b7a346164", "sources": ["arxiv", "semantic_scholar"], "title": "When Numbers Start Talking: Implicit Numerical Coordination Among LLM-Based Agents", "abstract": "LLMs-based agents increasingly operate in multi-agent environments where strategic interaction and coordination are required. While existing work has largely focused on individual agents or on interacting agents sharing explicit communication, less is known about how interacting agents coordinate implicitly. In particular, agents may engage in covert communication, relying on indirect or non-linguistic signals embedded in their actions rather than on explicit messages. This paper presents a game-theoretic study of covert communication in LLM-driven multi-agent systems. We analyse interactions across four canonical game-theoretic settings under different communication regimes, including explicit, restricted, and absent communication. Considering heterogeneous agent personalities and both one-shot and repeated games, we characterise when covert signals emerge and how they shape coordination and strategic outcomes.", "authors": ["Alessio Buscemi", "Daniele Proverbio", "Alessandro Di Stefano", "The-Anh Han", "German Castignani", "Pietro Liò"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-07", "url": "https://arxiv.org/abs/2601.03846", "pdf_url": "https://arxiv.org/pdf/2601.03846v2", "arxiv_id": "2601.03846", "doi": "10.48550/arXiv.2601.03846", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4354}
{"id": "717363ab874524401a133be57947700904a2ca2e3092aea43b3bb88be4a05406", "sources": ["arxiv", "semantic_scholar"], "title": "Understanding Multi-Agent Reasoning with Large Language Models for Cartoon VQA", "abstract": "Visual Question Answering (VQA) for stylised cartoon imagery presents challenges, such as interpreting exaggerated visual abstraction and narrative-driven context, which are not adequately addressed by standard large language models (LLMs) trained on natural images. To investigate this issue, a multi-agent LLM framework is introduced, specifically designed for VQA tasks in cartoon imagery. The proposed architecture consists of three specialised agents: visual agent, language agent and critic agent, which work collaboratively to support structured reasoning by integrating visual cues and narrative context. The framework was systematically evaluated on two cartoon-based VQA datasets: Pororo and Simpsons. Experimental results provide a detailed analysis of how each agent contributes to the final prediction, offering a deeper understanding of LLM-based multi-agent behaviour in cartoon VQA and multimodal inference.", "authors": ["Tong Wu", "Thanet Markchom"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-06", "url": "https://arxiv.org/abs/2601.03073", "pdf_url": "https://arxiv.org/pdf/2601.03073v1", "arxiv_id": "2601.03073", "doi": "10.48550/arXiv.2601.03073", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4343}
{"id": "0a5aab15e6e1d95c7369dcb60906c2733bcfe85cb9f5e7ec377fa4f7fa59cc29", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Enabled Multi-Agent Systems: Empirical Evaluation and Insights into Emerging Design Patterns & Paradigms", "abstract": "This paper formalises the literature on emerging design patterns and paradigms for Large Language Model (LLM)-enabled multi-agent systems (MAS), evaluating their practical utility across various domains. We define key architectural components, including agent orchestration, communication mechanisms, and control-flow strategies, and demonstrate how these enable rapid development of modular, domain-adaptive solutions. Three real-world case studies are tested in controlled, containerised pilots in telecommunications security, national heritage asset management, and utilities customer service automation. Initial empirical results show that, for these case studies, prototypes were delivered within two weeks and pilot-ready solutions within one month, suggesting reduced development overhead compared to conventional approaches and improved user accessibility. However, findings also reinforce limitations documented in the literature, including variability in LLM behaviour that leads to challenges in transitioning from prototype to production maturity. We conclude by outlining critical research directions for improving reliability, scalability, and governance in MAS architectures and the further work needed to mature MAS design patterns to mitigate the inherent challenges.", "authors": ["Harri Renney", "Maxim N Nethercott", "Nathan Renney", "Peter Hayes"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-06", "url": "https://arxiv.org/abs/2601.03328", "pdf_url": "https://arxiv.org/pdf/2601.03328v1", "arxiv_id": "2601.03328", "doi": "10.48550/arXiv.2601.03328", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4343}
{"id": "31c99d88b4876e11804c13c7772f0186e52573b42e0b5878ec21cb65d930da98", "sources": ["arxiv", "semantic_scholar"], "title": "AgentVNE: LLM-Augmented Graph Reinforcement Learning for Affinity-Aware Multi-Agent Placement in Edge Agentic AI", "abstract": "The Internet of Agents is propelling edge computing toward agentic AI and edge general intelligence (EGI). However, deploying multi-agent service (MAS) on resource-constrained edge infrastructure presents severe challenges. MAS service workflows are driven by complex cross-node interactions, dynamic memory accumulation, and collaborative tool usage. Exhibiting chain-like topological dependencies and strict affinity constraints, these workflows demand real-time responsiveness that exceeds the capabilities of traditional VNE algorithms designed for static resources. To address this, we propose AgentVNE, a cloud-edge collaborative framework utilizing a dual-layer architecture. First, AgentVNE employs a large language model (LLM) to identify implicit semantic constraints and generate affinity-based resource augmentation to resolve physical dependency issues. Second, it constructs a resource similarity-aware neural network, utilizing a pre-training and PPO fine-tuning strategy to precisely capture topological similarities between dynamic workflows and heterogeneous networks. By coupling semantic perception with topological reasoning, this mechanism effectively bridges the gap between dynamic service requirements and physical infrastructure. Simulation results demonstrate that AgentVNE reduces workflow communication latency to less than 40% of baselines and improves the service acceptance rate by approximately 5%-10% under high-load scenarios. Ultimately, this work provides a foundational solution for the semantic-aware deployment of agentic AI.", "authors": ["Runze Zheng", "Yuqing Zheng", "Zhengyi Cheng", "Long Luo", "Haoxiang Luo", "Gang Sun", "Hongfang Yu", "Dusit Niyato"], "categories": ["cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-05", "url": "https://arxiv.org/abs/2601.02021", "pdf_url": "https://arxiv.org/pdf/2601.02021v1", "arxiv_id": "2601.02021", "doi": "10.48550/arXiv.2601.02021", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4331}
{"id": "b3bd2fbe893eb09a3f8869f8ba89843c5e4b3f9c6aafd59e59f455e5525073a2", "sources": ["arxiv", "semantic_scholar"], "title": "The Rise of Agentic Testing: Multi-Agent Systems for Robust Software Quality Assurance", "abstract": "Software testing has progressed toward intelligent automation, yet current AI-based test generators still suffer from static, single-shot outputs that frequently produce invalid, redundant, or non-executable tests due to the lack of execution aware feedback. This paper introduces an agentic multi-model testing framework a closed-loop, self-correcting system in which a Test Generation Agent, an Execution and Analysis Agent, and a Review and Optimization Agent collaboratively generate, execute, analyze, and refine tests until convergence. By using sandboxed execution, detailed failure reporting, and iterative regeneration or patching of failing tests, the framework autonomously improves test quality and expands coverage. Integrated into a CI/CD-compatible pipeline, it leverages reinforcement signals from coverage metrics and execution outcomes to guide refinement. Empirical evaluations on microservice based applications show up to a 60% reduction in invalid tests, 30% coverage improvement, and significantly reduced human effort compared to single-model baselines demonstrating that multi-agent, feedback-driven loops can evolve software testing into an autonomous, continuously learning quality assurance ecosystem for self-healing, high-reliability codebases.", "authors": ["Saba Naqvi", "Mohammad Baqar", "Nawaz Ali Mohammad"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-05", "url": "https://arxiv.org/abs/2601.02454", "pdf_url": "https://arxiv.org/pdf/2601.02454v1", "arxiv_id": "2601.02454", "doi": "10.48550/arXiv.2601.02454", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4331}
{"id": "85ae3c876323153e4e98d7a091c73cb1fb060898f945877c6d44e401c79974f3", "sources": ["arxiv", "semantic_scholar"], "title": "ReliabilityBench: Evaluating LLM Agent Reliability Under Production-Like Stress Conditions", "abstract": "Existing benchmarks for tool-using LLM agents primarily report single-run success rates and miss reliability properties required in production. We introduce \\textbf{ReliabilityBench}, a benchmark for evaluating agent reliability across three dimensions: (i) consistency under repeated execution using $\\mathrm{pass}^k$, (ii) robustness to semantically equivalent task perturbations at intensity $ε$, and (iii) fault tolerance under controlled tool/API failures at intensity $λ$. ReliabilityBench contributes a unified reliability surface $R(k,ε,λ)$, \\textit{action metamorphic relations} that define correctness via end-state equivalence rather than text similarity, and a chaos-engineering-style fault injection framework (timeouts, rate limits, partial responses, schema drift). We evaluate two models (Gemini 2.0 Flash, GPT-4o) and two agent architectures (ReAct, Reflexion) across four domains (scheduling, travel, customer support, e-commerce) over 1,280 episodes. Perturbations alone reduce success from 96.9% at $ε=0$ to 88.1% at $ε=0.2$. Rate limiting is the most damaging fault in ablations. ReAct is more robust than Reflexion under combined stress, and Gemini 2.0 Flash achieves comparable reliability to GPT-4o at much lower cost. ReliabilityBench provides a systematic framework for assessing production readiness of LLM agents.", "authors": ["Aayush Gupta"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-03", "url": "https://arxiv.org/abs/2601.06112", "pdf_url": "https://arxiv.org/pdf/2601.06112v1", "arxiv_id": "2601.06112", "doi": "10.48550/arXiv.2601.06112", "citation_count": 10, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4308}
{"id": "794a9dde408cc068845f9bfb9805e1d0d31e90da5dd6f2da7862f1012ef1cf5c", "sources": ["arxiv", "semantic_scholar"], "title": "Space Debris Removal using Nano-Satellites controlled by Low-Power Autonomous Agents", "abstract": "Space debris is an ever-increasing problem in space travel. There are already many old, no longer functional spacecraft and debris orbiting the earth, which endanger both the safe operation of satellites and space travel. Small nano-satellite swarms can address this problem by autonomously de-orbiting debris safely into the Earth's atmosphere. This work builds on the recent advances of autonomous agents deployed in resource-constrained platforms and shows a first simplified approach how such intelligent and autonomous nano-satellite swarms can be realized. We implement our autonomous agent software on wireless microcontrollers and perform experiments on a specialized test-bed to show the feasibility and overall energy efficiency of our approach.", "authors": ["Dennis Christmann", "Juan F. Gutierrez", "Sthiti Padhi", "Patrick Plörer", "Aditya Takur", "Simona Silvestri", "Andres Gomez"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-01", "url": "https://arxiv.org/abs/2601.00465", "pdf_url": "https://arxiv.org/pdf/2601.00465v1", "arxiv_id": "2601.00465", "doi": "10.48550/arXiv.2601.00465", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4285}
{"id": "7ed0e2da42e443c909260207f6bb189843bb7434e03d6cc0ad217da0f3dcee1b", "sources": ["arxiv", "semantic_scholar"], "title": "Mapping Human Anti-collusion Mechanisms to Multi-agent AI Systems", "abstract": "As multi-agent AI systems become increasingly autonomous, evidence shows they can develop collusive strategies similar to those long observed in human markets and institutions. While human domains have accumulated centuries of anti-collusion mechanisms, it remains unclear how these can be adapted to AI settings. This paper addresses that gap by (i) developing a taxonomy of human anti-collusion mechanisms, including sanctions, leniency & whistleblowing, monitoring & auditing, market design, and governance and (ii) mapping them to potential interventions for multi-agent AI systems. For each mechanism, we propose implementation approaches. We also highlight open challenges, such as the attribution problem (difficulty attributing emergent coordination to specific agents), identity fluidity (agents being easily forked or modified), the boundary problem (distinguishing beneficial cooperation from harmful collusion), and adversarial adaptation (agents learning to evade detection).", "authors": ["Jamiu Idowu", "Ahmed Almasoud", "Ayman Alfahid"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-01", "url": "https://arxiv.org/abs/2601.00360", "pdf_url": "https://arxiv.org/pdf/2601.00360v3", "arxiv_id": "2601.00360", "doi": "10.1016/j.knosys.2026.116067", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Knowledge-Based Systems", "quality_score": 0.4285}
{"id": "83794b70e50688f12881394cda4660a3bffe3ab44b257d4dafda7881a5d78ef7", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Coordinated Rename Refactoring", "abstract": "The primary value of AI agents in software development lies in their ability to extend the developer's capacity for reasoning and action, not to supplant human involvement. To showcase how to use agents working in tandem with developers, we designed a novel approach for carrying out coordinated renaming. Coordinated renaming, where a single rename refactoring triggers refactorings in multiple, related identifiers, is a frequent yet challenging task. Developers must manually propagate these rename refactorings across numerous files and contexts, a process that is both tedious and highly error-prone. State-of-the-art heuristic-based approaches produce an overwhelming number of false positives, while vanilla Large Language Models (LLMs) provide incomplete suggestions due to their limited context and inability to interact with refactoring tools. This leaves developers with incomplete refactorings or burdens them with filtering too many false positives. Coordinated renaming is exactly the kind of repetitive task that agents can significantly reduce the developers' burden while keeping them in the driver's seat. We designed, implemented, and evaluated the first multi-agent framework that automates coordinated renaming. It operates on a key insight: a developer's initial refactoring is a clue to infer the scope of related refactorings. Our Scope Inference Agent first transforms this clue into an explicit, natural-language Declared Scope. The Planned Execution Agent then uses this as a strict plan to identify program elements that should undergo refactoring and safely executes the changes by invoking the IDE's own trusted refactoring APIs. Finally, the Replication Agent uses it to guide the project-wide search. We first conducted a formative study on the practice of coordinated renaming in 609K commits in 100 open-source projects and surveyed 205 developers ...", "authors": ["Abhiram Bellur", "Mohammed Raihan Ullah", "Fraol Batole", "Mohit Kansara", "Masaharu Morimoto", "Kai Ishikawa", "Haifeng Chen", "Yaroslav Zharov", "Timofey Bryksin", "Tien N. Nguyen", "Hridesh Rajan", "Danny Dig"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2026-01-01", "url": "https://arxiv.org/abs/2601.00482", "pdf_url": "https://arxiv.org/pdf/2601.00482v1", "arxiv_id": "2601.00482", "doi": "10.48550/arXiv.2601.00482", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6623}
{"id": "8e84c1754e0092fc5a6f99e70cf4df3437bb8f9cb1f4ba7601d086f339e89c79", "sources": ["arxiv", "semantic_scholar"], "title": "$α^3$-Bench: A Unified Benchmark of Safety, Robustness, and Efficiency for LLM-Based UAV Agents over 6G Networks", "abstract": "Large Language Models (LLMs) are increasingly used as high level controllers for autonomous Unmanned Aerial Vehicle (UAV) missions. However, existing evaluations rarely assess whether such agents remain safe, protocol compliant, and effective under realistic next generation networking constraints. This paper introduces $α^3$-Bench, a benchmark for evaluating LLM driven UAV autonomy as a multi turn conversational reasoning and control problem operating under dynamic 6G conditions. Each mission is formulated as a language mediated control loop between an LLM based UAV agent and a human operator, where decisions must satisfy strict schema validity, mission policies, speaker alternation, and safety constraints while adapting to fluctuating network slices, latency, jitter, packet loss, throughput, and edge load variations. To reflect modern agentic workflows, $α^3$-Bench integrates a dual action layer supporting both tool calls and agent to agent coordination, enabling evaluation of tool use consistency and multi agent interactions. We construct a large scale corpus of 113k conversational UAV episodes grounded in UAVBench scenarios and evaluate 17 state of the art LLMs using a fixed subset of 50 episodes per scenario under deterministic decoding. We propose a composite $α^3$ metric that unifies six pillars: Task Outcome, Safety Policy, Tool Consistency, Interaction Quality, Network Robustness, and Communication Cost, with efficiency normalized scores per second and per thousand tokens. Results show that while several models achieve high mission success and safety compliance, robustness and efficiency vary significantly under degraded 6G conditions, highlighting the need for network aware and resource efficient LLM based UAV agents. The dataset is publicly available on GitHub : https://github.com/maferrag/AlphaBench", "authors": ["Mohamed Amine Ferrag", "Abderrahmane Lakas", "Merouane Debbah"], "categories": ["eess.SY", "cs.AI"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2026-01-01", "url": "https://arxiv.org/abs/2601.03281", "pdf_url": "https://arxiv.org/pdf/2601.03281v1", "arxiv_id": "2601.03281", "doi": "10.48550/arXiv.2601.03281", "citation_count": 6, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/maferrag/AlphaBench", "venue": "arXiv.org", "quality_score": 0.6623}
{"id": "b560dd45231e76aa077429759a5882d4139cc4ca97917cf8793cbaab701f940e", "sources": ["arxiv", "semantic_scholar"], "title": "MCPAgentBench: A Real-world Task Benchmark for Evaluating LLM Agent MCP Tool Use", "abstract": "Large Language Models (LLMs) are increasingly serving as autonomous agents, and their utilization of external tools via the Model Context Protocol (MCP) is considered a future trend. Current MCP evaluation sets suffer from issues such as reliance on external MCP services and a lack of difficulty awareness. To address these limitations, we propose MCPAgentBench, a benchmark based on real-world MCP definitions designed to evaluate the tool-use capabilities of agents. We construct a dataset containing authentic tasks and simulated MCP tools. The evaluation employs a dynamic sandbox environment that presents agents with candidate tool lists containing distractors, thereby testing their tool selection and discrimination abilities. Furthermore, we introduce comprehensive metrics to measure both task completion rates and execution efficiency. Experiments conducted on various latest mainstream Large Language Models reveal significant performance differences in handling complex, multi-step tool invocations. All code is open-source at Github.", "authors": ["Wenrui Liu", "Zixiang Liu", "Elsie Dai", "Wenhan Yu", "Lei Yu", "Tong Yang", "Jinjun Han", "Hong Gao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-31", "url": "https://arxiv.org/abs/2512.24565", "pdf_url": "https://arxiv.org/pdf/2512.24565v3", "arxiv_id": "2512.24565", "doi": "10.48550/arXiv.2512.24565", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6605}
{"id": "e34cc61b5e2a72d27d3ca85c8b54adbb0beffdadad67d7f0bb1302f0ded69d3a", "sources": ["arxiv", "semantic_scholar"], "title": "AMAP Agentic Planning Technical Report", "abstract": "We present STAgent, an agentic large language model tailored for spatio-temporal understanding, designed to solve complex tasks such as constrained point-of-interest discovery and itinerary planning. STAgent is a specialized model capable of interacting with ten distinct tools within spatio-temporal scenarios, enabling it to explore, verify, and refine intermediate steps during complex reasoning. Notably, STAgent effectively preserves its general capabilities. We empower STAgent with these capabilities through three key contributions: (1) a stable tool environment that supports over ten domain-specific tools, enabling asynchronous rollout and training; (2) a hierarchical data curation framework that identifies high-quality data like a needle in a haystack, curating high-quality queries by retaining less than 1\\% of the raw data, emphasizing both diversity and difficulty; and (3) a cascaded training recipe that starts with a seed SFT stage acting as a guardian to measure query difficulty, followed by a second SFT stage fine-tuned on queries with high certainty, and an ultimate RL stage that leverages data of low certainty. Initialized with Qwen3-30B-A3B to establish a strong SFT foundation and leverage insights into sample difficulty, STAgent yields promising performance on TravelBench while maintaining its general capabilities across a wide range of general benchmarks, thereby demonstrating the effectiveness of our proposed agentic model.", "authors": [" AMAP AI Agent Team", "Yulan Hu", "Xiangwen Zhang", "Sheng Ouyang", "Hao Yi", "Lu Xu", "Qinglin Lang", "Lide Tan", "Xiang Cheng", "Tianchen Ye", "Zhicong Li", "Ge Chen", "Wenjin Yang", "Zheng Pan", "Shaopan Xiong", "Siran Yang", "Ju Huang", "Yan Zhang", "Jiamang Wang", "Yong Liu", "Yinfeng Huang", "Ning Wang", "Tucheng Lin", "Xin Li", "Ning Guo"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-31", "url": "https://arxiv.org/abs/2512.24957", "pdf_url": "https://arxiv.org/pdf/2512.24957v2", "arxiv_id": "2512.24957", "doi": "10.48550/arXiv.2512.24957", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4274}
{"id": "0a48fc6f1268fc37921ee808f8225b39c5f864e3cea191f8bbe23a8ebbf7edcc", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcement Learning-Augmented LLM Agents for Collaborative Decision Making and Performance Optimization", "abstract": "Large Language Models (LLMs) perform well in language tasks but often lack collaborative awareness and struggle to optimize global performance in multi-agent settings. We present a reinforcement learning-augmented LLM agent framework that formulates cooperation as a decentralized partially observable Markov decision process (Dec-POMDP) and adopts centralized training with decentralized execution (CTDE). We introduce Group Relative Policy Optimization (GRPO) to jointly optimize agent policies with access to global signals during training, together with a simplified joint reward that balances task quality, speed, and coordination cost. On collaborative writing and coding benchmarks, our framework delivers a 3x increase in task processing speed over single-agent baselines, 98.7% structural/style consistency in writing, and a 74.6% test pass rate in coding. The approach consistently outperforms strong multi-agent LLM baselines and provides a practical path toward reliable collaboration in complex workflows.", "authors": ["Dong Qiu", "Duo Xu", "Limengxi Yue"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-31", "url": "https://arxiv.org/abs/2512.24609", "pdf_url": "https://arxiv.org/pdf/2512.24609v1", "arxiv_id": "2512.24609", "doi": "10.1109/ICFTIC68075.2025.11324888", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.272}
{"id": "add0ab3edf98592eb6987e222b53f2655c42c13953674323b27695c70314483b", "sources": ["arxiv", "semantic_scholar"], "title": "Inter-Agent Relative Representations for Multi-Agent Option Discovery", "abstract": "Temporally extended actions improve the ability to explore and plan in single-agent settings. In multi-agent settings, the exponential growth of the joint state space with the number of agents makes coordinated behaviours even more valuable. Yet, this same exponential growth renders the design of multi-agent options particularly challenging. Existing multi-agent option discovery methods often sacrifice coordination by producing loosely coupled or fully independent behaviours. Toward addressing these limitations, we describe a novel approach for multi-agent option discovery. Specifically, we propose a joint-state abstraction that compresses the state space while preserving the information necessary to discover strongly coordinated behaviours. Our approach builds on the inductive bias that synchronisation over agent states provides a natural foundation for coordination in the absence of explicit objectives. We first approximate a fictitious state of maximal alignment with the team, the \\textit{Fermat} state, and use it to define a measure of \\textit{spreadness}, capturing team-level misalignment on each individual state dimension. Building on this representation, we then employ a neural graph Laplacian estimator to derive options that capture state synchronisation patterns between agents. We evaluate the resulting options across multiple scenarios in two simulated multi-agent domains, showing that they yield stronger downstream coordination capabilities compared to alternative option discovery methods.", "authors": ["Raul D. Steleac", "Mohan Sridharan", "David Abel"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-31", "url": "https://arxiv.org/abs/2512.24827", "pdf_url": "https://arxiv.org/pdf/2512.24827v3", "arxiv_id": "2512.24827", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.272}
{"id": "c9623b017706001c31e818abb5931afb5e5a2ac5f12b8c513a2aeb0c7379acfd", "sources": ["arxiv", "semantic_scholar"], "title": "Group Deliberation Oriented Multi-Agent Conversational Model for Complex Reasoning", "abstract": "This paper proposes a group deliberation oriented multi-agent conversational model to address the limitations of single large language models in complex reasoning tasks. The model adopts a three-level role division architecture consisting of generation, verification, and integration. An opinion generation agent produces diverse reasoning perspectives, an evidence verification agent retrieves external knowledge and quantifies factual support, and a consistency arbitration agent integrates logically coherent conclusions. A self-game mechanism is introduced to expand multi-path reasoning trajectories, while a retrieval enhancement module dynamically supplements external knowledge. A composite reward function combining factual consistency and logical coherence is designed, and an improved proximal policy optimization strategy is applied for collaborative training. Experimental results show that the proposed model improves multi-hop reasoning accuracy by 16.8 percent on HotpotQA, 14.3 percent on 2WikiMultihopQA, and 19.2 percent on MeetingBank, while improving consistency by 21.5 percent. The model achieves higher reasoning efficiency than mainstream multi-agent approaches, providing an effective and stable solution for complex reasoning tasks.", "authors": ["Zheyu Shi", "Dong Qiu", "Shanlong Yu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-31", "url": "https://arxiv.org/abs/2512.24613", "pdf_url": "https://arxiv.org/pdf/2512.24613v1", "arxiv_id": "2512.24613", "doi": "10.1109/IAECST68792.2025.11415036", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.272}
{"id": "5eb4cf6fc7345d6b5c1c074d4c3447482361cd72294b8edb8bd914659f332a98", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLMs for Generating Research Limitations", "abstract": "Identifying and articulating limitations is essential for transparent and rigorous scientific research. However, zero-shot large language models (LLMs) approach often produce superficial or general limitation statements (e.g., dataset bias or generalizability). They usually repeat limitations reported by authors without looking at deeper methodological issues and contextual gaps. This problem is made worse because many authors disclose only partial or trivial limitations. We propose, a multi-agent LLM framework for generating substantive limitations. It integrates OpenReview comments and author-stated limitations to provide stronger ground truth. It also uses cited and citing papers to capture broader contextual weaknesses. In this setup, different agents have specific roles as sequential role: some extract explicit limitations, others analyze methodological gaps, some simulate the viewpoint of a peer reviewer, and a citation agent places the work within the larger body of literature. A Judge agent refines their outputs, and a Master agent consolidates them into a clear set. This structure allows for systematic identification of explicit, implicit, peer review-focused, and literature-informed limitations. Moreover, traditional NLP metrics like BLEU, ROUGE, and cosine similarity rely heavily on n-gram or embedding overlap. They often overlook semantically similar limitations. To address this, we introduce a pointwise evaluation protocol that uses an LLM-as-a-Judge to measure coverage more accurately. Experiments show that our proposed model substantially improve performance. The RAG + multi-agent GPT-4o mini configuration achieves a +15.51\\% coverage gain over zero-shot baselines, while the Llama 3 8B multi-agent setup yields a +4.41\\% improvement.", "authors": ["Ibrahim Al Azher", "Zhishuai Guo", "Hamed Alhoori"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-30", "url": "https://arxiv.org/abs/2601.11578", "pdf_url": "https://arxiv.org/pdf/2601.11578v2", "arxiv_id": "2601.11578", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2712}
{"id": "5888a9b4051dbde04015ca3605c772eded298cd495589a326140a365245abef0", "sources": ["arxiv", "semantic_scholar"], "title": "Close the Loop: Synthesizing Infinite Tool-Use Data via Multi-Agent Role-Playing", "abstract": "Enabling Large Language Models (LLMs) to reliably invoke external tools remains a critical bottleneck for autonomous agents. Existing approaches suffer from three fundamental challenges: expensive human annotation for high-quality trajectories, poor generalization to unseen tools, and quality ceilings inherent in single-model synthesis that perpetuate biases and coverage gaps. We introduce InfTool, a fully autonomous framework that breaks these barriers through self-evolving multi-agent synthesis. Given only raw API specifications, InfTool orchestrates three collaborative agents (User Simulator, Tool-Calling Assistant, and MCP Server) to generate diverse, verified trajectories spanning single-turn calls to complex multi-step workflows. The framework establishes a closed loop: synthesized data trains the model via Group Relative Policy Optimization (GRPO) with gated rewards, the improved model generates higher-quality data targeting capability gaps, and this cycle iterates without human intervention. Experiments on the Berkeley Function-Calling Leaderboard (BFCL) demonstrate that InfTool transforms a base 32B model from 19.8% to 70.9% accuracy (+258%), surpassing models 10x larger and rivaling Claude-Opus, and entirely from synthetic data without human annotation.", "authors": ["Yuwen Li", "Wei Zhang", "Zelong Huang", "Mason Yang", "Jiajun Wu", "Shawn Guo", "Huahao Hu", "Lingyi Sun", "Jian Yang", "Mingjie Tang", "Byran Dai"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-29", "url": "https://arxiv.org/abs/2512.23611", "pdf_url": "https://arxiv.org/pdf/2512.23611v1", "arxiv_id": "2512.23611", "doi": "10.48550/arXiv.2512.23611", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4251}
{"id": "39ec1a669e30bb793ae7ef432f3764d06dfe6fd6c0d477bb62135317442accaa", "sources": ["arxiv", "semantic_scholar"], "title": "ChatGraPhT: A Visual Conversation Interface for Multi-Path Reflection with Agentic LLM Support", "abstract": "Large Language Models (LLMs) are increasingly used in complex knowledge work, yet linear transcript interfaces limit support for reflection. Schon's Reflective Practice distinguishes between reflection-in-action (during a task) and reflection-on-action (after a task), both benefiting from non-linear, revisitable representations of dialogue. ChatGraPhT is an interactive tool that shows dialogue as a visual map, allowing users to branch and merge ideas, edit past messages, and receive guidance that prompts deeper reflection. It supports non-linear, multi-path dialogue, while two agentic LLM assistants provide moment-to-moment and higher-level guidance. Our inquiry suggests that keeping the conversation structure visible, allowing branching and merging, and suggesting patterns or ways to combine ideas deepened user reflective engagement. Contributions are: (1) the design of a node-link, agentic LLM interface for reflective dialogue, and (2) transferable design knowledge on balancing structure and AI support to sustain reflection in complex, open-ended tasks.", "authors": ["Geoff Kimm", "Linus Tan"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-28", "url": "https://arxiv.org/abs/2512.22790", "pdf_url": "https://arxiv.org/pdf/2512.22790v1", "arxiv_id": "2512.22790", "doi": "10.48550/arXiv.2512.22790", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.424}
{"id": "17a385336ba5987d2193b78cb8993c636fde62f56b03491726443d67797f493b", "sources": ["arxiv", "semantic_scholar"], "title": "Heterogeneity in Multi-Agent Reinforcement Learning", "abstract": "Heterogeneity is a fundamental property in multi-agent reinforcement learning (MARL), which is closely related not only to the functional differences of agents, but also to policy diversity and environmental interactions. However, the MARL field currently lacks a rigorous definition and deeper understanding of heterogeneity. This paper systematically discusses heterogeneity in MARL from the perspectives of definition, quantification, and utilization. First, based on an agent-level modeling of MARL, we categorize heterogeneity into five types and provide mathematical definitions. Second, we define the concept of heterogeneity distance and propose a practical quantification method. Third, we design a heterogeneity-based multi-agent dynamic parameter sharing algorithm as an example of the application of our methodology. Case studies demonstrate that our method can effectively identify and quantify various types of agent heterogeneity. Experimental results show that the proposed algorithm, compared to other parameter sharing baselines, has better interpretability and stronger adaptability. The proposed methodology will help the MARL community gain a more comprehensive and profound understanding of heterogeneity, and further promote the development of practical algorithms.", "authors": ["Tianyi Hu", "Zhiqiang Pu", "Yuan Wang", "Tenghai Qiu", "Min Chen", "Xin Yu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-28", "url": "https://arxiv.org/abs/2512.22941", "pdf_url": "https://arxiv.org/pdf/2512.22941v1", "arxiv_id": "2512.22941", "doi": "10.48550/arXiv.2512.22941", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2698}
{"id": "ca8d2d438519ebf4177576d9fbbf30689ae6afaaf6845983b4258ab450b80b8d", "sources": ["arxiv", "semantic_scholar"], "title": "A Plan Reuse Mechanism for LLM-Driven Agent", "abstract": "Integrating large language models (LLMs) into personal assistants, like Xiao Ai and Blue Heart V, effectively enhances their ability to interact with humans, solve complex tasks, and manage IoT devices. Such assistants are also termed LLM-driven agents. Upon receiving user requests, the LLM-driven agent generates plans using an LLM, executes these plans through various tools, and then returns the response to the user. During this process, the latency for generating a plan with an LLM can reach tens of seconds, significantly degrading user experience. Real-world dataset analysis shows that about 30% of the requests received by LLM-driven agents are identical or similar, which allows the reuse of previously generated plans to reduce latency. However, it is difficult to accurately define the similarity between the request texts received by the LLM-driven agent through directly evaluating the original request texts. Moreover, the diverse expressions of natural language and the unstructured format of plan texts make implementing plan reuse challenging. To address these issues, we present and implement a plan reuse mechanism for LLM-driven agents called AgentReuse. AgentReuse leverages the similarities and differences among requests' semantics and uses intent classification to evaluate the similarities between requests and enable the reuse of plans. Experimental results based on a real-world dataset demonstrate that AgentReuse achieves a 93% effective plan reuse rate, an F1 score of 0.9718, and an accuracy of 0.9459 in evaluating request similarities, reducing latency by 93.12% compared with baselines without using the reuse mechanism.", "authors": ["Guopeng Li", "Ruiqi Wu", "Haisheng Tan"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-24", "url": "https://arxiv.org/abs/2512.21309", "pdf_url": "https://arxiv.org/pdf/2512.21309v2", "arxiv_id": "2512.21309", "doi": "10.48550/arXiv.2512.21309", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4194}
{"id": "8e010514472fb77525cb58e92a260036686c1e5106c6b9c9702a78f49964a129", "sources": ["arxiv", "semantic_scholar"], "title": "DAO-Agent: Zero Knowledge-Verified Incentives for Decentralized Multi-Agent Coordination", "abstract": "Autonomous Large Language Model (LLM)-based multi-agent systems have emerged as a promising paradigm for facilitating cross-application and cross-organization collaborations. These autonomous agents often operate in trustless environments, where centralized coordination faces significant challenges, such as the inability to ensure transparent contribution measurement and equitable incentive distribution. While blockchain is frequently proposed as a decentralized coordination platform, it inherently introduces high on-chain computation costs and risks exposing sensitive execution information of the agents. Consequently, the core challenge lies in enabling auditable task execution and fair incentive distribution for autonomous LLM agents in trustless environments, while simultaneously preserving their strategic privacy and minimizing on-chain costs. To address this challenge, we propose DAO-Agent, a novel framework that integrates three key technical innovations: (1) an on-chain decentralized autonomous organization (DAO) governance mechanism for transparent coordination and immutable logging; (2) a ZKP mechanism approach that enables Shapley-based contribution measurement off-chain, and (3) a hybrid on-chain/off-chain architecture that verifies ZKP-validated contribution measurements on-chain with minimal computational overhead. We implement DAO-Agent and conduct end-to-end experiments using a crypto trading task as a case study. Experimental results demonstrate that DAO-Agent achieves up to 99.9% reduction in verification gas costs compared to naive on-chain alternatives, with constant-time verification complexity that remains stable as coalition size increases, thereby establishing a scalable foundation for agent coordination in decentralized environments.", "authors": ["Yihan Xia", "Taotao Wang", "Wenxin Xu", "Shengli Zhang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-24", "url": "https://arxiv.org/abs/2512.20973", "pdf_url": "https://arxiv.org/pdf/2512.20973v1", "arxiv_id": "2512.20973", "doi": "10.48550/arXiv.2512.20973", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4194}
{"id": "70217b41b632e702bd6c13ca89bf3bfccb5f2c09386e6ed0077c03a3f037cbd0", "sources": ["arxiv", "semantic_scholar"], "title": "LongVideoAgent: Multi-Agent Reasoning with Long Videos", "abstract": "Recent advances in multimodal LLMs and systems that use tools for long-video QA point to the promise of reasoning over hour-long episodes. However, many methods still compress content into lossy summaries or rely on limited toolsets, weakening temporal grounding and missing fine-grained cues. We propose a multi-agent framework in which a master LLM coordinates a grounding agent to localize question-relevant segments and a vision agent to extract targeted textual observations. The master agent plans with a step limit, and is trained with reinforcement learning to encourage concise, correct, and efficient multi-agent cooperation. This design helps the master agent focus on relevant clips via grounding, complements subtitles with visual detail, and yields interpretable trajectories. On our proposed LongTVQA and LongTVQA+ which are episode-level datasets aggregated from TVQA/TVQA+, our multi-agent system significantly outperforms strong non-agent baselines. Experiments also show reinforcement learning further strengthens reasoning and planning for the trained agent. Code and data will be shared at https://longvideoagent.github.io/.", "authors": ["Runtao Liu", "Ziyi Liu", "Jiaqi Tang", "Yue Ma", "Renjie Pi", "Jipeng Zhang", "Qifeng Chen"], "categories": ["cs.AI", "cs.CV", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-23", "url": "https://arxiv.org/abs/2512.20618", "pdf_url": "https://arxiv.org/pdf/2512.20618v1", "arxiv_id": "2512.20618", "doi": "10.48550/arXiv.2512.20618", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4182}
{"id": "40fcd51782dd9444bfd2d98f6a48c28caca099e3ea0f3751b3e408f6fc4d9b76", "sources": ["arxiv", "semantic_scholar"], "title": "MAR:Multi-Agent Reflexion Improves Reasoning Abilities in LLMs", "abstract": "LLMs have shown the capacity to improve their performance on reasoning tasks through reflecting on their mistakes, and acting with these reflections in mind. However, continual reflections of the same LLM onto itself exhibit degeneration of thought, where the LLM continues to repeat the same errors again and again even with the knowledge that its wrong. To address this problem, we instead introduce multi-agent with multi-persona debators as the method to generate reflections. Through out extensive experimentation, we've found that the leads to better diversity of in the reflections generated by the llm agent. We demonstrate an accuracy of 47% EM HotPot QA (question answering) and 82.7% on HumanEval (programming), both performances surpassing reflection with a single llm.", "authors": ["Onat Ozer", "Yuchen Wang", "Grace Wu", "Daniel Dosti", "Honghao Zhang", "Vivi De La Rue"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-23", "url": "https://arxiv.org/abs/2512.20845", "pdf_url": "https://arxiv.org/pdf/2512.20845v2", "arxiv_id": "2512.20845", "doi": "10.48550/arXiv.2512.20845", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4182}
{"id": "93f787491f560405776d4bcd5a960f4e78a5cd041ef296d986d6f16c25ca6d23", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Hierarchical Procedural Memory for LLM Agents through Bayesian Selection and Contrastive Refinement", "abstract": "We present MACLA, a framework that decouples reasoning from learning by maintaining a frozen large language model while performing all adaptation in an external hierarchical procedural memory. MACLA extracts reusable procedures from trajectories, tracks reliability via Bayesian posteriors, selects actions through expected-utility scoring, and refines procedures by contrasting successes and failures. Across four benchmarks (ALFWorld, WebShop, TravelPlanner, InterCodeSQL), MACLA achieves 78.1 percent average performance, outperforming all baselines. On ALFWorld unseen tasks, MACLA reaches 90.3 percent with 3.1 percent positive generalization. The system constructs memory in 56 seconds, 2800 times faster than the state-of-the-art LLM parameter-training baseline, compressing 2851 trajectories into 187 procedures. Experimental results demonstrate that structured external memory with Bayesian selection and contrastive refinement enables sample-efficient, interpretable, and continually improving agents without LLM parameter updates.", "authors": ["Saman Forouzandeh", "Wei Peng", "Parham Moradi", "Xinghuo Yu", "Mahdi Jalili"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-22", "url": "https://arxiv.org/abs/2512.18950", "pdf_url": "https://arxiv.org/pdf/2512.18950v1", "arxiv_id": "2512.18950", "doi": "10.48550/arXiv.2512.18950", "citation_count": 10, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/S-Forouzandeh/MACLA-LLM-Agents-AAMAS-Conference", "venue": null, "quality_score": 0.4929}
{"id": "b00178561d019468609e9fb8e7af4210609aed520d469ded87aea98383a222c0", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Retrieval-Augmented Framework for Work-in-Progress Predictio", "abstract": "Work-in-Progress (WiP) prediction is critical for predictive process monitoring, enabling accurate anticipation of workload fluctuations and optimized operational planning. This paper proposes a retrieval-augmented, multi-agent framework that combines retrieval-augmented generation (RAG) and collaborative multi-agent reasoning for WiP prediction. The narrative generation component transforms structured event logs into semantically rich natural language stories, which are embedded into a semantic vector-based process memory to facilitate dynamic retrieval of historical context during inference. The framework includes predictor agents that independently leverage retrieved historical contexts and a decision-making assistant agent that extracts high-level descriptive signals from recent events. A fusion agent then synthesizes predictions using ReAct-style reasoning over agent outputs and retrieved narratives. We evaluate our framework on two real-world benchmark datasets. Results show that the proposed retrieval-augmented multi-agent approach achieves competitive prediction accuracy, obtaining a Mean Absolute Percentage Error (MAPE) of 1.50\\% on one dataset, and surpassing Temporal Convolutional Networks (TCN), Long Short-Term Memory (LSTM), and persistence baselines. The results highlight improved robustness, demonstrating the effectiveness of integrating retrieval mechanisms and multi-agent reasoning in WiP prediction.", "authors": ["Yousef Mehrdad Bibalan", "Behrouz Far", "Mohammad Moshirpour", "Bahareh Ghiyasian"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-22", "url": "https://arxiv.org/abs/2512.19841", "pdf_url": "https://arxiv.org/pdf/2512.19841v1", "arxiv_id": "2512.19841", "doi": "10.5121/csit.2025.152409", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2654}
{"id": "dd46657d3f1d2a21e75e813033c948fffa5aed9e6d0cf87ce8642c80e251f62d", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLM Committees for Autonomous Software Beta Testing", "abstract": "Manual software beta testing is costly and time-consuming, while single-agent large language model (LLM) approaches suffer from hallucinations and inconsistent behavior. We propose a multi-agent committee framework in which diverse vision-enabled LLMs collaborate through a three-round voting protocol to reach consensus on testing actions. The framework combines model diversity, persona-driven behavioral variation, and visual user interface understanding to systematically explore web applications. Across 84 experimental runs with 9 testing personas and 4 scenarios, multi-agent committees achieve an 89.5 percent overall task success rate. Configurations with 2 to 4 agents reach 91.7 to 100 percent success, compared to 78.0 percent for single-agent baselines, yielding improvements of 13.7 to 22.0 percentage points. At the action level, the system attains a 93.1 percent success rate with a median per-action latency of 0.71 seconds, enabling real-time and continuous integration testing. Vision-enabled agents successfully identify user interface elements, with navigation and reporting achieving 100 percent success and form filling achieving 99.2 percent success. We evaluate the framework on WebShop and OWASP benchmarks, achieving 74.7 percent success on WebShop compared to a 50.1 percent published GPT-3 baseline, and 82.0 percent success on OWASP Juice Shop security testing with coverage of 8 of the 10 OWASP Top 10 vulnerability categories. Across 20 injected regressions, the committee achieves an F1 score of 0.91 for bug detection, compared to 0.78 for single-agent baselines. The open-source implementation enables reproducible research and practical deployment of LLM-based software testing in CI/CD pipelines.", "authors": ["Sumanth Bharadwaj Hachalli Karanam", "Dhiwahar Adhithya Kennady"], "categories": ["cs.SE", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-21", "url": "https://arxiv.org/abs/2512.21352", "pdf_url": "https://arxiv.org/pdf/2512.21352v1", "arxiv_id": "2512.21352", "doi": "10.48550/arXiv.2512.21352", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6428}
{"id": "86ab4426667f830d736f0449d71e5653b8d082a0e9ce09b395b8b256381c26e6", "sources": ["arxiv", "semantic_scholar"], "title": "IntelliCode: A Multi-Agent LLM Tutoring System with Centralized Learner Modeling", "abstract": "LLM-based tutors are typically single-turn assistants that lack persistent representations of learner knowledge, making it difficult to provide principled, transparent, and long-term pedagogical support. We introduce IntelliCode, a multi-agent LLM tutoring system built around a centralized, versioned learner state that integrates mastery estimates, misconceptions, review schedules, and engagement signals. A StateGraph Orchestrator coordinates six specialized agents: skill assessment, learner profiling, graduated hinting, curriculum selection, spaced repetition, and engagement monitoring, each operating as a pure transformation over the shared state under a single-writer policy. This architecture enables auditable mastery updates, proficiency-aware hints, dependency-aware curriculum adaptation, and safety-aligned prompting. The demo showcases an end-to-end tutoring workflow: a learner attempts a DSA problem, receives a conceptual hint when stuck, submits a corrected solution, and immediately sees mastery updates and a personalized review interval. We report validation results with simulated learners, showing stable state updates, improved task success with graduated hints, and diverse curriculum coverage. IntelliCode demonstrates how persistent learner modeling, orchestrated multi-agent reasoning, and principled instructional design can be combined to produce transparent and reliable LLM-driven tutoring.", "authors": ["Jones David", "Shreya Ghosh"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-21", "url": "https://arxiv.org/abs/2512.18669", "pdf_url": "https://arxiv.org/pdf/2512.18669v1", "arxiv_id": "2512.18669", "doi": "10.48550/arXiv.2512.18669", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2647}
{"id": "b2a1d0c60751ae9aaf5fd7d4ffa6c8d0aabb85a64c3c9464ae7eb6532b61d34f", "sources": ["arxiv", "semantic_scholar"], "title": "Explainable and Fine-Grained Safeguarding of LLM Multi-Agent Systems via Bi-Level Graph Anomaly Detection", "abstract": "Large language model (LLM)-based multi-agent systems (MAS) have shown strong capabilities in solving complex tasks. As MAS become increasingly autonomous in various safety-critical tasks, detecting malicious agents has become a critical security concern. Although existing graph anomaly detection (GAD)-based defenses can identify anomalous agents, they mainly rely on coarse sentence-level information and overlook fine-grained lexical cues, leading to suboptimal performance. Moreover, the lack of interpretability in these methods limits their reliability and real-world applicability. To address these limitations, we propose XG-Guard, an explainable and fine-grained safeguarding framework for detecting malicious agents in MAS. To incorporate both coarse and fine-grained textual information for anomalous agent identification, we utilize a bi-level agent encoder to jointly model the sentence- and token-level representations of each agent. A theme-based anomaly detector further captures the evolving discussion focus in MAS dialogues, while a bi-level score fusion mechanism quantifies token-level contributions for explanation. Extensive experiments across diverse MAS topologies and attack scenarios demonstrate robust detection performance and strong interpretability of XG-Guard.", "authors": ["Junjun Pan", "Yixin Liu", "Rui Miao", "Kaize Ding", "Yu Zheng", "Quoc Viet Hung Nguyen", "Alan Wee-Chung Liew", "Shirui Pan"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-21", "url": "https://arxiv.org/abs/2512.18733", "pdf_url": "https://arxiv.org/pdf/2512.18733v1", "arxiv_id": "2512.18733", "doi": "10.48550/arXiv.2512.18733", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4159}
{"id": "195934b7b2f86f4049bbe3347884a5b333421d3cfd8e5b131bde48ee8005f9f8", "sources": ["arxiv", "semantic_scholar"], "title": "XAgen: An Explainability Tool for Identifying and Correcting Failures in Multi-Agent Workflows", "abstract": "As multi-agent systems powered by Large Language Models (LLMs) are increasingly adopted in real-world workflows, users with diverse technical backgrounds are now building and refining their own agentic processes. However, these systems can fail in opaque ways, making it difficult for users to observe, understand, and correct errors. We conducted formative interviews with 12 practitioners to identify mismatches between existing debugging tools and users' needs. Based on these insights, we designed XAgen, an explainability tool that supports users with varying AI expertise through three core capabilities: log visualization for glanceable workflow understanding, human-in-the-loop feedback to capture expert judgment, and automatic error detection via an LLM-as-a-judge. In a user study with 8 participants, XAgen helped users locate failures more easily, attribute to specific agents or steps, and iteratively improve configurations. Our findings surface human-centered design guidelines for explainable agentic AI development and highlight opportunities for more context-aware interactive debugging.", "authors": ["Xinru Wang", "Ming Yin", "Eunyee Koh", "Mustafa Doga Dogan"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-19", "url": "https://arxiv.org/abs/2512.17896", "pdf_url": "https://arxiv.org/pdf/2512.17896v2", "arxiv_id": "2512.17896", "doi": "10.1145/3772363.3798531", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2632}
{"id": "5f46285aa6a2f92448b50cade74f47ffa29b2f38676ca28fc32ee7bb2e949ba1", "sources": ["arxiv", "semantic_scholar"], "title": "Verifiability-First Agents: Provable Observability and Lightweight Audit Agents for Controlling Autonomous LLM Systems", "abstract": "As LLM-based agents grow more autonomous and multi-modal, ensuring they remain controllable, auditable, and faithful to deployer intent becomes critical. Prior benchmarks measured the propensity for misaligned behavior and showed that agent personalities and tool access significantly influence misalignment. Building on these insights, we propose a Verifiability-First architecture that (1) integrates run-time attestations of agent actions using cryptographic and symbolic methods, (2) embeds lightweight Audit Agents that continuously verify intent versus behavior using constrained reasoning, and (3) enforces challenge-response attestation protocols for high-risk operations. We introduce OPERA (Observability, Provable Execution, Red-team, Attestation), a benchmark suite and evaluation protocol designed to measure (i) detectability of misalignment, (ii) time to detection under stealthy strategies, and (iii) resilience of verifiability mechanisms to adversarial prompt and persona injection. Our approach shifts the evaluation focus from how likely misalignment is to how quickly and reliably misalignment can be detected and remediated.", "authors": ["Abhivansh Gupta"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-19", "url": "https://arxiv.org/abs/2512.17259", "pdf_url": "https://arxiv.org/pdf/2512.17259v1", "arxiv_id": "2512.17259", "doi": "10.48550/arXiv.2512.17259", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4136}
{"id": "66bd88c3d1fcec817c625f2eaa7733d21e81a728cd4af9109de21d64acb76609", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic Tool Dependency Retrieval for Lightweight Function Calling", "abstract": "Function calling agents powered by Large Language Models (LLMs) select external tools to automate complex tasks. On-device agents typically use a retrieval module to select relevant tools, improving performance and reducing context length. However, existing retrieval methods rely on static and limited inputs, failing to capture multi-step tool dependencies and evolving task context. This limitation often introduces irrelevant tools that mislead the agent, degrading efficiency and accuracy. We propose Dynamic Tool Dependency Retrieval (DTDR), a lightweight retrieval method that conditions on both the initial query and the evolving tool calling plan. DTDR models tool dependencies from function calling demonstrations, enabling adaptive retrieval as plans unfold. We benchmark DTDR against state-of-the-art retrieval methods across multiple datasets and LLM backbones, evaluating retrieval precision, downstream task accuracy, and computational efficiency. Additionally, we explore strategies to integrate retrieved tools into prompts. Our results show that DTDR improves function calling success rates between $23\\%$ and $104\\%$ compared to state-of-the-art static retrievers.", "authors": ["Bhrij Patel", "Davide Belli", "Amir Jalalirad", "Maximilian Arnold", "Aleksandr Ermolov", "Bence Major"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-18", "url": "https://arxiv.org/abs/2512.17052", "pdf_url": "https://arxiv.org/pdf/2512.17052v4", "arxiv_id": "2512.17052", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2625}
{"id": "167c73b2754bcf98d9f9fd8d582633665e5b2fc7d81cd8fcc1fbcc147cb4416e", "sources": ["arxiv", "semantic_scholar"], "title": "PDE-Agent: A toolchain-augmented multi-agent framework for PDE solving", "abstract": "Solving Partial Differential Equations (PDEs) is a cornerstone of engineering and scientific research. Traditional methods for PDE solving are cumbersome, relying on manual setup and domain expertise. While Physics-Informed Neural Network (PINNs) introduced end-to-end neural network-based solutions, and frameworks like DeepXDE further enhanced automation, these approaches still depend on expert knowledge and lack full autonomy. In this work, we frame PDE solving as tool invocation via LLM-driven agents and introduce PDE-Agent, the first toolchain-augmented multi-agent collaboration framework, inheriting the reasoning capacity of LLMs and the controllability of external tools and enabling automated PDE solving from natural language descriptions. PDE-Agent leverages the strengths of multi-agent and multi-tool collaboration through two key innovations: (1) A Prog-Act framework with graph memory for multi-agent collaboration, which enables effective dynamic planning and error correction via dual-loop mechanisms (localized fixes and global revisions). (2) A Resource-Pool integrated with a tool-parameter separation mechanism for multi-tool collaboration. This centralizes the management of runtime artifacts and resolves inter-tool dependency gaps in existing frameworks. To validate and evaluate this new paradigm for PDE solving , we develop PDE-Bench, a multi-type PDE Benchmark for agent-based tool collaborative solving, and propose multi-level metrics for assessing tool coordination. Evaluations verify that PDE-Agent exhibits superior applicability and performance in complex multi-step, cross-step dependent tasks. This new paradigm of toolchain-augmented multi-agent PDE solving will further advance future developments in automated scientific computing. Our source code and dataset will be made publicly available.", "authors": ["Jianming Liu", "Ren Zhu", "Jian Xu", "Kun Ding", "Xu-Yao Zhang", "Gaofeng Meng", "Cheng-Lin Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-18", "url": "https://arxiv.org/abs/2512.16214", "pdf_url": "https://arxiv.org/pdf/2512.16214v2", "arxiv_id": "2512.16214", "doi": "10.48550/arXiv.2512.16214", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4125}
{"id": "2041d185786238e9da68560f1351a8bceeba275443e68ac17e5ed8964629bb7e", "sources": ["arxiv", "semantic_scholar"], "title": "On the Role of Contextual Information and Ego States in LLM Agent Behavior for Transactional Analysis Dialogues", "abstract": "LLM-powered agents are now used in many areas, from customer support to education, and there is increasing interest in their ability to act more like humans. This includes fields such as social, political, and psychological research, where the goal is to model group dynamics and social behavior. However, current LLM agents often lack the psychological depth and consistency needed to capture the real patterns of human thinking. They usually provide direct or statistically likely answers, but they miss the deeper goals, emotional conflicts, and motivations that drive real human interactions. This paper proposes a Multi-Agent System (MAS) inspired by Transactional Analysis (TA) theory. In the proposed system, each agent is divided into three ego states - Parent, Adult, and Child. The ego states are treated as separate knowledge structures with their own perspectives and reasoning styles. To enrich their response process, they have access to an information retrieval mechanism that allows them to retrieve relevant contextual information from their vector stores. This architecture is evaluated through ablation tests in a simulated dialogue scenario, comparing agents with and without information retrieval. The results are promising and open up new directions for exploring how psychologically grounded structures can enrich agent behavior. The contribution is an agent architecture that integrates Transactional Analysis theory with contextual information retrieval to enhance the realism of LLM-based multi-agent simulations.", "authors": ["Monika Zamojska", "Jarosław A. Chudziak"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-18", "url": "https://arxiv.org/abs/2512.17060", "pdf_url": "https://arxiv.org/pdf/2512.17060v1", "arxiv_id": "2512.17060", "doi": "10.48550/arXiv.2512.17060", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4125}
{"id": "abfc1f3e19ec2ae18145c0f1c71bf2164c87d0adc0d65fadc0dde6c6a3bfa55f", "sources": ["arxiv", "semantic_scholar"], "title": "Ev-Trust: An Evolutionarily Stable Trust Mechanism for Decentralized LLM-Based Multi-Agent Service Economies", "abstract": "Decentralized LLM-based multi-agent service economies face three vulnerabilities that undermine traditional trust mechanisms: reduced cost of fraud, difficulty in evaluating service quality, and instability of service content. These compounding vulnerabilities can trigger population-level trust collapse and the proliferation of short-sighted strategies. We propose Ev-Trust, an evolutionarily stable trust mechanism that addresses these vulnerabilities through three targeted designs: a cross-validation gate leveraging requestor semantic comprehension to assess response validity, a variance-standardized drift measure filtering endogenous stochasticity from genuine behavioral anomalies, and an embedding of trust signals into the expected revenue function that converts trustworthiness into an evolutionary survival advantage. Based on replicator dynamics with a noisy best response micro-foundation, we prove the asymptotic stability of cooperative evolutionarily stable strategies and derive explicit threshold conditions for maintaining cooperative equilibria. We evaluate Ev-Trust through 100-round simulations with at least 100 heterogeneous LLM-driven agents covering seven behavioral types. The experiments are conducted on TruthfulQA and TriviaQA, two factual question-answering benchmarks. Compared to baselines based on transitive trust aggregation, reinforcement-learning reputation, and pure evolutionary imitation, Ev-Trust reduces malicious agent participation by approximately 60%, suppresses the fraudulent service rate by approximately 50%, and maintains stable trust differentiation under a 30% adversarial mutation. These results demonstrate that coupling semantic trust evaluation with evolutionary incentives provides a principled foundation for securing cooperation in decentralized LLM-based multi-agent systems.", "authors": ["Jiye Wang", "Shiduo Yang", "Ting Qiao", "Jiayu Qin", "Jianbin Li", "Yu Wang", "Yuanhe Zhao"], "categories": ["cs.MA", "cs.AI", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-18", "url": "https://arxiv.org/abs/2512.16167", "pdf_url": "https://arxiv.org/pdf/2512.16167v3", "arxiv_id": "2512.16167", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2625}
{"id": "74df30e91f10b6a8c0d284f0331cf00d24e9f3cf8a0186ca09215c24faf0b9bc", "sources": ["arxiv", "semantic_scholar"], "title": "SynthSeg-Agents: Multi-Agent Synthetic Data Generation for Zero-Shot Weakly Supervised Semantic Segmentation", "abstract": "Weakly Supervised Semantic Segmentation (WSSS) with image level labels aims to produce pixel level predictions without requiring dense annotations. While recent approaches have leveraged generative models to augment existing data, they remain dependent on real world training samples. In this paper, we introduce a novel direction, Zero Shot Weakly Supervised Semantic Segmentation (ZSWSSS), and propose SynthSeg Agents, a multi agent framework driven by Large Language Models (LLMs) to generate synthetic training data entirely without real images. SynthSeg Agents comprises two key modules, a Self Refine Prompt Agent and an Image Generation Agent. The Self Refine Prompt Agent autonomously crafts diverse and semantically rich image prompts via iterative refinement, memory mechanisms, and prompt space exploration, guided by CLIP based similarity and nearest neighbor diversity filtering. These prompts are then passed to the Image Generation Agent, which leverages Vision Language Models (VLMs) to synthesize candidate images. A frozen CLIP scoring model is employed to select high quality samples, and a ViT based classifier is further trained to relabel the entire synthetic dataset with improved semantic precision. Our framework produces high quality training data without any real image supervision. Experiments on PASCAL VOC 2012 and COCO 2014 show that SynthSeg Agents achieves competitive performance without using real training images. This highlights the potential of LLM driven agents in enabling cost efficient and scalable semantic segmentation.", "authors": ["Wangyu Wu", "Zhenhong Chen", "Xiaowei Huang", "Fei Ma", "Jimin Xiao"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-17", "url": "https://arxiv.org/abs/2512.15310", "pdf_url": "https://arxiv.org/pdf/2512.15310v1", "arxiv_id": "2512.15310", "doi": "10.48550/arXiv.2512.15310", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4114}
{"id": "52e052028049dd917b1ad729ce7c9b2e99b5142c5fa3be403aee9d25af760cac", "sources": ["arxiv", "semantic_scholar"], "title": "Mapis: A Knowledge-Graph Grounded Multi-Agent Framework for Evidence-Based PCOS Diagnosis", "abstract": "Polycystic Ovary Syndrome (PCOS) constitutes a significant public health issue affecting 10% of reproductive-aged women, highlighting the critical importance of developing effective diagnostic tools. Previous machine learning and deep learning detection tools are constrained by their reliance on large-scale labeled data and an lack of interpretability. Although multi-agent systems have demonstrated robust capabilities, the potential of such systems for PCOS detection remains largely unexplored. Existing medical multi-agent frameworks are predominantly designed for general medical tasks, suffering from insufficient domain integration and a lack of specific domain knowledge. To address these challenges, we propose Mapis, the first knowledge-grounded multi-agent framework explicitly designed for guideline-based PCOS diagnosis. Specifically, it built upon the 2023 International Guideline into a structured collaborative workflow that simulates the clinical diagnostic process. It decouples complex diagnostic tasks across specialized agents: a gynecological endocrine agent and a radiology agent collaborative to verify inclusion criteria, while an exclusion agent strictly rules out other causes. Furthermore, we construct a comprehensive PCOS knowledge graph to ensure verifiable, evidence-based decision-making. Extensive experiments on public benchmarks and specialized clinical datasets, benchmarking against nine diverse baselines, demonstrate that Mapis significantly outperforms competitive methods. On the clinical dataset, it surpasses traditional machine learning models by 13.56%, single-agent by 6.55%, and previous medical multi-agent systems by 7.05% in Accuracy.", "authors": ["Zanxiang He", "Meng Li", "Liyun Shi", "Weiye Daia", "Liming Nie"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-17", "url": "https://arxiv.org/abs/2512.15398", "pdf_url": "https://arxiv.org/pdf/2512.15398v1", "arxiv_id": "2512.15398", "doi": "10.48550/arXiv.2512.15398", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4114}
{"id": "ca609dfa54e8f393a5d3679b77092f4105bb7fc170c83ad02fa2ff289bc96be4", "sources": ["arxiv", "semantic_scholar"], "title": "MALCDF: A Distributed Multi-Agent LLM Framework for Real-Time Cyber", "abstract": "Traditional, centralized security tools often miss adaptive, multi-vector attacks. We present the Multi-Agent LLM Cyber Defense Framework (MALCDF), a practical setup where four large language model (LLM) agents-Detection, Intelligence, Response, and Analysis-work together in real time. Agents communicate over a Secure Communication Layer (SCL) with encrypted, ontology-aligned messages, and produce audit-friendly outputs (e.g., MITRE ATT&CK mappings). For evaluation, we keep the test simple and consistent: all reported metrics come from the same 50-record live stream derived from the CICIDS2017 feature schema. CICIDS2017 is used for configuration (fields/schema) and to train a practical ML baseline. The ML-IDS baseline is a Lightweight Random Forest IDS (LRF-IDS) trained on a subset of CICIDS2017 and tested on the 50-record stream, with no overlap between training and test records. In experiments, MALCDF reaches 90.0% detection accuracy, 85.7% F1-score, and 9.1% false-positive rate, with 6.8s average per-event latency. It outperforms the lightweight ML-IDS baseline and a single-LLM setup on accuracy while keeping end-to-end outputs consistent. Overall, this hands-on build suggests that coordinating simple LLM agents with secure, ontology-aligned messaging can improve practical, real-time cyber defense.", "authors": ["Arth Bhardwaj", "Sia Godika", "Yuvam Loonker"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-16", "url": "https://arxiv.org/abs/2512.14846", "pdf_url": "https://arxiv.org/pdf/2512.14846v1", "arxiv_id": "2512.14846", "doi": "10.48550/arXiv.2512.14846", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4102}
{"id": "8da19c290292d4d499cb419b88fd2fcadf406fd99c687ad4f0bfcafaccd0c739", "sources": ["arxiv", "semantic_scholar"], "title": "Grammar Search for Multi-Agent Systems", "abstract": "Automatic search for Multi-Agent Systems has recently emerged as a key focus in agentic AI research. Several prior approaches have relied on LLM-based free-form search over the code space. In this work, we propose a more structured framework that explores the same space through a fixed set of simple, composable components. We show that, despite lacking the generative flexibility of LLMs during the candidate generation stage, our method outperforms prior approaches on four out of five benchmarks across two domains: mathematics and question answering. Furthermore, our method offers additional advantages, including a more cost-efficient search process and the generation of modular, interpretable multi-agent systems with simpler logic.", "authors": ["Mayank Singh", "Vikas Yadav", "Shiva Krishna Reddy Malay", "Shravan Nayak", "Sai Rajeswar", "Sathwik Tejaswi Madhusudhan", "Eduardo Blanco"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-16", "url": "https://arxiv.org/abs/2512.14079", "pdf_url": "https://arxiv.org/pdf/2512.14079v1", "arxiv_id": "2512.14079", "doi": "10.48550/arXiv.2512.14079", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4102}
{"id": "8645c998f11c7ff6e589e304a3188e791e60e3ca66ddc8f9bacd120a93e1f50f", "sources": ["arxiv", "semantic_scholar"], "title": "Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Functional Materials Discovery", "abstract": "Artificial intelligence is reshaping scientific exploration, but most methods automate procedural tasks without engaging in scientific reasoning, limiting autonomy in discovery. We introduce Materials Agents for Simulation and Theory in Electronic-structure Reasoning (MASTER), an active learning framework where large language models autonomously design, execute, and interpret atomistic simulations. In MASTER, a multimodal system translates natural language into density functional theory workflows, while higher-level reasoning agents guide discovery through a hierarchy of strategies, including a single agent baseline and three multi-agent approaches: peer review, triage-ranking, and triage-forms. Across two chemical applications, CO adsorption on Cu-surface transition metal (M) adatoms and on M-N-C catalysts, reasoning-driven exploration reduces required atomistic simulations by up to 90% relative to trial-and-error selection. Reasoning trajectories reveal chemically grounded decisions that cannot be explained by stochastic sampling or semantic bias. Altogether, multi-agent collaboration accelerates materials discovery and marks a new paradigm for autonomous scientific exploration.", "authors": ["Samuel Rothfarb", "Megan C. Davis", "Ivana Matanovic", "Baikun Li", "Edward F. Holby", "Wilton J. M. Kort-Kamp"], "categories": ["cond-mat.mtrl-sci", "cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2025-12-15", "url": "https://arxiv.org/abs/2512.13930", "pdf_url": "https://arxiv.org/pdf/2512.13930v1", "arxiv_id": "2512.13930", "doi": "10.48550/arXiv.2512.13930", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "f9c74bc1cb8998fc3d9a0256325727e7c335ef0335c307a7469f4ad74488e651", "sources": ["arxiv", "semantic_scholar"], "title": "Bilevel Optimization for Covert Memory Tampering in Heterogeneous Multi-Agent Architectures (XAMT)", "abstract": "The increasing operational reliance on complex Multi-Agent Systems (MAS) across safety-critical domains necessitates rigorous adversarial robustness assessment. Modern MAS are inherently heterogeneous, integrating conventional Multi-Agent Reinforcement Learning (MARL) with emerging Large Language Model (LLM) agent architectures utilizing Retrieval-Augmented Generation (RAG). A critical shared vulnerability is reliance on centralized memory components: the shared Experience Replay (ER) buffer in MARL and the external Knowledge Base (K) in RAG agents. This paper proposes XAMT (Bilevel Optimization for Covert Memory Tampering in Heterogeneous Multi-Agent Architectures), a novel framework that formalizes attack generation as a bilevel optimization problem. The Upper Level minimizes perturbation magnitude (delta) to enforce covertness while maximizing system behavior divergence toward an adversary-defined target (Lower Level). We provide rigorous mathematical instantiations for CTDE MARL algorithms and RAG-based LLM agents, demonstrating that bilevel optimization uniquely crafts stealthy, minimal-perturbation poisons evading detection heuristics. Comprehensive experimental protocols utilize SMAC and SafeRAG benchmarks to quantify effectiveness at sub-percent poison rates (less than or equal to 1 percent in MARL, less than or equal to 0.1 percent in RAG). XAMT defines a new unified class of training-time threats essential for developing intrinsically secure MAS, with implications for trust, formal verification, and defensive strategies prioritizing intrinsic safety over perimeter-based detection.", "authors": ["Akhil Sharma", "Shaikh Yaser Arafat", "Jai Kumar Sharma", "Ken Huang"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-15", "url": "https://arxiv.org/abs/2512.15790", "pdf_url": "https://arxiv.org/pdf/2512.15790v1", "arxiv_id": "2512.15790", "doi": "10.48550/arXiv.2512.15790", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "4469ff01c317c8fc8fea5031230d45b312e765a27a264efe4bc145cd34749ab9", "sources": ["arxiv", "semantic_scholar"], "title": "AutoTool: Dynamic Tool Selection and Integration for Agentic Reasoning", "abstract": "Agentic reinforcement learning has advanced large language models (LLMs) to reason through long chain-of-thought trajectories while interleaving external tool use. Existing approaches assume a fixed inventory of tools, which limits the adaptability of LLM agents to new or evolving toolsets. We present AutoTool, a training framework that equips LLM agents with dynamic tool-selection capabilities throughout their reasoning trajectories. AutoTool employs a dual-phase optimization pipeline: (i) SFT and RL-based trajectory stabilization for coherent reasoning, and (ii) KL-regularized Plackett-Luce Ranking to refine consistent multi-step tool selection. We further build a 200k dataset with explicit tool-selection rationales across 1,000+ tools and 100+ tasks spanning mathematics, science, code generation, and multimodal reasoning. Across ten diverse benchmarks, we train two base models, Qwen3-8B and Qwen2.5-VL-7B, with AutoTool. With fewer parameters, AutoTool consistently outperforms advanced LLM agents and tool-integration methods, yielding average gains of 6.4% in math & science reasoning, 4.5% in search-based QA, 7.7% in code generation, and 6.9% in multimodal understanding. In addition, AutoTool exhibits stronger generalization by dynamically leveraging unseen tools from evolving toolsets during inference.", "authors": ["Jiaru Zou", "Ling Yang", "Yunzhe Qi", "Sirui Chen", "Mengting Ai", "Ke Shen", "Jingrui He", "Mengdi Wang"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-15", "url": "https://arxiv.org/abs/2512.13278", "pdf_url": "https://arxiv.org/pdf/2512.13278v2", "arxiv_id": "2512.13278", "doi": "10.48550/arXiv.2512.13278", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "b1c638ba98e7226b72dcac0594dbdc78956ccbed0814d2d4612fa88c07a1ee76", "sources": ["arxiv", "semantic_scholar"], "title": "Verification-Guided Context Optimization for Tool Calling via Hierarchical LLMs-as-Editors", "abstract": "Tool calling enables large language models (LLMs) to interact with external environments through tool invocation, providing a practical way to overcome the limitations of pretraining. However, the effectiveness of tool use depends heavily on the quality of the associated documentation and knowledge base context. These materials are usually written for human users and are often misaligned with how LLMs interpret information. This problem is even more pronounced in industrial settings, where hundreds of tools with overlapping functionality create challenges in scalability, variability, and ambiguity. We propose Verification-Guided Context Optimization (VGCO), a framework that uses LLMs as editors to automatically refine tool-related documentation and knowledge base context. VGCO works in two stages. First, Evaluation collects real-world failure cases and identifies mismatches between tools and their context. Second, Optimization performs hierarchical editing through offline learning with structure-aware, in-context optimization. The novelty of our LLM editors has three main aspects. First, they use a hierarchical structure that naturally integrates into the tool-calling workflow. Second, they are state-aware, action-specific, and verification-guided, which constrains the search space and enables efficient, targeted improvements. Third, they enable cost-efficient sub-task specialization, either by prompt engineering large editor models or by post-training smaller editor models. Unlike prior work that emphasizes multi-turn reasoning, VGCO focuses on the single-turn, large-scale tool-calling problem and achieves significant improvements in accuracy, robustness, and generalization across LLMs.", "authors": ["Henger Li", "Shuangjie You", "Flavio Di Palo", "Yiyue Qian", "Ayush Jain"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-15", "url": "https://arxiv.org/abs/2512.13860", "pdf_url": "https://arxiv.org/pdf/2512.13860v1", "arxiv_id": "2512.13860", "doi": "10.48550/arXiv.2512.13860", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "74a82de7b054c251943ee3ac66fa7fa52b0754c258402a29013d59cdf3a5b7b7", "sources": ["arxiv", "semantic_scholar"], "title": "neuralFOMO: Can LLMs Handle Being Second Best? Measuring Envy-Like Preferences in Multi-Agent Settings", "abstract": "Envy shapes competitiveness and cooperation in human groups, yet its role in large language model interactions remains largely unexplored. As LLMs increasingly operate in multi-agent settings, it is important to examine whether they exhibit envy-like preferences under social comparison. We evaluate LLM behavior across two scenarios: (1) a point-allocation game testing sensitivity to relative versus absolute payoff, and (2) comparative evaluations across general and contextual settings. To ground our analysis in psychological theory, we adapt four established psychometric questionnaires spanning general, domain-specific, workplace, and sibling-based envy. Our results reveal heterogeneous envy-like patterns across models and contexts, with some models sacrificing personal gain to reduce a peer's advantage, while others prioritize individual maximization. These findings highlight competitive dispositions as a design and safety consideration for multi-agent LLM systems.", "authors": ["Arnav Ramamoorthy", "Shrey Dhorajiya", "Ojas Pungalia", "Rashi Upadhyay", "Abhishek Mishra", "Abhiram H", "Tejasvi Alladi", "Sujan Yenuganti", "Dhruv Kumar"], "categories": ["cs.AI", "cs.CL", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-15", "url": "https://arxiv.org/abs/2512.13481", "pdf_url": "https://arxiv.org/pdf/2512.13481v2", "arxiv_id": "2512.13481", "doi": "10.48550/arXiv.2512.13481", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "354fa53a7c041f6f90a70ebac39b4f395f32a999f6496446396cc4c8a0142259", "sources": ["arxiv", "semantic_scholar"], "title": "AgentSHAP: Interpreting LLM Agent Tool Importance with Monte Carlo Shapley Value Estimation", "abstract": "LLM agents that use external tools can solve complex tasks, but understanding which tools actually contributed to a response remains a blind spot. No existing XAI methods address tool-level explanations. We introduce AgentSHAP, the first framework for explaining tool importance in LLM agents. AgentSHAP is model-agnostic: it treats the agent as a black box and works with any LLM (GPT, Claude, Llama, etc.) without needing access to internal weights or gradients. Using Monte Carlo Shapley values, AgentSHAP tests how an agent responds with different tool subsets and computes fair importance scores based on game theory. Our contributions are: (1) the first explainability method for agent tool attribution, grounded in Shapley values from game theory; (2) Monte Carlo sampling that reduces cost from O(2n) to practical levels; and (3) comprehensive experiments on API-Bank showing that AgentSHAP produces consistent scores across runs, correctly identifies which tools matter, and distinguishes relevant from irrelevant tools. AgentSHAP joins TokenSHAP (for tokens) and PixelSHAP (for image regions) to complete a family of Shapley-based XAI tools for modern generative AI. Code: https://github.com/GenAISHAP/TokenSHAP.", "authors": ["Miriam Horovicz"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-14", "url": "https://arxiv.org/abs/2512.12597", "pdf_url": "https://arxiv.org/pdf/2512.12597v1", "arxiv_id": "2512.12597", "doi": "10.48550/arXiv.2512.12597", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/GenAISHAP/TokenSHAP", "venue": "arXiv.org", "quality_score": 0.6304}
{"id": "47a7e5eb2c8873238c5b28101785d9e705d5724e6dc31aa5473cad6c1384d72a", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Trustworthy Multi-Turn LLM Agents via Behavioral Guidance", "abstract": "Large Language Models demonstrate strong reasoning and generation abilities, yet their behavior in multi-turn tasks often lacks reliability and verifiability. We present a task completion framework that enables LLM-based agents to act under explicit behavioral guidance in environments described by reinforcement learning formalisms with defined observation, action, and reward signals. The framework integrates three components: a lightweight task profiler that selects reasoning and generation strategies, a reasoning module that learns verifiable observation - action mappings, and a generation module that enforces constraint-compliant outputs through validation or deterministic synthesis. We show that as the agent interacts with the environment, these components co-evolve, yielding trustworthy behavior.", "authors": ["Gonca Gürsun"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-12", "url": "https://arxiv.org/abs/2512.11421", "pdf_url": "https://arxiv.org/pdf/2512.11421v1", "arxiv_id": "2512.11421", "doi": "10.48550/arXiv.2512.11421", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4056}
{"id": "0e6b0742f0c8ddb588206aaeec2717d2b9a7fb133f00d0e75fbc89b2e9992dc8", "sources": ["arxiv", "semantic_scholar"], "title": "FutureWeaver: Planning Test-Time Compute for Multi-Agent Systems with Modularized Collaboration", "abstract": "Scaling test-time computation has been shown to significantly improve large language model (LLM) performance without additional training. However, extending these techniques to multi-agent systems remains challenging: existing approaches lack principled mechanisms for allocating compute to enable effective collaboration, scaling coordination itself, or optimizing compute usage under explicit budget constraints. To address this gap, we propose FutureWeaver, a framework for planning and optimizing test-time compute allocation in multi-agent systems under fixed budgets. It introduces collaboration modules, formalized as modular, callable functions that encapsulate reusable multi-agent workflows and are automatically induced via self-play reflection from recurring interaction patterns. Building on these modules, it employs \\emph{a dual-level planning architecture} that jointly performs short-horizon action selection and long-horizon abstract lookahead to optimize inference trajectories under budget constraints. Experiments on complex agent benchmarks demonstrate that FutureWeaver consistently outperforms baselines across diverse budget settings, validating its effectiveness for multi-agent collaboration in inference-time optimization.", "authors": ["Dongwon Jung", "Peng Shi", "Muhao Chen", "Yi Zhang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-12", "url": "https://arxiv.org/abs/2512.11213", "pdf_url": "https://arxiv.org/pdf/2512.11213v2", "arxiv_id": "2512.11213", "doi": "10.48550/arXiv.2512.11213", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4056}
{"id": "e87b24eb216e2b0aaf2bc1d4068b631b7b11ef6626b6bd3774eb289c6852fd6f", "sources": ["arxiv", "semantic_scholar"], "title": "CompanionCast: Toward Social Collaboration with Multi-Agent Systems in Shared Experiences", "abstract": "Shared experiences are fundamental to social connection, yet media consumption is increasingly solitary. While AI companions offer real-time reactions and emotional regulation, existing systems either rely on single-agent designs or lack the social awareness and multi-party interaction required to replicate authentic group dynamics. We present CompanionCast, a general framework for orchestrating multiple specialized AI agents as social collaborators within a live shared context. CompanionCast integrates multimodal event detection, rolling context caching for improved grounding, and spatial audio to enhance co-presence. We validate CompanionCast through sports viewing, a domain with rich dynamics and strong social traditions. Pilot studies with soccer fans demonstrate that CompanionCast significantly improves perceived social presence and emotional sharing compared to solitary viewing. We conclude by discussing implications and open challenges for multi-agent systems as social collaborators in shared experiences.", "authors": ["Yiyang Wang", "Chen Chen", "Tica Lin", "Vishnu Raj", "Josh Kimball", "Alex Cabral", "Josiah Hester"], "categories": ["cs.HC", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-11", "url": "https://arxiv.org/abs/2512.10918", "pdf_url": "https://arxiv.org/pdf/2512.10918v2", "arxiv_id": "2512.10918", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2574}
{"id": "c8a6a73437a57e692016bab690875fbb6199e042f6ca75ba81d0008c72369339", "sources": ["arxiv", "semantic_scholar"], "title": "MiniScope: A Least Privilege Framework for Authorizing Tool Calling Agents", "abstract": "Tool calling agents are an emerging paradigm in LLM deployment, with major platforms such as ChatGPT, Claude, and Gemini adding connectors and autonomous capabilities. However, the inherent unreliability of LLMs introduces fundamental security risks when these agents operate over sensitive user services. Prior approaches either rely on manually written policies that require security expertise, or place LLMs in the confinement loop, which lacks rigorous security guarantees. We present MiniScope, a framework that enables tool calling agents to operate on user accounts while confining potential damage from unreliable LLMs. MiniScope introduces a novel way to automatically and rigorously enforce least privilege principles by reconstructing permission hierarchies that reflect relationships among tool calls and combining them with a mobile-style permission model to balance security and ease of use. To evaluate MiniScope, we create a synthetic dataset derived from ten popular real-world applications, capturing the complexity of realistic agentic tasks beyond existing simplified benchmarks. Our evaluation shows that MiniScope incurs only 1-6% latency overhead compared to vanilla tool calling agents, while significantly outperforming the LLM based baseline in minimizing permissions as well as computational and operational costs.", "authors": ["Jinhao Zhu", "Kevin Tseng", "Gil Vernik", "Xiao Huang", "Shishir G. Patil", "Vivian Fang", "Raluca Ada Popa"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-11", "url": "https://arxiv.org/abs/2512.11147", "pdf_url": "https://arxiv.org/pdf/2512.11147v1", "arxiv_id": "2512.11147", "doi": "10.48550/arXiv.2512.11147", "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4045}
{"id": "f470cf021752016d1dc9d6cf20b6f88f8dd9f7d3b4353ef3723b45d89be9cdf3", "sources": ["arxiv", "semantic_scholar"], "title": "On the Dynamics of Multi-Agent LLM Communities Driven by Value Diversity", "abstract": "As Large Language Models (LLM) based multi-agent systems become increasingly prevalent, the collective behaviors, e.g., collective intelligence, of such artificial communities have drawn growing attention. This work aims to answer a fundamental question: How does diversity of values shape the collective behavior of AI communities? Using naturalistic value elicitation grounded in the prevalent Schwartz's Theory of Basic Human Values, we constructed multi-agent simulations where communities with varying numbers of agents engaged in open-ended interactions and constitution formation. The results show that value diversity enhances value stability, fosters emergent behaviors, and brings more creative principles developed by the agents themselves without external guidance. However, these effects also show diminishing returns: extreme heterogeneity induces instability. This work positions value diversity as a new axis of future AI capability, bridging AI ability and sociological studies of institutional emergence.", "authors": ["Muhua Huang", "Qinlin Zhao", "Xiaoyuan Yi", "Xing Xie"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-11", "url": "https://arxiv.org/abs/2512.10665", "pdf_url": "https://arxiv.org/pdf/2512.10665v1", "arxiv_id": "2512.10665", "doi": "10.48550/arXiv.2512.10665", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4045}
{"id": "657563f9faf3842e0178aabed83b9d955ed034d0ae71c355b85d6ef21d8dc027", "sources": ["arxiv", "semantic_scholar"], "title": "SWEnergy: An Empirical Study on Energy Efficiency in Agentic Issue Resolution Frameworks with SLMs", "abstract": "Context. LLM-based autonomous agents in software engineering rely on large, proprietary models, limiting local deployment. This has spurred interest in Small Language Models (SLMs), but their practical effectiveness and efficiency within complex agentic frameworks for automated issue resolution remain poorly understood. Goal. We investigate the performance, energy efficiency, and resource consumption of four leading agentic issue resolution frameworks when deliberately constrained to using SLMs. We aim to assess the viability of these systems for this task in resource-limited settings and characterize the resulting trade-offs. Method. We conduct a controlled evaluation of four leading agentic frameworks (SWE-Agent, OpenHands, Mini SWE Agent, AutoCodeRover) using two SLMs (Gemma-3 4B, Qwen-3 1.7B) on the SWE-bench Verified Mini benchmark. On fixed hardware, we measure energy, duration, token usage, and memory over 150 runs per configuration. Results. We find that framework architecture is the primary driver of energy consumption. The most energy-intensive framework, AutoCodeRover (Gemma), consumed 9.4x more energy on average than the least energy-intensive, OpenHands (Gemma). However, this energy is largely wasted. Task resolution rates were near-zero, demonstrating that current frameworks, when paired with SLMs, consume significant energy on unproductive reasoning loops. The SLM's limited reasoning was the bottleneck for success, but the framework's design was the bottleneck for efficiency. Conclusions. Current agentic frameworks, designed for powerful LLMs, fail to operate efficiently with SLMs. We find that framework architecture is the primary driver of energy consumption, but this energy is largely wasted due to the SLMs' limited reasoning. Viable low-energy solutions require shifting from passive orchestration to architectures that actively manage SLM weaknesses.", "authors": ["Arihant Tripathy", "Ch Pavan Harshit", "Karthik Vaidhyanathan"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-10", "url": "https://arxiv.org/abs/2512.09543", "pdf_url": "https://arxiv.org/pdf/2512.09543v2", "arxiv_id": "2512.09543", "doi": "10.1145/3786167.3788406", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Proceedings of the 2026 International Workshop on Agentic Engineering (AGENT 2026), ACM, 2026, pp. 104-111", "quality_score": 0.4033}
{"id": "d0951e4e101a9ba28df0aeacfb0cb218bec4848a16bc5d9ca641880f0c64382a", "sources": ["arxiv", "semantic_scholar"], "title": "Query Optimization Beyond Data Systems: The Case for Multi-Agent Systems", "abstract": "The proliferation of large language models (LLMs) has accelerated the adoption of agent-based workflows, where multiple autonomous agents reason, invoke functions, and collaborate to compose complex data pipelines. However, current approaches to building such agentic architectures remain largely ad hoc, lacking generality, scalability, and systematic optimization. Existing systems often rely on fixed models and single execution engines and are unable to efficiently optimize multiple agents operating over heterogeneous data sources and query engines. This paper presents a vision for a next-generation query optimization framework tailored to multi-agent workflows. We argue that optimizing these workflows can benefit from redesigning query optimization principles to account for new challenges: orchestration of diverse agents, cost efficiency under expensive LLM calls and across heterogeneous engines, and redundancy across tasks. Led by a real-world example and building on an analysis of multi-agent workflows, we outline our envisioned architecture and the main research challenges of building a multi-agent query optimization framework, which aims at enabling automated model selection, workflow composition, and execution across heterogeneous engines. This vision establishes the groundwork for query optimization in emerging multi-agent architectures and opens up a set of future research directions.", "authors": ["Zoi Kaoudi", "Ioana Giurgiu"], "categories": ["cs.DB", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-10", "url": "https://arxiv.org/abs/2512.11001", "pdf_url": "https://arxiv.org/pdf/2512.11001v1", "arxiv_id": "2512.11001", "doi": "10.48550/arXiv.2512.11001", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4033}
{"id": "c358fdd27ae749991e48ead9ea4bf5428834ac173e2462780e3839a2b9ccae28", "sources": ["arxiv", "semantic_scholar"], "title": "Thinking with Images via Self-Calling Agent", "abstract": "Thinking-with-images paradigms have showcased remarkable visual reasoning capability by integrating visual information as dynamic elements into the Chain-of-Thought (CoT). However, optimizing interleaved multimodal CoT (iMCoT) through reinforcement learning remains challenging, as it relies on scarce high-quality reasoning data. In this study, we propose Self-Calling Chain-of-Thought (sCoT), a novel visual reasoning paradigm that reformulates iMCoT as a language-only CoT with self-calling. Specifically, a main agent decomposes the complex visual reasoning task to atomic subtasks and invokes its virtual replicas, i.e. parameter-sharing subagents, to solve them in isolated context. sCoT enjoys substantial training effectiveness and efficiency, as it requires no explicit interleaving between modalities. sCoT employs group-relative policy optimization to reinforce effective reasoning behavior to enhance optimization. Experiments on HR-Bench 4K show that sCoT improves the overall reasoning performance by up to $1.9\\%$ with $\\sim 75\\%$ fewer GPU hours compared to strong baseline approaches. Code is available at https://github.com/YWenxi/think-with-images-through-self-calling.", "authors": ["Wenxi Yang", "Yuzhong Zhao", "Fang Wan", "Qixiang Ye"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-09", "url": "https://arxiv.org/abs/2512.08511", "pdf_url": "https://arxiv.org/pdf/2512.08511v2", "arxiv_id": "2512.08511", "doi": "10.48550/arXiv.2512.08511", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/YWenxi/think-with-images-through-self-calling", "venue": "arXiv.org", "quality_score": 0.6216}
{"id": "a92446bb9ac0a40729082d9dc25cfa882ab5b141f79437aba12aab5c35e1c22c", "sources": ["arxiv", "semantic_scholar"], "title": "Single-Agent Scaling Fails Multi-Agent Intelligence: Towards Foundation Models with Native Multi-Agent Intelligence", "abstract": "Foundation models (FMs) are increasingly assuming the role of the ''brain'' of AI agents. While recent efforts have begun to equip FMs with native single-agent abilities -- such as GUI interaction or integrated tool use -- we argue that the next frontier is endowing FMs with native multi-agent intelligence. We identify four core capabilities of FMs in multi-agent contexts: understanding, planning, efficient communication, and adaptation. Contrary to assumptions about the spontaneous emergence of such abilities, we provide extensive empirical evidence, across 41 large language models and 7 challenging benchmarks, showing that scaling single-agent performance alone does not automatically yield robust multi-agent intelligence. To address this gap, we outline key research directions -- spanning dataset construction, evaluation, training paradigms, and safety considerations -- for building FMs with native multi-agent intelligence.", "authors": ["Shuyue Hu", "Haoyang Yan", "Yiqun Zhang", "Yang Chen", "Dongzhan Zhou", "Lei Bai"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-09", "url": "https://arxiv.org/abs/2512.08743", "pdf_url": "https://arxiv.org/pdf/2512.08743v3", "arxiv_id": "2512.08743", "doi": "10.48550/arXiv.2512.08743", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4022}
{"id": "46868371085bbfe2f3bab78e6fb60a98781434499e8f42b27fdda2de6d932bd9", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent LLM Framework for Design Space Exploration in Autonomous Driving Systems", "abstract": "Designing autonomous driving systems requires efficient exploration of large hardware/software configuration spaces under diverse environmental conditions, e.g., with varying traffic, weather, and road layouts. Traditional design space exploration (DSE) approaches struggle with multi-modal execution outputs and complex performance trade-offs, and often require human involvement to assess correctness based on execution outputs. This paper presents a multi-agent, large language model (LLM)-based DSE framework, which integrates multi-modal reasoning with 3D simulation and profiling tools to automate the interpretation of execution outputs and guide the exploration of system designs. Specialized LLM agents are leveraged to handle user input interpretation, design point generation, execution orchestration, and analysis of both visual and textual execution outputs, which enables identification of potential bottlenecks without human intervention. A prototype implementation is developed and evaluated on a robotaxi case study (an SAE Level 4 autonomous driving application). Compared with a genetic algorithm baseline, the proposed framework identifies more Pareto-optimal, cost-efficient solutions with reduced navigation time under the same exploration budget. Experimental results also demonstrate the efficiency of the adoption of the LLM-based approach for DSE. We believe that this framework paves the way to the design automation of autonomous driving systems.", "authors": ["Po-An Shih", "Shao-Hua Wang", "Yung-Che Li", "Chia-Heng Tu", "Chih-Han Chang"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-09", "url": "https://arxiv.org/abs/2512.08476", "pdf_url": "https://arxiv.org/pdf/2512.08476v1", "arxiv_id": "2512.08476", "doi": "10.48550/arXiv.2512.08476", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2559}
{"id": "eeabfb0a10a2d1e0e5d341ce2ebfc359f872a614c4f5b3f57094ddd499387223", "sources": ["arxiv", "semantic_scholar"], "title": "DART: Leveraging Multi-Agent Disagreement for Tool Recruitment in Multimodal Reasoning", "abstract": "Specialized visual tools can augment large language models or vision language models with expert knowledge (e.g., grounding, spatial reasoning, medical knowledge, etc.), but knowing which tools to call (and when to call them) can be challenging. We introduce DART, a multi-agent framework that uses disagreements between multiple debating visual agents to identify useful visual tools (e.g., object detection, OCR, spatial reasoning, etc.) that can resolve inter-agent disagreement. These tools allow for fruitful multi-agent discussion by introducing new information, and by providing tool-aligned agreement scores that highlight agents in agreement with expert tools, thereby facilitating discussion. We utilize an aggregator agent to select the best answer by providing the agent outputs and tool information. We test DART on four diverse benchmarks and show that our approach improves over multi-agent debate as well as over single agent tool-calling frameworks, beating the next-strongest baseline (multi-agent debate with a judge model) by 3.4% and 2.4% on A-OKVQA and MMMU respectively. We also find that DART adapts well to new tools in applied domains, with a 1.3% improvement on the M3D medical dataset over other strong tool-calling, single agent, and multi-agent baselines. Additionally, we measure text overlap across rounds to highlight the rich discussion in DART compared to existing multi-agent methods. Finally, we study the tool call distribution, finding that diverse tools are reliably used to help resolve disagreement.", "authors": ["Nithin Sivakumaran", "Justin Chih-Yao Chen", "David Wan", "Yue Zhang", "Jaehong Yoon", "Elias Stengel-Eskin", "Mohit Bansal"], "categories": ["cs.CL", "cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-08", "url": "https://arxiv.org/abs/2512.07132", "pdf_url": "https://arxiv.org/pdf/2512.07132v1", "arxiv_id": "2512.07132", "doi": "10.48550/arXiv.2512.07132", "citation_count": 3, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/nsivaku/dart", "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.6198}
{"id": "142243bb2a03a231de50dc3a541ea7839b118c9b659091ef3c99d9e67ab27539", "sources": ["arxiv", "semantic_scholar"], "title": "Understanding LLM Agent Behaviours via Game Theory: Strategy Recognition, Biases and Multi-Agent Dynamics", "abstract": "As Large Language Models (LLMs) increasingly operate as autonomous decision-makers in interactive and multi-agent systems and human societies, understanding their strategic behaviour has profound implications for safety, coordination, and the design of AI-driven social and economic infrastructures. Assessing such behaviour requires methods that capture not only what LLMs output, but the underlying intentions that guide their decisions. In this work, we extend the FAIRGAME framework to systematically evaluate LLM behaviour in repeated social dilemmas through two complementary advances: a payoff-scaled Prisoners Dilemma isolating sensitivity to incentive magnitude, and an integrated multi-agent Public Goods Game with dynamic payoffs and multi-agent histories. These environments reveal consistent behavioural signatures across models and languages, including incentive-sensitive cooperation, cross-linguistic divergence and end-game alignment toward defection. To interpret these patterns, we train traditional supervised classification models on canonical repeated-game strategies and apply them to FAIRGAME trajectories, showing that LLMs exhibit systematic, model- and language-dependent behavioural intentions, with linguistic framing at times exerting effects as strong as architectural differences. Together, these findings provide a unified methodological foundation for auditing LLMs as strategic agents and reveal systematic cooperation biases with direct implications for AI governance, collective decision-making, and the design of safe multi-agent systems.", "authors": ["Trung-Kiet Huynh", "Duy-Minh Dao-Sy", "Thanh-Bang Cao", "Phong-Hao Le", "Hong-Dan Nguyen", "Phu-Quy Nguyen-Lam", "Minh-Luan Nguyen-Vo", "Hong-Phat Pham", "Phu-Hoa Pham", "Thien-Kim Than", "Chi-Nguyen Tran", "Huy Tran", "Gia-Thoai Tran-Le", "Alessio Buscemi", "Le Hong Trang", "The Anh Han"], "categories": ["cs.MA", "cs.AI", "cs.GT", "cs.LG", "math.DS"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2025-12-08", "url": "https://arxiv.org/abs/2512.07462", "pdf_url": "https://arxiv.org/pdf/2512.07462v2", "arxiv_id": "2512.07462", "doi": "10.48550/arXiv.2512.07462", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.401}
{"id": "f808521c4c37a41c33b691aa7ff4657f19b0bdbf4a04fb9e05c60f88181317b8", "sources": ["arxiv", "semantic_scholar"], "title": "Experience-Evolving Multi-Turn Tool-Use Agent with Hybrid Episodic-Procedural Memory", "abstract": "As intents unfold and environments change, multi-turn agents face continuously shifting decision contexts. Although reusing past experience is intuitively appealing, existing approaches remain limited: full trajectories are often too context-specific to transfer, while tool-level reuse ignores the surrounding context and environment. In this paper, we introduce a hybrid episodic-procedural memory strategy (H-EPM) that enables experience-induced self-evolution of multi-turn tool-use policies by adaptively reusing partially overlapping successful experiences during both inference and training. Inspired by human episodic-procedural integration, we construct a tool graph from accumulated trajectories, where recurring tool-to-tool dependencies capture procedural routines and each edge is augmented with compact episodic summaries of relevant context. At inference time, the agent dynamically balances episodic recall for contextual reasoning with procedural execution for routine steps. Beyond inference, H-EPM introduces a memory-guided reinforcement learning paradigm that directly addresses a core challenge in multi-turn agent reinforcement learning, namely ineffective exploration over long trajectories. By biasing exploration toward historically successful tool transitions, H-EPM learns a stronger policy that generalizes at inference time without relying on domain-specific experience collection. Experiments show that H-EPM consistently delivers substantial inference-time gains over strong baselines across multi-turn tool-use benchmarks, reaching improvements of up to fifty percent. It also improves reinforcement learning policy performance, achieving gains of up to forty percent on out-of-distribution tasks.", "authors": ["Sijia Li", "Yuchen Huang", "Zifan Liu", "Zijian Li", "Jingjing fu", "Lei Song", "Jiang Bian", "Jun Zhang", "Rui Wang"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-08", "url": "https://arxiv.org/abs/2512.07287", "pdf_url": "https://arxiv.org/pdf/2512.07287v2", "arxiv_id": "2512.07287", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2552}
{"id": "10e12f3a4d8e4473fc409d100af6fbda5edc6e69903f1e4ab17d03a275e72358", "sources": ["arxiv", "semantic_scholar"], "title": "An Agent-Centric Dynamical Systems Perspective on Multi-Agent Reinforcement Learning", "abstract": "Analysing learning in Multi-Agent Reinforcement Learning (MARL) environments is challenging, in particular with respect to \\textit{individual} decision-making. Practitioners frequently struggle to compare training runs due to the inherent stochasticity in algorithms arising from random dithering exploration, environment transition noise, and stochastic gradient updates to name a few. Traditional analytical approaches, such as replicator dynamics, oft rely on mean-field approximations to remove stochastic effects, but this simplification, whilst able to provide general overall trends, can lead to dissonance between analytical predictions and actual agent realisations. We propose modelling MARL training as a \\textit{coupled stochastic dynamical systems}, capturing both agent interactions and environmental characteristics. Leveraging tools from dynamical systems theory, we pragmatically analyse the stability and sensitivity of agent behaviour, which are key dimensions for their practical deployments, for example, in presence of strict safety requirements. This framework allows us to rigorously study the inherent stochasticity of MARL, providing a deeper understanding of system behaviour.", "authors": ["James Rudd-Jones", "María Pérez-Ortiz", "Mirco Musolesi"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-08", "url": "https://arxiv.org/abs/2512.07588", "pdf_url": "https://arxiv.org/pdf/2512.07588v2", "arxiv_id": "2512.07588", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2552}
{"id": "bad3364261d236e330732eafc7c6d670dbefb0f9b2e7d79c378046476a603c9d", "sources": ["arxiv", "semantic_scholar"], "title": "The Evolution of Agentic AI in Cybersecurity: From Single LLM Reasoners to Multi-Agent Systems and Autonomous Pipelines", "abstract": "Cybersecurity has become one of the earliest adopters of agentic AI, as security operations centers increasingly rely on multi-step reasoning, tool-driven analysis, and rapid decision-making under pressure. While individual large language models can summarize alerts or interpret unstructured reports, they fall short in real SOC environments that require grounded data access, reproducibility, and accountable workflows. In response, the field has seen a rapid architectural evolution from single-model helpers toward tool-augmented agents, distributed multi-agent systems, schema-bound tool ecosystems, and early explorations of semi-autonomous investigative pipelines. This survey presents a five-generation taxonomy of agentic AI in cybersecurity. It traces how capabilities and risks change as systems advance from text-only LLM reasoners to multi-agent collaboration frameworks and constrained-autonomy pipelines. We compare these generations across core dimensions - reasoning depth, tool use, memory, reproducibility, and safety. In addition, we also synthesize emerging benchmarks used to evaluate cyber-oriented agents. Finally, we outline the unresolved challenges that accompany this evolution, such as response validation, tool-use correctness, multi-agent coordination, long-horizon reasoning, and safeguards for high-impact actions. Collectively, this work provides a structured perspective on how agentic AI is taking shape within cybersecurity and what is required to ensure its safe and reliable deployment.", "authors": ["Vaishali Vinay"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-07", "url": "https://arxiv.org/abs/2512.06659", "pdf_url": "https://arxiv.org/pdf/2512.06659v1", "arxiv_id": "2512.06659", "doi": "10.1109/ICAIC67076.2026.11395809", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Applied Informatics and Communication", "quality_score": 0.3999}
{"id": "f7e69cf7594a1c7e84968a8a8ded3aca5ba531bdbef35b3a3731e9f4dc46fb72", "sources": ["arxiv", "semantic_scholar"], "title": "DoVer: Intervention-Driven Auto Debugging for LLM Multi-Agent Systems", "abstract": "Large language model (LLM)-based multi-agent systems are challenging to debug because failures often arise from long, branching interaction traces. The prevailing practice is to leverage LLMs for log-based failure localization, attributing errors to a specific agent and step. However, this paradigm has two key limitations: (i) log-only debugging lacks validation, producing untested hypotheses, and (ii) single-step or single-agent attribution is often ill-posed, as we find that multiple distinct interventions can independently repair the failed task. To address the first limitation, we introduce DoVer, an intervention-driven debugging framework, which augments hypothesis generation with active verification through targeted interventions (e.g., editing messages, altering plans). For the second limitation, rather than evaluating on attribution accuracy, we focus on measuring whether the system resolves the failure or makes quantifiable progress toward task success, reflecting a more outcome-oriented view of debugging. Within the Magnetic-One agent framework, on the datasets derived from GAIA and AssistantBench, DoVer flips 18-28% of failed trials into successes, achieves up to 16% milestone progress, and validates or refutes 30-60% of failure hypotheses. DoVer also performs effectively on a different dataset (GSMPlus) and agent framework (AG2), where it recovers 49% of failed trials. These results highlight intervention as a practical mechanism for improving reliability in agentic systems and open opportunities for more robust, scalable debugging methods for LLM-based multi-agent systems. Project website and code will be available at https://aka.ms/DoVer.", "authors": ["Ming Ma", "Jue Zhang", "Fangkai Yang", "Yu Kang", "Qingwei Lin", "Saravan Rajmohan", "Dongmei Zhang"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-07", "url": "https://arxiv.org/abs/2512.06749", "pdf_url": "https://arxiv.org/pdf/2512.06749v3", "arxiv_id": "2512.06749", "doi": "10.48550/arXiv.2512.06749", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3999}
{"id": "4875ea83418fabf0d70b97c18cab877551c78574597c6fb06f02797dd2553f8f", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Driven Composite Neural Architecture Search for Multi-Source RL State Encoding", "abstract": "Designing state encoders for reinforcement learning (RL) with multiple information sources -- such as sensor measurements, time-series signals, image observations, and textual instructions -- remains underexplored and often requires manual design. We formalize this challenge as a problem of composite neural architecture search (NAS), where multiple source-specific modules and a fusion module are jointly optimized. Existing NAS methods overlook useful side information from the intermediate outputs of these modules -- such as their representation quality -- limiting sample efficiency in multi-source RL settings. To address this, we propose an LLM-driven NAS pipeline in which the LLM serves as a neural architecture design agent, leveraging language-model priors and intermediate-output signals to guide sample-efficient search for high-performing composite state encoders. On a mixed-autonomy traffic control task, our approach discovers higher-performing architectures with fewer candidate evaluations than traditional NAS baselines and the LLM-based GENIUS framework.", "authors": ["Yu Yu", "Qian Xie", "Nairen Cao", "Li Jin"], "categories": ["cs.LG", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-12-07", "url": "https://arxiv.org/abs/2512.06982", "pdf_url": "https://arxiv.org/pdf/2512.06982v2", "arxiv_id": "2512.06982", "doi": "10.48550/arXiv.2512.06982", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3999}
{"id": "caf963156a6dc319cedef952031c931b903db9b6ff3ef57864fe2aae32b686e5", "sources": ["arxiv", "semantic_scholar"], "title": "HiveMind: Contribution-Guided Online Prompt Optimization of LLM Multi-Agent Systems", "abstract": "Recent advances in LLM-based multi-agent systems have demonstrated remarkable capabilities in complex decision-making scenarios such as financial trading and software engineering. However, evaluating each individual agent's effectiveness and online optimization of underperforming agents remain open challenges. To address these issues, we present HiveMind, a self-adaptive framework designed to optimize LLM multi-agent collaboration through contribution analysis. At its core, HiveMind introduces Contribution-Guided Online Prompt Optimization (CG-OPO), which autonomously refines agent prompts based on their quantified contributions. We first propose the Shapley value as a grounded metric to quantify each agent's contribution, thereby identifying underperforming agents in a principled manner for automated prompt refinement. To overcome the computational complexity of the classical Shapley value, we present DAG-Shapley, a novel and efficient attribution algorithm that leverages the inherent Directed Acyclic Graph structure of the agent workflow to axiomatically prune non-viable coalitions. By hierarchically reusing intermediate outputs of agents in the DAG, our method further reduces redundant computations, and achieving substantial cost savings without compromising the theoretical guarantees of Shapley values. Evaluated in a multi-agent stock-trading scenario, HiveMind achieves superior performance compared to static baselines. Notably, DAG-Shapley reduces LLM calls by over 80\\% while maintaining attribution accuracy comparable to full Shapley values, establishing a new standard for efficient credit assignment and enabling scalable, real-world optimization of multi-agent collaboration.", "authors": ["Yihan Xia", "Taotao Wang", "Shengli Zhang", "Zhangyuhua Weng", "Bin Cao", "Soung Chang Liew"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-06", "url": "https://arxiv.org/abs/2512.06432", "pdf_url": "https://arxiv.org/pdf/2512.06432v1", "arxiv_id": "2512.06432", "doi": "10.48550/arXiv.2512.06432", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3987}
{"id": "3f800b1869dde3fd8bb163a1ce10f14939ab516dae4f7bd808ab80b5a2e63934", "sources": ["arxiv", "semantic_scholar"], "title": "ARCANE: A Multi-Agent Framework for Interpretable and Configurable Alignment", "abstract": "As agents based on large language models are increasingly deployed to long-horizon tasks, maintaining their alignment with stakeholder preferences becomes critical. Effective alignment in such settings requires reward models that are interpretable so that stakeholders can understand and audit model objectives. Moreover, reward models must be capable of steering agents at interaction time, allowing preference shifts to be incorporated without retraining. We introduce ARCANE, a framework that frames alignment as a multi-agent collaboration problem that dynamically represents stakeholder preferences as natural-language rubrics: weighted sets of verifiable criteria that can be generated on-the-fly from task context. Inspired by utility theory, we formulate rubric learning as a reconstruction problem and apply a regularized Group-Sequence Policy Optimization (GSPO) procedure that balances interpretability, faithfulness, and computational efficiency. Using a corpus of 219 labeled rubrics derived from the GDPVal benchmark, we evaluate ARCANE on challenging tasks requiring multi-step reasoning and tool use. The learned rubrics produce compact, legible evaluations and enable configurable trade-offs (e.g., correctness vs. conciseness) without retraining. Our results show that rubric-based reward models offer a promising path toward interpretable, test-time adaptive alignment for complex, long-horizon AI systems.", "authors": ["Charlie Masters", "Marta Grześkiewicz", "Stefano V. Albrecht"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-05", "url": "https://arxiv.org/abs/2512.06196", "pdf_url": "https://arxiv.org/pdf/2512.06196v1", "arxiv_id": "2512.06196", "doi": "10.48550/arXiv.2512.06196", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3976}
{"id": "0da1e2f957d1b822de36cf88f82c7f5373a7e09d0c88a30112dbf55db285bd8c", "sources": ["arxiv", "semantic_scholar"], "title": "Topology Matters: Measuring Memory Leakage in Multi-Agent LLMs", "abstract": "Graph topology is a fundamental determinant of memory leakage in multi-agent LLM systems, yet its effects remain poorly quantified. We introduce MAMA (Multi-Agent Memory Attack), a controlled evaluation framework for comparing topology-conditioned memory leakage in multi-agent LLM systems. MAMA operates on synthetic documents containing labeled Personally Identifiable Information (PII) entities, from which we generate sanitized task instructions. We execute a two-phase protocol: Engram (seeding private information into a target agent's memory) and Resonance (multi-round interaction where an attacker attempts extraction). Over 10 rounds, we measure leakage using a two-stage recovery criterion that combines exact-match extraction with LLM-based inference over the attacker's final output. We evaluate six canonical topologies (complete, circle, chain, tree, star, star-ring) across $n\\in\\{4,5,6\\}$, attacker-target placements, and base models. Results are consistent: denser connectivity, shorter attacker-target distance, and higher target centrality increase leakage; most leakage occurs in early rounds and then plateaus; model choice shifts absolute rates but preserves broad structural trends; spatiotemporal/location attributes leak more readily than identity credentials or regulated identifiers. We distill practical guidance for system design: favor sparse or hierarchical connectivity, maximize attacker-target separation, and restrict hub/shortcut pathways via topology-aware access control. Our code is available at https://github.com/llll121/mama-eval.", "authors": ["Jinbo Liu", "Defu Cao", "Yifei Wei", "Tianyao Su", "Yuan Liang", "Yushun Dong", "Yan Liu", "Yue Zhao", "Xiyang Hu"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-04", "url": "https://arxiv.org/abs/2512.04668", "pdf_url": "https://arxiv.org/pdf/2512.04668v4", "arxiv_id": "2512.04668", "doi": "10.48550/arXiv.2512.04668", "citation_count": 4, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/llll121/mama-eval", "venue": "arXiv.org", "quality_score": 0.6127}
{"id": "30736469b53d7cb3e4c8f668b9aae16772dcb81ed4045439f82269fddb8ca30c", "sources": ["arxiv", "semantic_scholar"], "title": "Detecting Perspective Shifts in Multi-agent Systems", "abstract": "Generative models augmented with external tools and update mechanisms (or \\textit{agents}) have demonstrated capabilities beyond intelligent prompting of base models. As agent use proliferates, dynamic multi-agent systems have naturally emerged. Recent work has investigated the theoretical and empirical properties of low-dimensional representations of agents based on query responses at a single time point. This paper introduces the Temporal Data Kernel Perspective Space (TDKPS), which jointly embeds agents across time, and proposes several novel hypothesis tests for detecting behavioral change at the agent- and group-level in black-box multi-agent systems. We characterize the empirical properties of our proposed tests, including their sensitivity to key hyperparameters, in simulations motivated by a multi-agent system of evolving digital personas. Finally, we demonstrate via natural experiment that our proposed tests detect changes that correlate sensitively, specifically, and significantly with a real exogenous event. As far as we are aware, TDKPS is the first principled framework for monitoring behavioral dynamics in black-box multi-agent systems -- a critical capability as generative agent deployment continues to scale.", "authors": ["Eric Bridgeford", "Hayden Helm"], "categories": ["cs.AI", "cs.MA", "stat.ME"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2025-12-04", "url": "https://arxiv.org/abs/2512.05013", "pdf_url": "https://arxiv.org/pdf/2512.05013v2", "arxiv_id": "2512.05013", "doi": "10.48550/arXiv.2512.05013", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3965}
{"id": "b25f3c372d71d0320383a126889c19c6fd042e55d1e51996884004ed58f93c98", "sources": ["arxiv", "semantic_scholar"], "title": "Norm-Governed Multi-Agent Decision-Making in Simulator-Coupled Environments:The Reinsurance Constrained Multi-Agent Simulation Process (R-CMASP)", "abstract": "Reinsurance decision-making exhibits the core structural properties that motivate multi-agent models: distributed and asymmetric information, partial observability, heterogeneous epistemic responsibilities, simulator-driven environment dynamics, and binding prudential and regulatory constraints. Deterministic workflow automation cannot meet these requirements, as it lacks the epistemic flexibility, cooperative coordination mechanisms, and norm-sensitive behaviour required for institutional risk-transfer. We propose the Reinsurance Constrained Multi-Agent Simulation Process (R-CMASP), a formal model that extends stochastic games and Dec-POMDPs by adding three missing elements: (i) simulator-coupled transition dynamics grounded in catastrophe, capital, and portfolio engines; (ii) role-specialized agents with structured observability, belief updates, and typed communication; and (iii) a normative feasibility layer encoding solvency, regulatory, and organizational rules as admissibility constraints on joint actions. Using LLM-based agents with tool access and typed message protocols, we show in a domain-calibrated synthetic environment that governed multi-agent coordination yields more stable, coherent, and norm-adherent behaviour than deterministic automation or monolithic LLM baselines--reducing pricing variance, improving capital efficiency, and increasing clause-interpretation accuracy. Embedding prudential norms as admissibility constraints and structuring communication into typed acts measurably enhances equilibrium stability. Overall, the results suggest that regulated, simulator-driven decision environments are most naturally modelled as norm-governed, simulator-coupled multi-agent systems.", "authors": ["Stella C. Dong"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-04", "url": "https://arxiv.org/abs/2512.09939", "pdf_url": "https://arxiv.org/pdf/2512.09939v1", "arxiv_id": "2512.09939", "doi": "10.48550/arXiv.2512.09939", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3965}
{"id": "fb02b998c12d6326ae0d0e06b709151f378165425e31d5c69030e84596db142d", "sources": ["arxiv", "semantic_scholar"], "title": "Persona-based Multi-Agent Collaboration for Brainstorming", "abstract": "We demonstrate the importance of persona-based multi-agents brainstorming for both diverse topics and subject matter ideation. Prior work has shown that generalized multi-agent collaboration often provides better reasoning than a single agent alone. In this paper, we propose and develop a framework for persona-based agent selection, showing how persona domain curation can improve brainstorming outcomes. Using multiple experimental setups, we evaluate brainstorming outputs across different persona pairings (e.g., Doctor vs VR Engineer) and A2A (agent-to-agent) dynamics (separate, together, separate-then-together). Our results show that (1) persona choice shapes idea domains, (2) collaboration mode shifts diversity of idea generation, and (3) multi-agent persona-driven brainstorming produces idea depth and cross-domain coverage.", "authors": ["Nate Straub", "Saara Khan", "Katharina Jay", "Brian Cabral", "Oskar Linde"], "categories": ["cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-04", "url": "https://arxiv.org/abs/2512.04488", "pdf_url": "https://arxiv.org/pdf/2512.04488v2", "arxiv_id": "2512.04488", "doi": "10.48550/arXiv.2512.04488", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3965}
{"id": "c818aceddf22baacbec7ef446278c7785c171429163769564384ccdc0728c652", "sources": ["arxiv", "semantic_scholar"], "title": "AsymPuzl: An Asymmetric Puzzle for multi-agent cooperation", "abstract": "Large Language Model (LLM) agents are increasingly studied in multi-turn, multi-agent scenarios, yet most existing setups emphasize open-ended role-play rather than controlled evaluation. We introduce AsymPuzl, a minimal but expressive two-agent puzzle environment designed to isolate communication under information asymmetry. Each agent observes complementary but incomplete views of a symbolic puzzle and must exchange messages to solve it cooperatively. Using a diverse set of current-generation and open-source LLMs, we show that (i) strong models such as GPT-5 and Claude-4.0 reliably converge across puzzle sizes on the solution by sharing complete information in two turns, (ii) weaker models often ignore partner messages or over-correct their hypotheses, and (iii) feedback design is non-trivial: simple self-feedback improves success rates, while detailed joint feedback can hurt performance. These findings show that even in simple cooperative tasks, LLM communication strategies diverge and depend on the granularity of feedback signals. AsymPuzl thus provides a testbed for probing the limits of multi-turn cooperation and opens avenues for studying coordination mechanisms.", "authors": ["Xavier Cadet", "Edward Koh", "Peter Chin"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-03", "url": "https://arxiv.org/abs/2512.03466", "pdf_url": "https://arxiv.org/pdf/2512.03466v1", "arxiv_id": "2512.03466", "doi": "10.48550/arXiv.2512.03466", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6109}
{"id": "7dcead0dda3cdfdf6c3355177f712f094ec13e9c8c70b0253af0ac7878852e1a", "sources": ["arxiv", "semantic_scholar"], "title": "Reason-Plan-ReAct: A Reasoner-Planner Supervising a ReAct Executor for Complex Enterprise Tasks", "abstract": "Despite recent advances, autonomous agents often struggle to solve complex tasks in enterprise domains that require coordinating multiple tools and processing diverse data sources. This struggle is driven by two main limitations. First, single-agent architectures enforce a monolithic plan-execute loop, which directly causes trajectory instability. Second, the requirement to use local open-weight models for data privacy introduces smaller context windows leading to the rapid consumption of context from large tool outputs. To solve this problem we introduce RP-ReAct (Reasoner Planner-ReAct), a novel multi-agent approach that fundamentally decouples strategic planning from low-level execution to achieve superior reliability and efficiency. RP-ReAct consists of a Reasoner Planner Agent (RPA), responsible for planning each sub-step, continuously analysing the execution results using the strong reasoning capabilities of a Large Reasoning Model, and one or multiple Proxy-Execution Agent (PEA) that translates sub-steps into concrete tool interactions using a ReAct approach. Crucially, we incorporate a context-saving strategy within the PEA to mitigate context window overflow by managing large tool outputs via external storage and on-demand access. We evaluate RP-ReAct, on the challenging, multi-domain ToolQA benchmark using a diverse set of six open-weight reasoning models. Our empirical results show that RP-ReAct achieves superior performance and improved generalization ability over state-of-the-art baselines when addressing diverse complex tasks across the evaluated domains. Furthermore we establish the enhanced robustness and stability of our approach across different model scales, paving the way for effective and deployable agentic solutions for enterprises.", "authors": ["Gianni Molinari", "Fabio Ciravegna"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-03", "url": "https://arxiv.org/abs/2512.03560", "pdf_url": "https://arxiv.org/pdf/2512.03560v1", "arxiv_id": "2512.03560", "doi": "10.48550/arXiv.2512.03560", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3953}
{"id": "3d9e29b6c83a5b544a73ce70f9ddc7223d9d45adb22da39a73aba62f873f4e22", "sources": ["arxiv", "semantic_scholar"], "title": "Don't Trust Your Upstream: Exploiting LLM Multi-Agent System via Topology-Guided Adversarial Propagation", "abstract": "The digital world is witnessing the rapid rise of LLM-based multi-agent systems (MASs) and their powerful applications. However, their security remains insufficiently understood, as existing evaluations are largely limited to narrow attack settings and may substantially underestimate the real risks of MAS deployments. Inspired by the MAS inter-agent dependencies, where upstream outputs are reinterpreted and executed by downstream agents, we propose a topology-aware attack scheme that propagates adversarial contamination from exposed edge agents to high-privilege agents to induce malicious behaviors. By combining topology reconnaissance, contamination propagation modeling, and hierarchical payload encapsulation, our approach overcomes the key challenges of black-box attacks and makes such multi-hop compromise practical. Experiments show that our approach achieves success rates of 40\\%--78\\% on three widely-used MAS frameworks under five topologies, and 85\\% on two real-world MAS applications across 20 representative scenarios. The results reveal fundamental vulnerabilities in MASs that have been overlooked by prior studies. Based on these findings, we propose a topology-trust mitigation that blocks 94.8\\% of such composite attacks.", "authors": ["Ruichao Liang", "Le Yin", "Jing Chen", "Yebo Feng", "Cong Wu", "Xiaoyu Zhang", "Huangpeng Gu", "Zijian Zhang", "Yang Liu"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-03", "url": "https://arxiv.org/abs/2512.04129", "pdf_url": "https://arxiv.org/pdf/2512.04129v2", "arxiv_id": "2512.04129", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2516}
{"id": "68f884867ceedd7ec7e407615f54104d179d6f40619017136af4ef6e2ecd1a1d", "sources": ["arxiv", "semantic_scholar"], "title": "Thucy: An LLM-based Multi-Agent System for Claim Verification across Relational Databases", "abstract": "In today's age, it is becoming increasingly difficult to decipher truth from lies. Every day, politicians, media outlets, and public figures make conflicting claims -- often about topics that can, in principle, be verified against structured data. For instance, statements about crime rates, economic growth or healthcare can all be verified against official public records and structured datasets. Building a system that can automatically do that would have sounded like science fiction just a few years ago. Yet, with the extraordinary progress in LLMs and agentic AI, this is now within reach. Still, there remains a striking gap between what is technically possible and what is being demonstrated by recent work. Most existing verification systems operate only on small, single-table databases -- typically a few hundred rows -- that conveniently fit within an LLM's context window. In this paper we report our progress on Thucy, the first cross-database, cross-table multi-agent claim verification system that also provides concrete evidence for each verification verdict. Thucy remains completely agnostic to the underlying data sources before deployment and must therefore autonomously discover, inspect, and reason over all available relational databases to verify claims. Importantly, Thucy also reports the exact SQL queries that support its verdict (whether the claim is accurate or not) offering full transparency to expert users familiar with SQL. When evaluated on the TabFact dataset -- the standard benchmark for fact verification over structured data -- Thucy surpasses the previous state of the art by 5.6 percentage points in accuracy (94.3% vs. 88.7%).", "authors": ["Michael Theologitis", "Dan Suciu"], "categories": ["cs.DB", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-02", "url": "https://arxiv.org/abs/2512.03278", "pdf_url": "https://arxiv.org/pdf/2512.03278v2", "arxiv_id": "2512.03278", "doi": "10.48550/arXiv.2512.03278", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3942}
{"id": "19ec99a03aec8e1850bdc8c239758780578e584250f1b35e2fb5fb4abdfc809d", "sources": ["arxiv", "semantic_scholar"], "title": "When Refusals Fail: Unstable Safety Mechanisms in Long-Context LLM Agents", "abstract": "Solving complex or long-horizon problems often requires large language models (LLMs) to use external tools and operate over a significantly longer context window. New LLMs enable longer context windows and support tool calling capabilities. Prior works have focused mainly on evaluation of LLMs on long-context prompts, leaving agentic setup relatively unexplored, both from capability and safety perspectives. Our work addresses this gap. We find that LLM agents could be sensitive to length, type, and placement of the context, exhibiting unexpected and inconsistent shifts in task performance and in refusals to execute harmful requests. Models with 1M-2M token context windows show severe degradation already at 100K tokens, with performance drops exceeding 50\\% for both benign and harmful tasks. Refusal rates shift unpredictably: GPT-4.1-nano increases from $\\sim$5\\% to $\\sim$40\\% while Grok 4 Fast decreases from $\\sim$80\\% to $\\sim$10\\% at 200K tokens. Our work shows potential safety issues with agents operating on longer context and opens additional questions on the current metrics and paradigm for evaluating LLM agent safety on long multi-step tasks. In particular, our results on LLM agents reveal a notable divergence in both capability and safety performance compared to prior evaluations of LLMs on similar criteria.", "authors": ["Tsimur Hadeliya", "Mohammad Ali Jauhar", "Nidhi Sakpal", "Diogo Cruz"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-02", "url": "https://arxiv.org/abs/2512.02445", "pdf_url": "https://arxiv.org/pdf/2512.02445v1", "arxiv_id": "2512.02445", "doi": "10.48550/arXiv.2512.02445", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3942}
{"id": "951b54c026d34493ddc129fbb73e28d4ca369a6294662b68c8df9e596d304740", "sources": ["arxiv", "semantic_scholar"], "title": "Decentralized Multi-Agent System with Trust-Aware Communication", "abstract": "The emergence of Large Language Models (LLMs) is rapidly accelerating the development of autonomous multi-agent systems (MAS), paving the way for the Internet of Agents. However, traditional centralized MAS architectures present significant challenges, including single points of failure, vulnerability to censorship, inherent scalability limitations, and critical trust issues. We propose a novel Decentralized Multi-Agent System (DMAS) architecture designed to overcome these fundamental problems by enabling trust-aware, scalable, and censorship-resistant interactions among autonomous agents. Our DMAS features a decentralized agent runtime underpinned by a blockchain-based architecture. We formalize a trust-aware communication protocol that leverages cryptographic primitives and on-chain operations to provide security properties: verifiable interaction cycles, communication integrity, authenticity, non-repudiation, and conditional confidentiality, which we further substantiate through a comprehensive security analysis. Our performance analysis validates the DMAS as a scalable and efficient solution for building trustworthy multi-agent systems.", "authors": ["Yepeng Ding", "Ahmed Twabi", "Junwei Yu", "Lingfeng Zhang", "Tohru Kondo", "Hiroyuki Sato"], "categories": ["cs.MA", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-02", "url": "https://arxiv.org/abs/2512.02410", "pdf_url": "https://arxiv.org/pdf/2512.02410v1", "arxiv_id": "2512.02410", "doi": "10.1109/ISPA67752.2025.00198", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Symposium on Image and Signal Processing and Analysis", "quality_score": 0.3942}
{"id": "3c7f67292e5e82ccf58ac060d282cfca5fe89b722fcd8c68953d1f61dc7030ce", "sources": ["arxiv", "semantic_scholar"], "title": "InEx: Hallucination Mitigation via Introspection and Cross-Modal Multi-Agent Collaboration", "abstract": "Hallucination remains a critical challenge in large language models (LLMs), hindering the development of reliable multimodal LLMs (MLLMs). Existing solutions often rely on human intervention or underutilize the agent's ability to autonomously mitigate hallucination. To address these limitations, we draw inspiration from how humans make reliable decisions in the real world. They begin with introspective reasoning to reduce uncertainty and form an initial judgment, then rely on external verification from diverse perspectives to reach a final decision. Motivated by this cognitive paradigm, we propose InEx, a training-free, multi-agent framework designed to autonomously mitigate hallucination. InEx introduces internal introspective reasoning, guided by entropy-based uncertainty estimation, to improve the reliability of the decision agent's reasoning process. The agent first generates a response, which is then iteratively verified and refined through external cross-modal multi-agent collaboration with the editing agent and self-reflection agents, further enhancing reliability and mitigating hallucination. Extensive experiments show that InEx consistently outperforms existing methods, achieving 4%-27% gains on general and hallucination benchmarks, and demonstrating strong robustness.", "authors": ["Zhongyu Yang", "Yingfang Yuan", "Xuanming Jiang", "Baoyi An", "Wei Pang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-02", "url": "https://arxiv.org/abs/2512.02981", "pdf_url": "https://arxiv.org/pdf/2512.02981v1", "arxiv_id": "2512.02981", "doi": "10.48550/arXiv.2512.02981", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3942}
{"id": "32341ff99c5e1b0a0d364b32050aef7fda9de2d92dac07e25982f479c7b53f91", "sources": ["arxiv", "semantic_scholar"], "title": "A Flexible Multi-Agent LLM-Human Framework for Fast Human Validated Tool Building", "abstract": "We introduce CollabToolBuilder, a flexible multiagent LLM framework with expert-in-the-loop (HITL) guidance that iteratively learns to create tools for a target goal, aligning with human intent and process, while minimizing time for task/domain adaptation effort and human feedback capture. The architecture generates and validates tools via four specialized agents (Coach, Coder, Critic, Capitalizer) using a reinforced dynamic prompt and systematic human feedback integration to reinforce each agent's role toward goals and constraints. This work is best viewed as a system-level integration and methodology combining multi-agent in-context learning, HITL controls, and reusable tool capitalization for complex iterative problems such as scientific document generation. We illustrate it with preliminary experiments (e.g., generating state-of-the-art research papers or patents given an abstract) and discuss its applicability to other iterative problem-solving.", "authors": ["Daull Xavier", "Patrice Bellot", "Emmanuel Bruno", "Vincent Martin", "Elisabeth Murisasco"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-01", "url": "https://arxiv.org/abs/2512.01434", "pdf_url": "https://arxiv.org/pdf/2512.01434v1", "arxiv_id": "2512.01434", "doi": "10.48550/arXiv.2512.01434", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.393}
{"id": "bddc9817e104a7c5eb58043ef3166d27bcbb1c6913e5ab67144cf722386abf68", "sources": ["arxiv", "semantic_scholar"], "title": "STRIDE: A Systematic Framework for Selecting AI Modalities -- Agentic AI, AI Assistants, or LLM Calls", "abstract": "The rapid shift from stateless large language models (LLMs) to autonomous, goal-driven agents raises a central question: When is agentic AI truly necessary? While agents enable multi-step reasoning, persistent memory, and tool orchestration, deploying them indiscriminately leads to higher cost, complexity, and risk. We present STRIDE (Systematic Task Reasoning Intelligence Deployment Evaluator), a framework that provides principled recommendations for selecting between three modalities: (i) direct LLM calls, (ii) guided AI assistants, and (iii) fully autonomous agentic AI. STRIDE integrates structured task decomposition, dynamism attribution, and self-reflection requirement analysis to produce an Agentic Suitability Score, ensuring that full agentic autonomy is reserved for tasks with inherent dynamism or evolving context. Evaluated across 30 real-world tasks spanning SRE, compliance, and enterprise automation, STRIDE achieved 92% accuracy in modality selection, reduced unnecessary agent deployments by 45%, and cut resource costs by 37%. Expert validation over six months in SRE and compliance domains confirmed its practical utility, with domain specialists agreeing that STRIDE effectively distinguishes between tasks requiring simple LLM calls, guided assistants, or full agentic autonomy. This work reframes agent adoption as a necessity-driven design decision, ensuring autonomy is applied only when its benefits justify the costs.", "authors": ["Shubhi Asthana", "Bing Zhang", "Chad DeLuca", "Ruchi Mahindru", "Hima Patel"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-01", "url": "https://arxiv.org/abs/2512.02228", "pdf_url": "https://arxiv.org/pdf/2512.02228v1", "arxiv_id": "2512.02228", "doi": "10.48550/arXiv.2512.02228", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.393}
{"id": "657b7aedd18979aea9f63c52030ec3fb65278f2951d254c5736060e3c3c88c01", "sources": ["arxiv", "semantic_scholar"], "title": "Extending NGU to Multi-Agent RL: A Preliminary Study", "abstract": "The Never Give Up (NGU) algorithm has proven effective in reinforcement learning tasks with sparse rewards by combining episodic novelty and intrinsic motivation. In this work, we extend NGU to multi-agent environments and evaluate its performance in the simple_tag environment from the PettingZoo suite. Compared to a multi-agent DQN baseline, NGU achieves moderately higher returns and more stable learning dynamics. We investigate three design choices: (1) shared replay buffer versus individual replay buffers, (2) sharing episodic novelty among agents using different k thresholds, and (3) using heterogeneous values of the beta parameter. Our results show that NGU with a shared replay buffer yields the best performance and stability, highlighting that the gains come from combining NGU intrinsic exploration with experience sharing. Novelty sharing performs comparably when k = 1 but degrades learning for larger values. Finally, heterogeneous beta values do not improve over a small common value. These findings suggest that NGU can be effectively applied in multi-agent settings when experiences are shared and intrinsic exploration signals are carefully tuned.", "authors": ["Juan Hernandez", "Diego Fernández", "Manuel Cifuentes", "Denis Parra", "Rodrigo Toro Icarte"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-01", "url": "https://arxiv.org/abs/2512.01321", "pdf_url": "https://arxiv.org/pdf/2512.01321v1", "arxiv_id": "2512.01321", "doi": "10.48550/arXiv.2512.01321", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.393}
{"id": "c76524b3000520171cdce4b628ee83b2729b7f47d1b4bf66b4ff60f2102fa5a9", "sources": ["arxiv", "semantic_scholar"], "title": "DialogGuard: Multi-Agent Psychosocial Safety Evaluation of Sensitive LLM Responses", "abstract": "Large language models (LLMs) now mediate many web-based mental-health, crisis, and other emotionally sensitive services, yet their psychosocial safety in these settings remains poorly understood and weakly evaluated. We present DialogGuard, a multi-agent framework for assessing psychosocial risks in LLM-generated responses along five high-severity dimensions: privacy violations, discriminatory behaviour, mental manipulation, psychological harm, and insulting behaviour. DialogGuard can be applied to diverse generative models through four LLM-as-a-judge pipelines, including single-agent scoring, dual-agent correction, multi-agent debate, and stochastic majority voting, grounded in a shared three-level rubric usable by both human annotators and LLM judges. Using PKU-SafeRLHF with human safety annotations, we show that multi-agent mechanisms detect psychosocial risks more accurately than non-LLM baselines and single-agent judging; dual-agent correction and majority voting provide the best trade-off between accuracy, alignment with human ratings, and robustness, while debate attains higher recall but over-flags borderline cases. We release Dialog-Guard as open-source software with a web interface that provides per-dimension risk scores and explainable natural-language rationales. A formative study with 12 practitioners illustrates how it supports prompt design, auditing, and supervision of web-facing applications for vulnerable users.", "authors": ["Han Luo", "Guy Laban"], "categories": ["cs.AI", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-01", "url": "https://arxiv.org/abs/2512.02282", "pdf_url": "https://arxiv.org/pdf/2512.02282v1", "arxiv_id": "2512.02282", "doi": "10.48550/arXiv.2512.02282", "citation_count": 3, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6074}
{"id": "2418a6abfd9b9923f8a62d4a401982fd65b9ded0eec24ca68a6712d33b588c7c", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-Kernel: A MicroKernel Multi-Agent System Framework for Adaptive Social Simulation Powered by LLMs", "abstract": "Multi-Agent System (MAS) developing frameworks serve as the foundational infrastructure for social simulations powered by Large Language Models (LLMs). However, existing frameworks fail to adequately support large-scale simulation development due to inherent limitations in adaptability, configurability, reliability, and code reusability. For example, they cannot simulate a society where the agent population and profiles change over time. To fill this gap, we propose Agent-Kernel, a framework built upon a novel society-centric modular microkernel architecture. It decouples core system functions from simulation logic and separates cognitive processes from physical environments and action execution. Consequently, Agent-Kernel achieves superior adaptability, configurability, reliability, and reusability. We validate the framework's superiority through two distinct applications: a simulation of the Universe 25 (Mouse Utopia) experiment, which demonstrates the handling of rapid population dynamics from birth to death; and a large-scale simulation of the Zhejiang University Campus Life, successfully coordinating 10,000 heterogeneous agents, including students and faculty.", "authors": ["Yuren Mao", "Peigen Liu", "Xinjian Wang", "Rui Ding", "Jing Miao", "Hui Zou", "Mingjie Qi", "Wanxiang Luo", "Longbin Lai", "Kai Wang", "Zhengping Qian", "Peilun Yang", "Yunjun Gao", "Ying Zhang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-12-01", "url": "https://arxiv.org/abs/2512.01610", "pdf_url": "https://arxiv.org/pdf/2512.01610v1", "arxiv_id": "2512.01610", "doi": "10.48550/arXiv.2512.01610", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.393}
{"id": "c3e15757059d5eb464d545ce16c1ae99db7858aa3f34f76bd72fbe3cc34108f5", "sources": ["arxiv", "semantic_scholar"], "title": "Assertion-Conditioned Compliance: A Provenance-Aware Vulnerability in Multi-Turn Tool-Calling Agents", "abstract": "Multi-turn tool-calling LLMs (models capable of invoking external APIs or tools across several user turns) have emerged as a key feature in modern AI assistants, enabling extended dialogues from benign tasks to critical business, medical, and financial operations. Yet implementing multi-turn pipelines remains difficult for many safety-critical industries due to ongoing concerns regarding model resilience. While standardized benchmarks such as the Berkeley Function-Calling Leaderboard (BFCL) have underpinned confidence concerning advanced function-calling models (like Salesforce's xLAM V2), there is still a lack of visibility into multi-turn conversation-level robustness, especially given their exposure to real-world systems. In this paper, we introduce Assertion-Conditioned Compliance (A-CC), a novel evaluation paradigm for multi-turn function-calling dialogues. A-CC provides holistic metrics that evaluate a model's behavior when confronted with misleading assertions originating from two distinct vectors: (1) user-sourced assertions (USAs), which measure sycophancy toward plausible but misinformed user beliefs, and (2) function-sourced assertions (FSAs), which measure compliance with plausible but contradictory system policies (e.g., stale hints from unmaintained tools). Our results show that models are highly vulnerable to both USA sycophancy and FSA policy conflicts, confirming A-CC as a critical, latent vulnerability in deployed agents.", "authors": ["Daud Waqas", "Aaryamaan Golthi", "Erika Hayashida", "Huanzhi Mao"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-29", "url": "https://arxiv.org/abs/2512.00332", "pdf_url": "https://arxiv.org/pdf/2512.00332v2", "arxiv_id": "2512.00332", "doi": "10.48550/arXiv.2512.00332", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2486}
{"id": "b7d0105bc84db6a5a631850f9445ceef580502863473daaed7d1777d0ac98dbb", "sources": ["arxiv", "semantic_scholar"], "title": "ART: Adaptive Response Tuning Framework -- A Multi-Agent Tournament-Based Approach to LLM Response Optimization", "abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language understanding and generation. However, single-model responses often exhibit inconsistencies, hallucinations, and varying quality across different query domains. This paper presents ART (Adaptive Response Tuning), a novel framework that employs tournament-style ELO ranking and multi-agent reasoning to systematically optimize LLM outputs. By enabling multiple LLM agents to compete, critique, and collaborate through structured tournament workflows, ART produces consensus responses that outperform individual model outputs. Our framework introduces configurable tournament parameters, dynamic agent selection, and multiple consensus fusion strategies. Experimental evaluations demonstrate significant improvements in response accuracy, coherence, and reliability compared to baseline single-model approaches. The ART framework provides a scalable, production-ready solution for applications requiring high-quality, vetted LLM responses, achieving an 8.4% improvement in overall quality metrics and R^2 values exceeding 0.96 in ELO rating convergence.", "authors": ["Omer Jauhar Khan"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-29", "url": "https://arxiv.org/abs/2512.00617", "pdf_url": "https://arxiv.org/pdf/2512.00617v2", "arxiv_id": "2512.00617", "doi": "10.48550/arXiv.2512.00617", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3907}
{"id": "0f03ba565dc1ddc4cd94c6bd101d451de913b68c09c95472c45137234a132f44", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Evolving Latent Strategies for Multi-Agent Language Systems without Model Fine-Tuning", "abstract": "This study proposes a multi-agent language framework that enables continual strategy evolution without fine-tuning the language model's parameters. The core idea is to liberate the latent vectors of abstract concepts from traditional static semantic representations, allowing them to be continuously updated through environmental interaction and reinforcement feedback. We construct a dual-loop architecture: the behavior loop adjusts action preferences based on environmental rewards, while the language loop updates the external latent vectors by reflecting on the semantic embeddings of generated text. Together, these mechanisms allow agents to develop stable and disentangled strategic styles over long-horizon multi-round interactions. Experiments show that agents' latent spaces exhibit clear convergence trajectories under reflection-driven updates, along with structured shifts at critical moments. Moreover, the system demonstrates an emergent ability to implicitly infer and continually adapt to emotional agents, even without shared rewards. These results indicate that, without modifying model parameters, an external latent space can provide language agents with a low-cost, scalable, and interpretable form of abstract strategic representation.", "authors": ["Wenlong Tang"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-28", "url": "https://arxiv.org/abs/2512.20629", "pdf_url": "https://arxiv.org/pdf/2512.20629v3", "arxiv_id": "2512.20629", "doi": "10.48550/arXiv.2512.20629", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wltang-dev/Latent-Strategy-RL-Agent", "venue": "arXiv.org", "quality_score": 0.6021}
{"id": "f276b8fb7dab3ce7905d591be9e215d52e6a9f94c33cd9fa9ef6abc1629f27b1", "sources": ["arxiv", "semantic_scholar"], "title": "Peer-to-Peer Energy Trading in Dairy Farms using Multi-Agent Reinforcement Learning", "abstract": "The integration of renewable energy resources in rural areas, such as dairy farming communities, enables decentralized energy management through Peer-to-Peer (P2P) energy trading. This research highlights the role of P2P trading in efficient energy distribution and its synergy with advanced optimization techniques. While traditional rule-based methods perform well under stable conditions, they struggle in dynamic environments. To address this, Multi-Agent Reinforcement Learning (MARL), specifically Proximal Policy Optimization (PPO) and Deep Q-Networks (DQN), is combined with community/distributed P2P trading mechanisms. By incorporating auction-based market clearing, a price advisor agent, and load and battery management, the approach achieves significant improvements. Results show that, compared to baseline models, DQN reduces electricity costs by 14.2% in Ireland and 5.16% in Finland, while increasing electricity revenue by 7.24% and 12.73%, respectively. PPO achieves the lowest peak hour demand, reducing it by 55.5% in Ireland, while DQN reduces peak hour demand by 50.0% in Ireland and 27.02% in Finland. These improvements are attributed to both MARL algorithms and P2P energy trading, which together results in electricity cost and peak hour demand reduction, and increase electricity selling revenue. This study highlights the complementary strengths of DQN, PPO, and P2P trading in achieving efficient, adaptable, and sustainable energy management in rural communities.", "authors": ["Mian Ibad Ali Shah", "Marcos Eduardo Cruz Victorio", "Maeve Duffy", "Enda Barrett", "Karl Mason"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-28", "url": "https://arxiv.org/abs/2511.23148", "pdf_url": "https://arxiv.org/pdf/2511.23148v1", "arxiv_id": "2511.23148", "doi": "10.1016/j.apenergy.2025.127041", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Applied Energy", "quality_score": 0.3896}
{"id": "b563be89d3da530e0477ff4e0df10a99d668cd7d7da4205074cc5bf22f0d3b5f", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic AI Framework for Individuals with Disabilities and Neurodivergence: A Multi-Agent System for Healthy Eating, Daily Routines, and Inclusive Well-Being", "abstract": "The paper presents a detailed Agentic Artificial Intelligence (AI) model that would enable people with disabilities and neurodivergence to lead healthier lives and have more regular days. The system will use a multi-layer structure; it will include an Application and Interface Layer, an Agents Layer, and a Data Source Layer to provide adaptive, transparent, and inclusive support. Fundamentally, a hybrid reasoning engine will synchronize four special-purpose agents, which include: a personalized-nutrition-based, called a Meal Planner Agent; an adaptive-scheduling-based, called a Reminder Agent; interactive assistance during grocery shopping and cooking, called a Food Guidance Agent; and a continuous-intake-and-physiological-tracking, called a Monitoring Agent. All the agents interact through a central communicative system called the Blackboard/Event Bus, which allows autonomous interaction and real-time feedback loops with multimedia user interfaces. Privacy-sensitive data sources, including electronic health records (EHRs), nutritional databases, wearable sensors, and smart kitchen Internet of Things, are also included in the framework and placed into a policy-controlled layer, which ensures data safety and compliance with consent. Collaborative care and clinician dashboards allow common supervision, and discussable artificial intelligence (XAI) modules give brief explanations of why a decision was made, making users responsible and reliant. The proposed agentic AI framework is an extension beyond traditional assistive systems since it incorporates inclusiveness, personalization, and accessibility at all levels. It displays the intersection of multi-agent reasoning, multi-modal interfaces, and human-centered design that will enable the development of autonomy, health, and digital equity among people with disabilities and neurodivergence.", "authors": ["Salman Jan", "Toqeer Ali Syed", "Gohar Ali", "Ali Akarma", "Mohammad Riyaz Belgaum", "Ahmad Ali"], "categories": ["cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-27", "url": "https://arxiv.org/abs/2511.22737", "pdf_url": "https://arxiv.org/pdf/2511.22737v1", "arxiv_id": "2511.22737", "doi": "10.48550/arXiv.2511.22737", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3884}
{"id": "e34995ca2e854e5fec4fb7ede4344bcdf4fa0de4bb002ed34713785f4f6d5090", "sources": ["arxiv", "semantic_scholar"], "title": "NOMAD: A Multi-Agent LLM System for UML Class Diagram Generation from Natural Language Requirements", "abstract": "Large Language Models (LLMs) are increasingly utilised in software engineering, yet their ability to generate structured artefacts such as UML diagrams remains underexplored. In this work we present NOMAD, a cognitively inspired, modular multi-agent framework that decomposes UML generation into a series of role-specialised subtasks. Each agent handles a distinct modelling activity, such as entity extraction, relationship classification, and diagram synthesis, mirroring the goal-directed reasoning processes of an engineer. This decomposition improves interpretability and allows for targeted verification strategies. We evaluate NOMAD through a mixed design: a large case study (Northwind) for in-depth probing and error analysis, and human-authored UML exercises for breadth and realism. NOMAD outperforms all selected baselines, while revealing persistent challenges in fine-grained attribute extraction. Building on these observations, we introduce the first systematic taxonomy of errors in LLM-generated UML diagrams, categorising structural, relationship, and semantic/logical. Finally, we examine verification as a design probe, showing its mixed effects and outlining adaptive strategies as promising directions. Together, these contributions position NOMAD as both an effective framework for UML class diagram generation and a lens onto the broader research challenges of reliable language-to-model workflows.", "authors": ["Polydoros Giannouris", "Sophia Ananiadou"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-27", "url": "https://arxiv.org/abs/2511.22409", "pdf_url": "https://arxiv.org/pdf/2511.22409v2", "arxiv_id": "2511.22409", "doi": "10.5220/0014301900004058", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2472}
{"id": "affb51c9d21afe0b740e5a3561e67cafc2fd7062d2348d9be41b36f615435ac8", "sources": ["arxiv", "semantic_scholar"], "title": "Tool-RoCo: An Agent-as-Tool Self-organization Large Language Model Benchmark in Multi-robot Cooperation", "abstract": "This study proposes Tool-RoCo, a novel benchmark for evaluating large language models (LLMs) in long-term multi-agent cooperation based on RoCo, a multi-robot cooperative benchmark. Recent research on LLM-based multi-agent systems has relied on predefined orchestration, while ignoring agent autonomy. Tool-RoCo treats other agents as tools and introduces cooperative tools, leveraging tool usage to evaluate multi-agent cooperation and self-organization. Tool usage means that each agent (LLM) selects a tool from a candidate set based on the current state, receives feedback, and adjusts its selection in subsequent rounds. To evaluate different autonomy levels, we propose four LLM paradigms: (1) centralized cooperation, where a single LLM allocates tools to all agents; (2) centralized self-organization, where a central LLM autonomously activates agents while keeping others inactive; (3) decentralized cooperation, where each agent has its own LLM and calls tools based on local information; and (4) self-organization, where a randomly chosen initial agent can request collaboration, activating additional agents via tool calls. Tool-RoCo includes three multi-robot tasks, SORT, PACK, and CABINET, to measure format and parameter accuracy and agent coordination through tool usage. The results using several LLMs showed that cooperative tools accounted for only 7.09% of all tools, indicating that LLM-based agents rarely invoked others as assistants. Moreover, activation tools accounted for 96.42%, suggesting that current LLMs tend to maintain active agents while seldom deactivating them for adaptive coordination. Tool-RoCo provides a systematic benchmark to evaluate LLM autonomy and cooperation in multi-agent tasks. Code and Demo: https://github.com/ColaZhang22/Tool-Roco", "authors": ["Ke Zhang", "Xiaoning Zhao", "Ce Zheng", "Jiahong Ning", "Dandan Zhu", "Wenqi Zhang", "Chen Sun", "Toshiharu Sugawara"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-26", "url": "https://arxiv.org/abs/2511.21510", "pdf_url": "https://arxiv.org/pdf/2511.21510v2", "arxiv_id": "2511.21510", "doi": "10.48550/arXiv.2511.21510", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ColaZhang22/Tool-Roco", "venue": "arXiv.org", "quality_score": 0.5985}
{"id": "b4c338d483a6885493beea3ef33a1c01cc0edddda766c9baedae777cecb1ac9a", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Systems for Dataset Adaptation in Software Engineering: Capabilities, Limitations, and Future Directions", "abstract": "Automating the adaptation of software engineering (SE) research artifacts across datasets is essential for scalability and reproducibility, yet it remains largely unstudied. Recent advances in large language model (LLM)-based multi-agent systems, such as GitHub Copilot's agent mode, promise to automate complex development workflows through coordinated reasoning, code generation, and tool interaction. This paper presents the first empirical study on how state-of-the-art multi-agent systems perform in dataset adaptation tasks. We evaluate Copilot, backed by GPT-4.1 and Claude Sonnet 4, on adapting SE research artifacts from benchmark repositories including ROCODE and LogHub2.0. Through a five-stage evaluation pipeline (file comprehension, code editing, command generation, validation, and final execution), we measure success rates, analyze failure patterns, and assess prompt-based interventions designed to enhance agent performance. Results show that current systems can identify key files and generate partial adaptations but rarely produce functionally correct implementations. Prompt-level interventions, especially providing execution error messages and reference code, substantially improve structural similarity to ground truth (from 7.25% to 67.14%), highlighting the importance of contextual and feedback-driven guidance. Our findings reveal both the promise and limitations of today's multi-agent LLM systems for dataset adaptation, and suggest concrete directions for building more reliable, self-correcting agents in future SE research.", "authors": ["Jingyi Chen", "Xiaoyan Guo", "Songqiang Chen", "Shing-Chi Cheung", "Jiasi Shen"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-26", "url": "https://arxiv.org/abs/2511.21380", "pdf_url": "https://arxiv.org/pdf/2511.21380v1", "arxiv_id": "2511.21380", "doi": "10.48550/arXiv.2511.21380", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3873}
{"id": "84497c759198aa536b70ea9ae4aa01a643c64327f1f3803ae4af40c23cdd29fa", "sources": ["arxiv", "semantic_scholar"], "title": "BAMAS: Structuring Budget-Aware Multi-Agent Systems", "abstract": "Large language model (LLM)-based multi-agent systems have emerged as a powerful paradigm for enabling autonomous agents to solve complex tasks. As these systems scale in complexity, cost becomes an important consideration for practical deployment. However, existing work rarely addresses how to structure multi-agent systems under explicit budget constraints. In this paper, we propose BAMAS, a novel approach for building multi-agent systems with budget awareness. BAMAS first selects an optimal set of LLMs by formulating and solving an Integer Linear Programming problem that balances performance and cost. It then determines how these LLMs should collaborate by leveraging a reinforcement learning-based method to select the interaction topology. Finally, the system is instantiated and executed based on the selected agents and their collaboration topology. We evaluate BAMAS on three representative tasks and compare it with state-of-the-art agent construction methods. Results show that BAMAS achieves comparable performance while reducing cost by up to 86%.", "authors": ["Liming Yang", "Junyu Luo", "Xuanzhe Liu", "Yiling Lou", "Zhenpeng Chen"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-26", "url": "https://arxiv.org/abs/2511.21572", "pdf_url": "https://arxiv.org/pdf/2511.21572v1", "arxiv_id": "2511.21572", "doi": "10.48550/arXiv.2511.21572", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3873}
{"id": "4bbe1d4750b0a73b25d844261688a6542ea1a8cd503729716319fbc87e97f1ef", "sources": ["arxiv", "semantic_scholar"], "title": "DRAFT-RL: Multi-Agent Chain-of-Draft Reasoning for Reinforcement Learning-Enhanced LLMs", "abstract": "Large Language Models (LLMs) have shown impressive capabilities in multi-step reasoning and problem-solving.Recent works introduce multi-agent reflection frameworks where multiple LLM agents critique and refine each other's outputs using reinforcement learning (RL). However, these approaches often rely on single-shot responses and lack structural diversity in reasoning exploration. In this paper, we propose DRAFT-RL, a novel framework that integrates Chain-of-Draft (CoD) reasoning into multi-agent RL training. Instead of generating single responses, each agent produces multiple drafts per query, which are then evaluated by peer agents and a learned reward model to identify the most promising trajectory. These selected drafts are used to refine future reasoning strategies through actor-critic learning.DRAFT-RL enables explicit multi-path exploration, peer-guided reflection, and reward-aligned selection, resulting in more robust and interpretable LLM agent behavior. We evaluate our method on complex reasoning tasks including code synthesis, symbolic math, and knowledge-intensive QA,demonstrating that DRAFT-RL outperforms existing reflective and RL-based agents by significant margins in both accuracy and convergence speed", "authors": ["Yuanhao Li", "Mingshan Liu", "Hongbo Wang", "Yiding Zhang", "Yifei Ma", "Wei Tan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-25", "url": "https://arxiv.org/abs/2511.20468", "pdf_url": "https://arxiv.org/pdf/2511.20468v1", "arxiv_id": "2511.20468", "doi": "10.48550/arXiv.2511.20468", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3861}
{"id": "e19d14b64e75f34eba10939801ce086f6f83ff7ee958bd75e17cfb5b6456018d", "sources": ["arxiv", "semantic_scholar"], "title": "Latent Collaboration in Multi-Agent Systems", "abstract": "Multi-agent systems (MAS) extend large language models (LLMs) from independent single-model reasoning to coordinative system-level intelligence. While existing LLM agents depend on text-based mediation for reasoning and communication, we take a step forward by enabling models to collaborate directly within the continuous latent space. We introduce LatentMAS, an end-to-end training-free framework that enables pure latent collaboration among LLM agents. In LatentMAS, each agent first performs auto-regressive latent thoughts generation through last-layer hidden embeddings instead of text. Then, a shared latent working memory preserves and transfers each agent's internal representations and latent thoughts, ensuring lossless information exchange without re-encoding. We provide detailed theoretical analyses showing that LatentMAS achieves higher expressiveness and lossless information preservation with lower overall complexity than standard text-based MAS. In addition, empirical evaluations across 9 comprehensive benchmarks spanning math and science reasoning, commonsense understanding, and code generation show that LatentMAS outperforms advanced single agents and text-based MAS baselines, achieving up to 14.6% higher accuracy, reducing output token usage by 70.8%-83.7%, and providing 4$\\times$-4.3$\\times$ faster end-to-end inference. Code and data are fully open-sourced at https://github.com/Gen-Verse/LatentMAS.", "authors": ["Jiaru Zou", "Ruizhong Qiu", "Gaotang Li", "Xiyuan Yang", "Katherine Tieu", "Pan Lu", "Ke Shen", "Hanghang Tong", "Yejin Choi", "Jingrui He", "James Zou", "Mengdi Wang", "Ling Yang"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-25", "url": "https://arxiv.org/abs/2511.20639", "pdf_url": "https://arxiv.org/pdf/2511.20639v3", "arxiv_id": "2511.20639", "doi": "10.48550/arXiv.2511.20639", "citation_count": 30, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/Gen-Verse/LatentMAS", "venue": "arXiv.org", "quality_score": 0.5968}
{"id": "3671818abe1a7dceb59814e1c5ea6d4531621be60a09bfc4ad4cf490e1eea163", "sources": ["arxiv", "semantic_scholar"], "title": "An Adaptive, Data-Integrated Agent-Based Modeling Framework for Explainable and Contestable Policy Design", "abstract": "Multi-agent systems often operate under feedback, adaptation, and non-stationarity, yet many simulation studies retain static decision rules and fixed control parameters. This paper introduces a general adaptive multi-agent learning framework that integrates: (i) four dynamic regimes distinguishing static versus adaptive agents and fixed versus adaptive system parameters; (ii) information-theoretic diagnostics (entropy rate, statistical complexity, and predictive information) to assess predictability and structure; (iii) structural causal models for explicit intervention semantics; (iv) procedures for generating agent-level priors from aggregate or sample data; and (v) unsupervised methods for identifying emergent behavioral regimes. The framework offers a domain-neutral architecture for analyzing how learning agents and adaptive controls jointly shape system trajectories, enabling systematic comparison of stability, performance, and interpretability across non-equilibrium, oscillatory, or drifting dynamics. Mathematical definitions, computational operators, and an experimental design template are provided, yielding a structured methodology for developing explainable and contestable multi-agent decision processes.", "authors": ["Roberto Garrone"], "categories": ["cs.MA", "cs.AI", "cs.LG", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-11-24", "url": "https://arxiv.org/abs/2511.19726", "pdf_url": "https://arxiv.org/pdf/2511.19726v1", "arxiv_id": "2511.19726", "doi": "10.48550/arXiv.2511.19726", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.385}
{"id": "256754cd2a190f9b58c460cb7cb64a38c372713d72220fde2244e1df7788919b", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Driven Stationarity-Aware Expert Demonstrations for Multi-Agent Reinforcement Learning in Mobile Systems", "abstract": "Multi-agent reinforcement learning (MARL) has been increasingly adopted in many real-world applications. While MARL enables decentralized deployment on resource-constrained edge devices, it suffers from severe non-stationarity due to the synchronous updates of agent policies. This non stationarity results in unstable training and poor policy con vergence, especially as the number of agents increases. In this paper, we propose RELED, a scalable MARL framework that integrates large language model (LLM)-driven expert demonstrations with autonomous agent exploration. RELED incorporates a Stationarity-Aware Expert Demonstration module, which leverages theoretical non-stationarity bounds to enhance the quality of LLM-generated expert trajectories, thus providing high reward and training-stable samples for each agent. Moreover, a Hybrid Expert-Agent Policy Optimization module adaptively balances each agent's learning from both expert-generated and agent-generated trajectories, accelerating policy convergence and improving generalization. Extensive experiments with real city networks based on OpenStreetMap demonstrate that RELED achieves superior performance compared to state-of-the-art MARL methods.", "authors": ["Tianyang Duan", "Zongyuan Zhang", "Zheng Lin", "Songxiao Guo", "Xiuxian Guan", "Guangyu Wu", "Zihan Fang", "Haotian Meng", "Xia Du", "Ji-Zhe Zhou", "Heming Cui", "Jun Luo", "Yue Gao"], "categories": ["cs.LG", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-24", "url": "https://arxiv.org/abs/2511.19368", "pdf_url": "https://arxiv.org/pdf/2511.19368v1", "arxiv_id": "2511.19368", "doi": "10.48550/arXiv.2511.19368", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Mobile Computing", "quality_score": 0.385}
{"id": "7460281660a4c77330fb3c7fa24f2da6b81f4f101435c692314f7311da910998", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent LLM Framework for Multi-Domain Low-Resource In-Context NER via Knowledge Retrieval, Disambiguation and Reflective Analysis", "abstract": "In-context learning (ICL) with large language models (LLMs) has emerged as a promising paradigm for named entity recognition (NER) in low-resource scenarios. However, existing ICL-based NER methods suffer from three key limitations: (1) reliance on dynamic retrieval of annotated examples, which is problematic when annotated data is scarce; (2) limited generalization to unseen domains due to the LLM's insufficient internal domain knowledge; and (3) failure to incorporate external knowledge or resolve entity ambiguities. To address these challenges, we propose KDR-Agent, a novel multi-agent framework for multi-domain low-resource in-context NER that integrates Knowledge retrieval, Disambiguation, and Reflective analysis. KDR-Agent leverages natural-language type definitions and a static set of entity-level contrastive demonstrations to reduce dependency on large annotated corpora. A central planner coordinates specialized agents to (i) retrieve factual knowledge from Wikipedia for domain-specific mentions, (ii) resolve ambiguous entities via contextualized reasoning, and (iii) reflect on and correct model predictions through structured self-assessment. Experiments across ten datasets from five domains demonstrate that KDR-Agent significantly outperforms existing zero-shot and few-shot ICL baselines across multiple LLM backbones. The code and data can be found at https://github.com/MWXGOD/KDR-Agent.", "authors": ["Wenxuan Mu", "Jinzhong Ning", "Di Zhao", "Yijia Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-24", "url": "https://arxiv.org/abs/2511.19083", "pdf_url": "https://arxiv.org/pdf/2511.19083v1", "arxiv_id": "2511.19083", "doi": "10.48550/arXiv.2511.19083", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/MWXGOD/KDR-Agent", "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.595}
{"id": "52fec5abbfc98122b6146133f9790efd9e58c4f7c8fab08a0839485b7c65cd37", "sources": ["arxiv", "semantic_scholar"], "title": "Z-Space: A Multi-Agent Tool Orchestration Framework for Enterprise-Grade LLM Automation", "abstract": "Large Language Models can break through knowledge and timeliness limitations by invoking external tools within the Model Context Protocol framework to achieve automated execution of complex tasks. However, with the rapid growth of enterprise-scale MCP services, efficiently and accurately matching target functionalities among thousands of heterogeneous tools has become a core challenge restricting system practicality. Existing approaches generally rely on full-prompt injection or static semantic retrieval, facing issues including semantic disconnection between user queries and tool descriptions, context inflation in LLM input, and high inference latency. To address these challenges, this paper proposes Z-Space, a data-generation-oriented multi-agent collaborative tool invocation framework Z-Space. The Z-Space framework establishes a multi-agent collaborative architecture and tool filtering algorithm: (1) A structured semantic understanding of user queries is achieved through an intent parsing model; (2) A tool filtering module (FSWW) based on fused subspace weighted algorithm realizes fine-grained semantic alignment between intents and tools without parameter tuning; (3) An inference execution agent is constructed to support dynamic planning and fault-tolerant execution for multi-step tasks. This framework has been deployed in the Eleme platform's technical division, serving large-scale test data generation scenarios across multiple business units including Taotian, Gaode, and Hema. Production data demonstrates that the system reduces average token consumption in tool inference by 96.26\\% while achieving a 92\\% tool invocation accuracy rate, significantly enhancing the efficiency and reliability of intelligent test data generation systems.", "authors": ["Qingsong He", "Jing Nan", "Jiayu Jiao", "Liangjie Tang", "Xiaodong Xu", "Mengmeng Sun", "Qingyao Wang", "Minghui Yan"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-23", "url": "https://arxiv.org/abs/2511.19483", "pdf_url": "https://arxiv.org/pdf/2511.19483v1", "arxiv_id": "2511.19483", "doi": "10.48550/arXiv.2511.19483", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3839}
{"id": "010d1fc0d7c1c1184259cbe38cb1c40b2c1ff125fa06746fdc4ebeef0e4f4c15", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaborative Filtering: Orchestrating Users and Items for Agentic Recommendations", "abstract": "Agentic recommendations cast recommenders as large language model (LLM) agents that can plan, reason, use tools, and interact with users of varying preferences in web applications. However, most existing agentic recommender systems focus on generic single-agent plan-execute workflows or multi-agent task decomposition pipelines. Without recommendation-oriented design, they often underuse the collaborative signals in the user-item interaction history, leading to unsatisfying recommendation results. To address this, we propose the Multi-Agent Collaborative Filtering (MACF) framework for agentic recommendations, drawing an analogy between traditional collaborative filtering algorithms and LLM-based multi-agent collaboration. Specifically, given a target user and query, we instantiate similar users and relevant items as LLM agents with unique profiles. Each agent is able to call retrieval tools, suggest candidate items, and interact with other agents. Different from the static preference aggregation in traditional collaborative filtering, MACF employs a central orchestrator agent to adaptively manage the collaboration between user and item agents via dynamic agent recruitment and personalized collaboration instruction. Experimental results on datasets from three different domains show the advantages of our MACF framework compared to strong agentic recommendation baselines.", "authors": ["Yu Xia", "Sungchul Kim", "Tong Yu", "Ryan A. Rossi", "Julian McAuley"], "categories": ["cs.CL", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-23", "url": "https://arxiv.org/abs/2511.18413", "pdf_url": "https://arxiv.org/pdf/2511.18413v3", "arxiv_id": "2511.18413", "doi": "10.1145/3774904.3792931", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2443}
{"id": "527f51956bdf9adf3eee17b3ecfd3d772506d8088a8cf4a4d77414daa7004c4a", "sources": ["arxiv", "semantic_scholar"], "title": "Hybrid Agentic AI and Multi-Agent Systems in Smart Manufacturing", "abstract": "The convergence of Agentic AI and MAS enables a new paradigm for intelligent decision making in SMS. Traditional MAS architectures emphasize distributed coordination and specialized autonomy, while recent advances in agentic AI driven by LLMs introduce higher order reasoning, planning, and tool orchestration capabilities. This paper presents a hybrid agentic AI and multi agent framework for a Prescriptive Maintenance use case, where LLM based agents provide strategic orchestration and adaptive reasoning, complemented by rule based and SLMs agents performing efficient, domain specific tasks on the edge. The proposed framework adopts a layered architecture that consists of perception, preprocessing, analytics, and optimization layers, coordinated through an LLM Planner Agent that manages workflow decisions and context retention. Specialized agents autonomously handle schema discovery, intelligent feature analysis, model selection, and prescriptive optimization, while a HITL interface ensures transparency and auditability of generated maintenance recommendations. This hybrid design supports dynamic model adaptation, cost efficient maintenance scheduling, and interpretable decision making. An initial proof of concept implementation is validated on two industrial manufacturing datasets. The developed framework is modular and extensible, supporting seamless integration of new agents or domain modules as capabilities evolve. The results demonstrate the system capability to automatically detect schema, adapt preprocessing pipelines, optimize model performance through adaptive intelligence, and generate actionable, prioritized maintenance recommendations. The framework shows promise in achieving improved robustness, scalability, and explainability for RxM in smart manufacturing, bridging the gap between high level agentic reasoning and low level autonomous execution.", "authors": ["Mojtaba A. Farahani", "Md Irfan Khan", "Thorsten Wuest"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-23", "url": "https://arxiv.org/abs/2511.18258", "pdf_url": "https://arxiv.org/pdf/2511.18258v1", "arxiv_id": "2511.18258", "doi": "10.1016/j.jmsy.2026.04.002", "citation_count": 3, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Journal of manufacturing systems", "quality_score": 0.3839}
{"id": "5631c3d275e7f2f1bcc971beb0aae8292280db4900cd241f2fdbb10a11ff155e", "sources": ["arxiv", "semantic_scholar"], "title": "Shadows in the Code: Exploring the Risks and Defenses of LLM-based Multi-Agent Software Development Systems", "abstract": "The rapid advancement of Large Language Model (LLM)-driven multi-agent systems has significantly streamlined software developing tasks, enabling users with little technical expertise to develop executable applications. While these systems democratize software creation through natural language requirements, they introduce significant security risks that remain largely unexplored. We identify two risky scenarios: Malicious User with Benign Agents (MU-BA) and Benign User with Malicious Agents (BU-MA). We introduce the Implicit Malicious Behavior Injection Attack (IMBIA), demonstrating how multi-agent systems can be manipulated to generate software with concealed malicious capabilities beneath seemingly benign applications, and propose Adv-IMBIA as a defense mechanism. Evaluations across ChatDev, MetaGPT, and AgentVerse frameworks reveal varying vulnerability patterns, with IMBIA achieving attack success rates of 93%, 45%, and 71% in MU-BA scenarios, and 71%, 84%, and 45% in BU-MA scenarios. Our defense mechanism reduced attack success rates significantly, particularly in the MU-BA scenario. Further analysis reveals that compromised agents in the coding and testing phases pose significantly greater security risks, while also identifying critical agents that require protection against malicious user exploitation. Our findings highlight the urgent need for robust security measures in multi-agent software development systems and provide practical guidelines for implementing targeted, resource-efficient defensive strategies.", "authors": ["Xiaoqing Wang", "Keman Huang", "Bin Liang", "Hongyu Li", "Xiaoyong Du"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-23", "url": "https://arxiv.org/abs/2511.18467", "pdf_url": "https://arxiv.org/pdf/2511.18467v1", "arxiv_id": "2511.18467", "doi": "10.48550/arXiv.2511.18467", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3839}
{"id": "bf6f9176750b284588ea69e24e5195764420090f5414c9bfb3bd9f4db40a4095", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-as-a-Graph: Knowledge Graph-Based Tool and Agent Retrieval for LLM Multi-Agent Systems", "abstract": "Recent advances in Large Language Model Multi-Agent Systems enable scalable orchestration and retrieval of specialized, parallelized subagents, each equipped with hundreds or thousands of Model Context Protocol (MCP) servers and tools. However, existing agent, MCP, and retrieval methods typically match queries against a single agent description, obscuring fine-grained tool capabilities of each agent, resulting in suboptimal agent selection. We introduce Agent-as-a-Graph retrieval, a knowledge graph retrieval augmented generation approach that represents both tools and their parent agents as nodes and edges in a knowledge graph. During retrieval, i) relevant agents and tool nodes are first retrieved through vector search, ii) we apply a type-specific weighted reciprocal rank fusion (wRRF) for reranking tools and agents, and iii) parent agents are traversed in the knowledge graph for the final set of agents. We evaluate Agent-as-a-Graph on the LiveMCPBenchmark, achieving 14.9% and 14.6% improvements in Recall@5 and nDCG@5 over prior state-of-the-art retrievers, and 2.4% improvements in wRRF optimizations.", "authors": ["Faheem Nizar", "Elias Lumer", "Anmol Gulati", "Pradeep Honaganahalli Basavaraju", "Vamse Kumar Subbiah"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-22", "url": "https://arxiv.org/abs/2511.18194", "pdf_url": "https://arxiv.org/pdf/2511.18194v1", "arxiv_id": "2511.18194", "doi": "10.48550/arXiv.2511.18194", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2435}
{"id": "cf69db109776e5044debae94a3422ce274701fbbe594166f219c14e597ecba09", "sources": ["arxiv", "semantic_scholar"], "title": "MASTEST: A LLM-Based Multi-Agent System For RESTful API Tests", "abstract": "Testing RESTful API is increasingly important in quality assurance of cloud-native applications. Recent advances in machine learning (ML) techniques have demonstrated that various testing activities can be performed automatically by large language models (LLMs) with reasonable accuracy. This paper develops a multi-agent system called MASTEST that combines LLM-based and programmed agents to form a complete tool chain that covers the whole workflow of API test starting from generating unit and system test scenarios from API specification in the OpenAPI Swagger format, to generating of Pytest test scripts, executing test scripts to interact with web services, to analysing web service response messages to determine test correctness and calculate test coverage. The system also supports the incorporation of human testers in reviewing and correcting LLM generated test artefacts to ensure the quality of testing activities. MASTEST system is evaluated on two LLMs, GPT-4o and DeepSeek V3.1 Reasoner with five public APIs. The performances of LLMs on various testing activities are measured by a wide range of metrics, including unit and system test scenario coverage and API operation coverage for the quality of generated test scenarios, data type correctness, status code coverage and script syntax correctness for the quality of LLM generated test scripts, as well as bug detection ability and usability of LLM generated test scenarios and scripts. Experiment results demonstrated that both DeepSeek and GPT-4o achieved a high overall performance. DeepSeek excels in data type correctness and status code detection, while GPT-4o performs best in API operation coverage. For both models, LLM generated test scripts maintained 100\\% syntax correctness and only required minimal manual edits for semantic correctness. These findings indicate the effectiveness and feasibility of MASTEST.", "authors": ["Xiaoke Han", "Hong Zhu"], "categories": ["cs.SE", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-22", "url": "https://arxiv.org/abs/2511.18038", "pdf_url": "https://arxiv.org/pdf/2511.18038v1", "arxiv_id": "2511.18038", "doi": "10.48550/arXiv.2511.18038", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3827}
{"id": "b21f6f2f005e62ed7723fd5cb193673fa8887f7d1afe25e356f334acb91e84c1", "sources": ["arxiv", "semantic_scholar"], "title": "Bridging Symbolic Control and Neural Reasoning in LLM Agents: Structured Cognitive Loop with a Governance Layer", "abstract": "Large language model agents suffer from fundamental architectural problems: entangled reasoning and execution, memory volatility, and uncontrolled action sequences. We introduce Structured Cognitive Loop (SCL), a modular architecture that explicitly separates agent cognition into five phases: Retrieval, Cognition, Control, Action, and Memory (R-CCAM). Soft Symbolic Control constitutes a dedicated governance layer within SCL, applying symbolic constraints to probabilistic inference while preserving the flexibility of neural reasoning and restoring the explainability and controllability of classical symbolic systems. Through empirical validation on multi-step conditional reasoning tasks, we demonstrate that SCL achieves zero policy violations, eliminates redundant tool calls, and maintains complete decision traceability. These results address critical gaps in existing frameworks such as ReAct, AutoGPT, and memory-augmented approaches. Our contributions are threefold: (1) we situate SCL within the taxonomy of hybrid intelligence, differentiating it from prompt-centric and memory-only approaches; (2) we formally define Soft Symbolic Control and contrast it with neuro-symbolic AI; and (3) we derive three design principles for trustworthy agents: modular decomposition, adaptive symbolic governance, and transparent state management. We provide a complete open-source implementation demonstrating the R-CCAM loop architecture, alongside a live GPT-4o-powered travel planning agent. By connecting expert system principles with modern LLM capabilities, this work offers a practical and theoretically grounded path toward reliable, explainable, and governable AI agents.", "authors": ["Myung Ho Kim"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-21", "url": "https://arxiv.org/abs/2511.17673", "pdf_url": "https://arxiv.org/pdf/2511.17673v5", "arxiv_id": "2511.17673", "doi": "10.48550/arXiv.2511.17673", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5897}
{"id": "244e4a7d81bef1dae609b1eabad1f6085f143d3aae08b61c8b1c93a15d03dd8c", "sources": ["arxiv", "semantic_scholar"], "title": "Optimizing PyTorch Inference with LLM-Based Multi-Agent Systems", "abstract": "Maximizing performance on available GPU hardware is an ongoing challenge for modern AI inference systems. Traditional approaches include writing custom GPU kernels and using specialized model compilers to tune high-level code for specific GPU targets. Recent work shows that LLM-based multi-agent systems can effectively perform such tuning, often outperforming existing compilers and eliminating the need for manual kernel development. However, the dynamics of multi-agent systems for this task remain unexplored. In this work, we present a logical framework for comparing multi-agent PyTorch optimization systems. Our evaluation shows that exploit-heavy strategies perform best when paired with error-fixing agents, and that performance correlates with the granularity of optimization steps. The best implementation achieves an average 2.88x speedup over PyTorch Eager (1.85x over torch.compile) on an H100 GPU across diverse tasks in KernelBench, a benchmark suite covering a range of machine learning architectures in PyTorch. Code is publicly available at: https://github.com/pike-project/pike", "authors": ["Kirill Nagaitsev", "Luka Grbcic", "Samuel Williams", "Costin Iancu"], "categories": ["cs.MA", "cs.AI", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-21", "url": "https://arxiv.org/abs/2511.16964", "pdf_url": "https://arxiv.org/pdf/2511.16964v2", "arxiv_id": "2511.16964", "doi": "10.48550/arXiv.2511.16964", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/pike-project/pike", "venue": "arXiv.org", "quality_score": 0.5897}
{"id": "cedd2177b0008752b6a025102276e6b0d27736378dad8c5af14b8903e3df7e15", "sources": ["arxiv", "semantic_scholar"], "title": "Designing Domain-Specific Agents via Hierarchical Task Abstraction Mechanism", "abstract": "LLM-driven agents, particularly those using general frameworks like ReAct or human-inspired role-playing, often struggle in specialized domains that necessitate rigorously structured workflows. Fields such as remote sensing, requiring specialized tools (e.g., correction, spectral indices calculation), and multi-step procedures (e.g., numerous intermediate products and optional steps), significantly challenge generalized approaches. To address this gap, we introduce a novel agent design framework centered on a Hierarchical Task Abstraction Mechanism (HTAM). Specifically, HTAM moves beyond emulating social roles, instead structuring multi-agent systems into a logical hierarchy that mirrors the intrinsic task-dependency graph of a given domain. This task-centric architecture thus enforces procedural correctness and decomposes complex problems into sequential layers, where each layer's sub-agents operate on the outputs of the preceding layers. We instantiate this framework as EarthAgent, a multi-agent system tailored for complex geospatial analysis. To evaluate such complex planning capabilities, we build GeoPlan-bench, a comprehensive benchmark of realistic, multi-step geospatial planning tasks. It is accompanied by a suite of carefully designed metrics to evaluate tool selection, path similarity, and logical completeness. Experiments show that EarthAgent substantially outperforms a range of established single- and multi-agent systems. Our work demonstrates that aligning agent architecture with a domain's intrinsic task structure is a critical step toward building robust and reliable specialized autonomous systems.", "authors": ["Kaiyu Li", "Jiayu Wang", "Zhi Wang", "Hui Qiao", "Weizhan Zhang", "Deyu Meng", "Xiangyong Cao"], "categories": ["cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-21", "url": "https://arxiv.org/abs/2511.17198", "pdf_url": "https://arxiv.org/pdf/2511.17198v1", "arxiv_id": "2511.17198", "doi": "10.48550/arXiv.2511.17198", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3816}
{"id": "be47d4fa335685c0c30c75746939c548dc54797d835a417bd99af6fd9514a1ee", "sources": ["arxiv", "semantic_scholar"], "title": "SkyRL-Agent: Efficient RL Training for Multi-turn LLM Agent", "abstract": "We introduce SkyRL-Agent, a framework for efficient, multi-turn, long-horizon agent training and evaluation. It provides efficient asynchronous dispatching, lightweight tool integration, and flexible backend interoperability, enabling seamless use with existing RL frameworks such as SkyRL-train, VeRL, and Tinker. Using SkyRL-Agent, we train SA-SWE-32B, a software engineering agent trained from Qwen3-32B (24.4% Pass@1) purely with reinforcement learning. We introduce two key components: an optimized asynchronous pipeline dispatcher that achieves a 1.55x speedup over naive asynchronous batching, and a tool-enhanced training recipe leveraging an AST-based search tool to facilitate code navigation, boost rollout Pass@K, and improve training efficiency. Together, these optimizations enable SA-SWE-32B to reach 39.4% Pass@1 on SWE-Bench Verified with more than 2x cost reduction compared to prior models reaching similar performance. Despite being trained solely on SWE tasks, SA-SWE-32B generalizes effectively to other agentic tasks, including Terminal-Bench, BrowseComp-Plus, and WebArena. We further demonstrate SkyRL-Agent's extensibility through case studies on deep research, computer use, and memory agents, each trained using a different training backend.", "authors": ["Shiyi Cao", "Dacheng Li", "Fangzhou Zhao", "Shuo Yuan", "Sumanth R. Hegde", "Connor Chen", "Charlie Ruan", "Tyler Griggs", "Shu Liu", "Eric Tang", "Richard Liaw", "Philipp Moritz", "Matei Zaharia", "Joseph E. Gonzalez", "Ion Stoica"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-20", "url": "https://arxiv.org/abs/2511.16108", "pdf_url": "https://arxiv.org/pdf/2511.16108v1", "arxiv_id": "2511.16108", "doi": "10.48550/arXiv.2511.16108", "citation_count": 27, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3804}
{"id": "97c7b7f75d67b8889e3fd9d320453624159ca27f5f15b1e0eac059e081c0fe44", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Code Verification via Information Theory", "abstract": "LLMs generate buggy code: 29.6% of SWE-bench solved patches fail, 62% of BaxBench solutions have vulnerabilities, and existing tools only catch 65% of bugs with 35% false positives. We built CodeX-Verify, a multi-agent system that uses four specialized agents to detect different types of bugs. We prove mathematically that combining agents with different detection patterns finds more bugs than any single agent when the agents look for different problems, using submodularity of mutual information under conditional independence. Measuring agent correlation of rho = 0.05 to 0.25 confirms they detect different bugs. Testing on 99 code samples with verified labels shows our system catches 76.1% of bugs, matching the best existing method (Meta Prompt Testing: 75%) while running faster and without test execution. We tested all 15 agent combinations and found that using multiple agents improves accuracy by 39.7 percentage points (from 32.8% to 72.4%) compared to single agents, with diminishing returns of +14.9pp, +13.5pp, and +11.2pp for agents 2, 3, and 4, validating our theoretical model. The best two-agent combination (Correctness + Performance) reaches 79.3% accuracy. Testing on 300 real patches from Claude Sonnet 4.5 runs in under 200ms per sample, making this practical for production use.", "authors": ["Shreshth Rajan"], "categories": ["cs.SE", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-20", "url": "https://arxiv.org/abs/2511.16708", "pdf_url": "https://arxiv.org/pdf/2511.16708v3", "arxiv_id": "2511.16708", "doi": "10.48550/arXiv.2511.16708", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3804}
{"id": "5cae7704a50cecb3f142c89692c1c305ae0a938494b032a2941b49d6dd8746d6", "sources": ["arxiv", "semantic_scholar"], "title": "Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated Reasoning", "abstract": "Large Language Model (LLM) Agents, often trained with Reinforcement Learning (RL), are constrained by a dependency on human-curated data, limiting scalability and tethering AI to human knowledge. Existing self-evolution frameworks offer an alternative but are typically restricted by the model's inherent capabilities and single-round interactions, hindering the development of complex curricula involving tool use or dynamic reasoning. We introduce Agent0, a fully autonomous framework that evolves high-performing agents without external data through multi-step co-evolution and seamless tool integration. Agent0 establishes a symbiotic competition between two agents initialized from the same base LLM: a curriculum agent that proposes increasingly challenging frontier tasks, and an executor agent that learns to solve them. We integrate external tools to enhance the executor's problem-solving capacity; this improvement, in turn, pressures the curriculum agent to construct more complex, tool-aware tasks. Through this iterative process, Agent0 establishes a self-reinforcing cycle that continuously produces high-quality curricula. Empirically, Agent0 substantially boosts reasoning capabilities, improving the Qwen3-8B-Base model by 18% on mathematical reasoning and 24% on general reasoning benchmarks. Code is available at https://github.com/aiming-lab/Agent0.", "authors": ["Peng Xia", "Kaide Zeng", "Jiaqi Liu", "Can Qin", "Fang Wu", "Yiyang Zhou", "Caiming Xiong", "Huaxiu Yao"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-20", "url": "https://arxiv.org/abs/2511.16043", "pdf_url": "https://arxiv.org/pdf/2511.16043v1", "arxiv_id": "2511.16043", "doi": "10.48550/arXiv.2511.16043", "citation_count": 42, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/aiming-lab/Agent0", "venue": "arXiv.org", "quality_score": 0.5879}
{"id": "d1e4f43dd4d71e7f9f665a47a043032fa07e6d59aefcd1715a0c35740b54e327", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLM Orchestration Achieves Deterministic, High-Quality Decision Support for Incident Response", "abstract": "Large language models (LLMs) promise to accelerate incident response in production systems, yet single-agent approaches generate vague, unusable recommendations. We present MyAntFarm.ai, a reproducible containerized framework demonstrating that multi-agent orchestration fundamentally transforms LLM-based incident response quality. Through 348 controlled trials comparing single-agent copilot versus multi-agent systems on identical incident scenarios, we find that multi-agent orchestration achieves 100% actionable recommendation rate versus 1.7% for single-agent approaches, an 80 times improvement in action specificity and 140 times improvement in solution correctness. Critically, multi-agent systems exhibit zero quality variance across all trials, enabling production SLA commitments impossible with inconsistent single-agent outputs. Both architectures achieve similar comprehension latency (approx.40s), establishing that the architectural value lies in deterministic quality, not speed. We introduce Decision Quality (DQ), a novel metric capturing validity, specificity, and correctness properties essential for operational deployment that existing LLM metrics do not address. These findings reframe multi-agent orchestration from a performance optimization to a production-readiness requirement for LLM-based incident response. All code, Docker configurations, and trial data are publicly available for reproduction.", "authors": ["Philip Drammeh"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-19", "url": "https://arxiv.org/abs/2511.15755", "pdf_url": "https://arxiv.org/pdf/2511.15755v2", "arxiv_id": "2511.15755", "doi": "10.48550/arXiv.2511.15755", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3793}
{"id": "68982f81594a7a44f75e344c1f8326a5a2776d4b15844c4d5ff56f710064c68c", "sources": ["arxiv", "semantic_scholar"], "title": "The Subtle Art of Defection: Understanding Uncooperative Behaviors in LLM based Multi-Agent Systems", "abstract": "This paper introduces a novel framework for simulating and analyzing how uncooperative behaviors can destabilize or collapse LLM-based multi-agent systems. Our framework includes two key components: (1) a game theory-based taxonomy of uncooperative agent behaviors, addressing a notable gap in the existing literature; and (2) a structured, multi-stage simulation pipeline that dynamically generates and refines uncooperative behaviors as agents' states evolve. We evaluate the framework via a collaborative resource management setting, measuring system stability using metrics such as survival time and resource overuse rate. Empirically, our framework achieves 96.7% accuracy in generating realistic uncooperative behaviors, validated by human evaluations. Our results reveal a striking contrast: cooperative agents maintain perfect system stability (100% survival over 12 rounds with 0% resource overuse), while any uncooperative behavior can trigger rapid system collapse within 1 to 7 rounds. We also evaluate LLM-based defense methods, finding they detect some uncooperative behaviors, but some behaviors remain largely undetectable. These gaps highlight how uncooperative agents degrade collective outcomes and underscore the need for more resilient multi-agent systems.", "authors": ["Devang Kulshreshtha", "Wanyu Du", "Raghav Jain", "Srikanth Doss", "Hang Su", "Sandesh Swamy", "Yanjun Qi"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-19", "url": "https://arxiv.org/abs/2511.15862", "pdf_url": "https://arxiv.org/pdf/2511.15862v2", "arxiv_id": "2511.15862", "doi": "10.48550/arXiv.2511.15862", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2414}
{"id": "c1a14b3fb710b2a3160033fc650a4990bb20a1bc5b30de3534c1d95ade2c0b1c", "sources": ["arxiv", "semantic_scholar"], "title": "Know Your Intent: An Autonomous Multi-Perspective LLM Agent Framework for DeFi User Transaction Intent Mining", "abstract": "As Decentralized Finance (DeFi) develops, understanding user intent behind DeFi transactions is crucial yet challenging due to complex smart contract interactions, multifaceted on-/off-chain factors, and opaque hex logs. Existing methods lack deep semantic insight. To address this, we propose the Transaction Intent Mining (TIM) framework. TIM leverages a DeFi intent taxonomy built on grounded theory and a multi-agent Large Language Model (LLM) system to robustly infer user intents. A Meta-Level Planner dynamically coordinates domain experts to decompose multiple perspective-specific intent analyses into solvable subtasks. Question Solvers handle the tasks with multi-modal on/off-chain data. While a Cognitive Evaluator mitigates LLM hallucinations and ensures verifiability. Experiments show that TIM significantly outperforms machine learning models, single LLMs, and single Agent baselines. We also analyze core challenges in intent inference. This work helps provide a more reliable understanding of user motivations in DeFi, offering context-aware explanations for complex blockchain activity.", "authors": ["Qian'ang Mao", "Yuxuan Zhang", "Jiaman Chen", "Wenjun Zhou", "Jiaqi Yan"], "categories": ["cs.AI", "q-fin.GN"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2025-11-19", "url": "https://arxiv.org/abs/2511.15456", "pdf_url": "https://arxiv.org/pdf/2511.15456v1", "arxiv_id": "2511.15456", "doi": "10.48550/arXiv.2511.15456", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3793}
{"id": "7e9b2d51eb4759250a5810757f7ed99ded644df5e31083815bf0c6b1b7e175d2", "sources": ["arxiv", "semantic_scholar"], "title": "From Competition to Coordination: Market Making as a Scalable Framework for Safe and Aligned Multi-Agent LLM Systems", "abstract": "As foundation models are increasingly deployed as interacting agents in multi-agent systems, their collective behavior raises new challenges for trustworthiness, transparency, and accountability. Traditional coordination mechanisms, such as centralized oversight or adversarial adjudication, struggle to scale and often obscure how decisions emerge. We introduce a market-making framework for multi-agent large language model (LLM) coordination that organizes agent interactions as structured economic exchanges. In this setup, each agent acts as a market participant, updating and trading probabilistic beliefs, to converge toward shared, truthful outcomes. By aligning local incentives with collective epistemic goals, the framework promotes self-organizing, verifiable reasoning without requiring external enforcement. Empirically, we evaluate this approach across factual reasoning, ethical judgment, and commonsense inference tasks. Market-based coordination yields accuracy gains of up to 10% over single-shot baselines while preserving interpretability and transparency of intermediate reasoning steps. Beyond these improvements, our findings demonstrate that economic coordination principles can operationalize accountability and robustness in multi-agent LLM systems, offering a scalable pathway toward self-correcting, socially responsible AI capable of maintaining trust and oversight in real world deployment scenarios.", "authors": ["Brendan Gho", "Suman Muppavarapu", "Afnan Shaik", "Tyson Tsay", "Atharva Mohan", "James Begin", "Kevin Zhu", "Archana Vaidheeswaran", "Vasu Sharma"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-18", "url": "https://arxiv.org/abs/2511.17621", "pdf_url": "https://arxiv.org/pdf/2511.17621v2", "arxiv_id": "2511.17621", "doi": "10.48550/arXiv.2511.17621", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3781}
{"id": "94aef9c40f8f62d219865df888c711b2a7e96ab90144c3f1e588b30f38900b8a", "sources": ["arxiv", "semantic_scholar"], "title": "AutoTool: Efficient Tool Selection for Large Language Model Agents", "abstract": "Large Language Model (LLM) agents have emerged as powerful tools for automating complex tasks by leveraging the reasoning and decision-making abilities of LLMs. However, a major bottleneck in current agent frameworks lies in the high inference cost of tool selection, especially in approaches like ReAct that repeatedly invoke the LLM to determine which tool to use at each step. In this work, we propose AutoTool, a novel graph-based framework that bypasses repeated LLM inference by exploiting a key empirical observation: tool usage inertia - the tendency of tool invocations to follow predictable sequential patterns. AutoTool constructs a directed graph from historical agent trajectories, where nodes represent tools and edges capture transition probabilities, effectively modeling the inertia in tool selection. It further integrates parameter-level information to refine tool input generation. By traversing this structured representation, AutoTool efficiently selects tools and their parameters with minimal reliance on LLM inference. Extensive experiments across diverse agent tasks demonstrate that AutoTool reduces inference costs by up to 30% while maintaining competitive task completion rates, offering a practical and scalable enhancement for inference-heavy frameworks. Our work highlights the promise of integrating statistical structure into LLM agent design for greater efficiency without sacrificing performance.", "authors": ["Jingyi Jia", "Qinbin Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-18", "url": "https://arxiv.org/abs/2511.14650", "pdf_url": "https://arxiv.org/pdf/2511.14650v1", "arxiv_id": "2511.14650", "doi": "10.48550/arXiv.2511.14650", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/jiajingyyyyyy/AutoTool", "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.5844}
{"id": "31bb002987b07b185aca88f90834c5285f62e8a5e050e6ed79baf642b04d3d6a", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-Oriented Visual Programming for the Web of Things", "abstract": "In this paper we introduce and discuss an approach for multi-agent-oriented visual programming. This aims at enabling individuals without programming experience but with knowledge in specific target domains to design and (re)configure autonomous software. We argue that, compared to procedural programming, it should be simpler for users to create programs when agent abstractions are employed. The underlying rationale is that these abstractions, and specifically the belief-desire-intention architecture that is aligned with human practical reasoning, match more closely with people's everyday experience in interacting with other agents and artifacts in the real world. On top of this, we designed and implemented a visual programming system for agents that hides the technicalities of agent-oriented programming using a blocks-based visual development environment that is built on the JaCaMo platform. To further validate the proposed solution, we integrate the Web of Things (WoT) to let users create autonomous behaviour on top of physical mashups of devices, following the trends in industrial end-user programming. Finally, we report on a pilot user study where we verified that novice users are indeed able to make use of this development environment to create multi-agent systems to solve simple automation tasks.", "authors": ["Samuele Burattini", "Alessandro Ricci", "Simon Mayer", "Danai Vachtsevanou", "Jeremy Lemee", "Andrei Ciortea", "Angelo Croatti"], "categories": ["cs.HC", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-17", "url": "https://arxiv.org/abs/2511.13158", "pdf_url": "https://arxiv.org/pdf/2511.13158v1", "arxiv_id": "2511.13158", "doi": "10.48550/arXiv.2511.13158", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.377}
{"id": "aab36679e2edde2ee90cf997e59265d5480756bc8c8ee4b7412838632178ad53", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO", "abstract": "Multi-agent systems perform well on general reasoning tasks. However, the lack of training in specialized areas hinders their accuracy. Current training methods train a unified large language model (LLM) for all agents in the system. This may limit the performances due to different distributions underlying for different agents. Therefore, training multi-agent systems with distinct LLMs should be the next step to solve. However, this approach introduces optimization challenges. For example, agents operate at different frequencies, rollouts involve varying sub-agent invocations, and agents are often deployed across separate servers, disrupting end-to-end gradient flow. To address these issues, we propose M-GRPO, a hierarchical extension of Group Relative Policy Optimization designed for vertical Multi-agent systems with a main agent (planner) and multiple sub-agents (multi-turn tool executors). M-GRPO computes group-relative advantages for both main and sub-agents, maintaining hierarchical credit assignment. It also introduces a trajectory-alignment scheme that generates fixed-size batches despite variable sub-agent invocations. We deploy a decoupled training pipeline in which agents run on separate servers and exchange minimal statistics via a shared store. This enables scalable training without cross-server backpropagation. In experiments on real-world benchmarks (e.g., GAIA, XBench-DeepSearch, and WebWalkerQA), M-GRPO consistently outperforms both single-agent GRPO and multi-agent GRPO with frozen sub-agents, demonstrating improved stability and sample efficiency. These results show that aligning heterogeneous trajectories and decoupling optimization across specialized agents enhances tool-augmented reasoning tasks.", "authors": ["Haoyang Hong", "Jiajun Yin", "Yuan Wang", "Jingnan Liu", "Zhe Chen", "Ailing Yu", "Ji Li", "Zhiling Ye", "Hansong Xiao", "Yefei Chen", "Hualei Zhou", "Yun Yue", "Minghui Yang", "Chunxiao Guo", "Junwei Liu", "Peng Wei", "Jinjie Gu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-17", "url": "https://arxiv.org/abs/2511.13288", "pdf_url": "https://arxiv.org/pdf/2511.13288v2", "arxiv_id": "2511.13288", "doi": "10.48550/arXiv.2511.13288", "citation_count": 12, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.377}
{"id": "14c9abb244c05b7002708aae686f7bf2b9fd945e98503498ef4316624d5ff2b0", "sources": ["arxiv", "semantic_scholar"], "title": "LoCoBench-Agent: An Interactive Benchmark for LLM Agents in Long-Context Software Engineering", "abstract": "As large language models (LLMs) evolve into sophisticated autonomous agents capable of complex software development tasks, evaluating their real-world capabilities becomes critical. While existing benchmarks like LoCoBench~\\cite{qiu2025locobench} assess long-context code understanding, they focus on single-turn evaluation and cannot capture the multi-turn interactive nature, tool usage patterns, and adaptive reasoning required by real-world coding agents. We introduce \\textbf{LoCoBench-Agent}, a comprehensive evaluation framework specifically designed to assess LLM agents in realistic, long-context software engineering workflows. Our framework extends LoCoBench's 8,000 scenarios into interactive agent environments, enabling systematic evaluation of multi-turn conversations, tool usage efficiency, error recovery, and architectural consistency across extended development sessions. We also introduce an evaluation methodology with 9 metrics across comprehension and efficiency dimensions. Our framework provides agents with 8 specialized tools (file operations, search, code analysis) and evaluates them across context lengths ranging from 10K to 1M tokens, enabling precise assessment of long-context performance. Through systematic evaluation of state-of-the-art models, we reveal several key findings: (1) agents exhibit remarkable long-context robustness; (2) comprehension-efficiency trade-off exists with negative correlation, where thorough exploration increases comprehension but reduces efficiency; and (3) conversation efficiency varies dramatically across models, with strategic tool usage patterns differentiating high-performing agents. As the first long-context LLM agent benchmark for software engineering, LoCoBench-Agent establishes a rigorous foundation for measuring agent capabilities, identifying performance gaps, and advancing autonomous software development at scale.", "authors": ["Jielin Qiu", "Zuxin Liu", "Zhiwei Liu", "Rithesh Murthy", "Jianguo Zhang", "Haolin Chen", "Shiyu Wang", "Ming Zhu", "Liangwei Yang", "Juntao Tan", "Roshan Ram", "Akshara Prabhakar", "Tulika Awalgaonkar", "Zixiang Chen", "Zhepeng Cen", "Cheng Qian", "Shelby Heinecke", "Weiran Yao", "Silvio Savarese", "Caiming Xiong", "Huan Wang"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-17", "url": "https://arxiv.org/abs/2511.13998", "pdf_url": "https://arxiv.org/pdf/2511.13998v1", "arxiv_id": "2511.13998", "doi": "10.48550/arXiv.2511.13998", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.377}
{"id": "1d77bbdb6483f7be0135bfe2f42e4d7c2f1fdf0e6a94efd672d3ea279416c5d6", "sources": ["arxiv", "semantic_scholar"], "title": "Emergent Convergence in Multi-Agent LLM Annotation", "abstract": "Large language models (LLMs) are increasingly deployed in collaborative settings, yet little is known about how they coordinate when treated as black-box agents. We simulate 7500 multi-agent, multi-round discussions in an inductive coding task, generating over 125000 utterances that capture both final annotations and their interactional histories. We introduce process-level metrics: code stability, semantic self-consistency, and lexical confidence alongside sentiment and convergence measures, to track coordination dynamics. To probe deeper alignment signals, we analyze the evolving geometry of output embeddings, showing that intrinsic dimensionality declines over rounds, suggesting semantic compression. The results reveal that LLM groups converge lexically and semantically, develop asymmetric influence patterns, and exhibit negotiation-like behaviors despite the absence of explicit role prompting. This work demonstrates how black-box interaction analysis can surface emergent coordination strategies, offering a scalable complement to internal probe-based interpretability methods.", "authors": ["Angelina Parfenova", "Alexander Denzler", "Juergen Pfeffer"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-17", "url": "https://arxiv.org/abs/2512.00047", "pdf_url": "https://arxiv.org/pdf/2512.00047v1", "arxiv_id": "2512.00047", "doi": "10.18653/v1/2025.blackboxnlp-1.12", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "EMNLP2025", "quality_score": 0.377}
{"id": "75384cf43443ff73a79f7a384cf132e7495995e7191908cb0a5d6ba1b250c43c", "sources": ["arxiv", "semantic_scholar"], "title": "A novel strategy for multi-resource load balancing in agent-based systems", "abstract": "The paper presents a multi-resource load balancing strategy which can be utilised within an agent-based system. This approach can assist system designers in their attempts to optimise the structure for complex enterprise architectures. In this system, the social behaviour of the agent and its adaptation abilities are applied to determine an optimal setup for a given configuration. All the methods have been developed to allow the agent's self-assessment. The proposed agent system has been implemented and the experiment results are presented here.", "authors": ["Leszek Sliwko", "Aleksander Zgrzywa"], "categories": ["cs.MA", "cs.AI", "cs.DC", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-15", "url": "https://arxiv.org/abs/2511.17580", "pdf_url": "https://arxiv.org/pdf/2511.17580v1", "arxiv_id": "2511.17580", "doi": "10.1504/IJIIDS.2009.025162", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Journal of Intelligent Information and Database Systems", "quality_score": 0.3747}
{"id": "318fea5c30a266cdf4a0f8bd58604cd7e74328711faf2cf408695ed86a3f5ca0", "sources": ["arxiv", "semantic_scholar"], "title": "Goal-Oriented Multi-Agent Reinforcement Learning for Decentralized Agent Teams", "abstract": "Connected and autonomous vehicles across land, water, and air must often operate in dynamic, unpredictable environments with limited communication, no centralized control, and partial observability. These real-world constraints pose significant challenges for coordination, particularly when vehicles pursue individual objectives. To address this, we propose a decentralized Multi-Agent Reinforcement Learning (MARL) framework that enables vehicles, acting as agents, to communicate selectively based on local goals and observations. This goal-aware communication strategy allows agents to share only relevant information, enhancing collaboration while respecting visibility limitations. We validate our approach in complex multi-agent navigation tasks featuring obstacles and dynamic agent populations. Results show that our method significantly improves task success rates and reduces time-to-goal compared to non-cooperative baselines. Moreover, task performance remains stable as the number of agents increases, demonstrating scalability. These findings highlight the potential of decentralized, goal-driven MARL to support effective coordination in realistic multi-vehicle systems operating across diverse domains.", "authors": ["Hung Du", "Hy Nguyen", "Srikanth Thudumu", "Rajesh Vasa", "Kon Mouzakis"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-15", "url": "https://arxiv.org/abs/2511.11992", "pdf_url": "https://arxiv.org/pdf/2511.11992v1", "arxiv_id": "2511.11992", "doi": "10.1109/CCNC65079.2026.11366321", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Consumer Communications and Networking Conference", "quality_score": 0.3747}
{"id": "b55bcd07ce30488e0af2181fd1198bdef8bdb87c09f359228626459920d00ee4", "sources": ["arxiv", "semantic_scholar"], "title": "MALBO: Optimizing LLM-Based Multi-Agent Teams via Multi-Objective Bayesian Optimization", "abstract": "The optimal assignment of Large Language Models (LLMs) to specialized roles in multi-agent systems is a significant challenge, defined by a vast combinatorial search space, expensive black-box evaluations, and an inherent trade-off between performance and cost. Current optimization methods focus on single-agent settings and lack a principled framework for this multi-agent, multi-objective problem. This thesis introduces MALBO (Multi-Agent LLM Bayesian Optimization), a systematic framework designed to automate the efficient composition of LLM-based agent teams. We formalize the assignment challenge as a multi-objective optimization problem, aiming to identify the Pareto front of configurations between task accuracy and inference cost. The methodology employs multi-objective Bayesian Optimization (MOBO) with independent Gaussian Process surrogate models. By searching over a continuous feature-space representation of the LLMs, this approach performs a sample-efficient exploration guided by the expected hypervolume improvement. The primary contribution is a principled and automated methodology that yields a Pareto front of optimal team configurations. Our results demonstrate that the Bayesian optimization phase, compared to an initial random search, maintained a comparable average performance while reducing the average configuration cost by over 45%. Furthermore, MALBO identified specialized, heterogeneous teams that achieve cost reductions of up to 65.8% compared to homogeneous baselines, all while maintaining maximum performance. The framework thus provides a data-driven tool for deploying cost-effective and highly specialized multi-agent AI systems.", "authors": ["Antonio Sabbatella"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-14", "url": "https://arxiv.org/abs/2511.11788", "pdf_url": "https://arxiv.org/pdf/2511.11788v1", "arxiv_id": "2511.11788", "doi": "10.48550/arXiv.2511.11788", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3735}
{"id": "9adebc14f051fb5ef158631914762ddcbad1fe067f656ac012f939062e858dcc", "sources": ["arxiv", "semantic_scholar"], "title": "iMAD: Intelligent Multi-Agent Debate for Efficient and Accurate LLM Inference", "abstract": "Large Language Model (LLM) agent systems have advanced rapidly, driven by their strong generalization in zero-shot settings. To further enhance reasoning and accuracy on complex tasks, Multi-Agent Debate (MAD) has emerged as a promising framework that engages multiple LLM agents in structured debates to encourage diverse reasoning. However, triggering MAD for every query is inefficient, as it incurs substantial computational (token) cost and may even degrade accuracy by overturning correct single-agent answers. To address these limitations, we propose intelligent Multi-Agent Debate (iMAD), a token-efficient framework that selectively triggers MAD only when it is likely to be beneficial (i.e., correcting an initially wrong answer). To achieve this goal, iMAD learns generalizable model behaviors to make accurate debate decisions. Specifically, iMAD first prompts a single agent to produce a structured self-critique response, from which we extract 41 interpretable linguistic and semantic features capturing hesitation cues. Then, iMAD uses a lightweight debate-decision classifier, trained using our proposed FocusCal loss, to determine whether to trigger MAD, enabling robust debate decisions without test dataset-specific tuning. Through extensive experiments using six (visual) question answering datasets against five competitive baselines, we have shown that iMAD significantly reduces token usage (by up to 92%) while also improving final answer accuracy (by up to 13.5%).", "authors": ["Wei Fan", "JinYi Yoon", "Bo Ji"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-14", "url": "https://arxiv.org/abs/2511.11306", "pdf_url": "https://arxiv.org/pdf/2511.11306v2", "arxiv_id": "2511.11306", "doi": "10.48550/arXiv.2511.11306", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3735}
{"id": "f86e91712d65afcb803fff1f5552045eba306b37b0dc4e3c901fed1b883bb4bf", "sources": ["arxiv", "semantic_scholar"], "title": "Key Decision-Makers in Multi-Agent Debates: Who Holds the Power?", "abstract": "Recent studies on LLM agent scaling have highlighted the potential of Multi-Agent Debate (MAD) to enhance reasoning abilities. However, the critical aspect of role allocation strategies remains underexplored. In this study, we demonstrate that allocating roles with differing viewpoints to specific positions significantly impacts MAD's performance in reasoning tasks. Specifically, we find a novel role allocation strategy, \"Truth Last\", which can improve MAD performance by up to 22% in reasoning tasks. To address the issue of unknown truth in practical applications, we propose the Multi-Agent Debate Consistency (MADC) strategy, which systematically simulates and optimizes its core mechanisms. MADC incorporates path consistency to assess agreement among independent roles, simulating the role with the highest consistency score as the truth. We validated MADC across a range of LLMs (9 models), including the DeepSeek-R1 Distilled Models, on challenging reasoning tasks. MADC consistently demonstrated advanced performance, effectively overcoming MAD's performance bottlenecks and providing a crucial pathway for further improvements in LLM agent scaling.", "authors": ["Qian Zhang", "Yan Zheng", "Jinyi Liu", "Hebin Liang", "Lanjun Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-14", "url": "https://arxiv.org/abs/2511.11040", "pdf_url": "https://arxiv.org/pdf/2511.11040v1", "arxiv_id": "2511.11040", "doi": "10.48550/arXiv.2511.11040", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3735}
{"id": "7556ada5ff20704ae51865973a9afcba12a741dadc87c790a7314dce55c411cf", "sources": ["arxiv", "semantic_scholar"], "title": "Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting", "abstract": "LLM-based agents are increasingly deployed in multi-agent systems (MAS). As these systems move toward real-world applications, their security becomes paramount. Existing research largely evaluates single-agent security, leaving a critical gap in understanding the vulnerabilities introduced by multi-agent design. However, existing systems fall short due to lack of unified frameworks and metrics focusing on unique rejection modes in MAS. We present SafeAgents, a unified and extensible framework for fine-grained security assessment of MAS. SafeAgents systematically exposes how design choices such as plan construction strategies, inter-agent context sharing, and fallback behaviors affect susceptibility to adversarial prompting. We introduce Dharma, a diagnostic measure that helps identify weak links within multi-agent pipelines. Using SafeAgents, we conduct a comprehensive study across five widely adopted multi-agent architectures (centralized, decentralized, and hybrid variants) on four datasets spanning web tasks, tool use, and code generation. Our findings reveal that common design patterns carry significant vulnerabilities. For example, centralized systems that delegate only atomic instructions to sub-agents obscure harmful objectives, reducing robustness. Our results highlight the need for security-aware design in MAS. Link to code is https://github.com/microsoft/SafeAgents", "authors": ["Nirmit Arora", "Sathvik Joel", "Ishan Kavathekar", " Palak", "Rohan Gandhi", "Yash Pandya", "Tanuja Ganu", "Aditya Kanade", "Akshay Nambi"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-14", "url": "https://arxiv.org/abs/2511.10949", "pdf_url": "https://arxiv.org/pdf/2511.10949v1", "arxiv_id": "2511.10949", "doi": "10.48550/arXiv.2511.10949", "citation_count": 5, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/microsoft/SafeAgents", "venue": "arXiv.org", "quality_score": 0.5773}
{"id": "f86cc1ff4fd81520535d0bdaeaebe7747621c6eed424e79ff6cf8b214130ac14", "sources": ["arxiv", "semantic_scholar"], "title": "From Single to Societal: Analyzing Persona-Induced Bias in Multi-Agent Interactions", "abstract": "Large Language Model (LLM)-based multi-agent systems are increasingly used to simulate human interactions and solve collaborative tasks. A common practice is to assign agents with personas to encourage behavioral diversity. However, this raises a critical yet underexplored question: do personas introduce biases into multi-agent interactions? This paper presents a systematic investigation into persona-induced biases in multi-agent interactions, with a focus on social traits like trustworthiness (how an agent's opinion is received by others) and insistence (how strongly an agent advocates for its opinion). Through a series of controlled experiments in collaborative problem-solving and persuasion tasks, we reveal that (1) LLM-based agents exhibit biases in both trustworthiness and insistence, with personas from historically advantaged groups (e.g., men and White individuals) perceived as less trustworthy and demonstrating less insistence; and (2) agents exhibit significant in-group favoritism, showing a higher tendency to conform to others who share the same persona. These biases persist across various LLMs, group sizes, and numbers of interaction rounds, highlighting an urgent need for awareness and mitigation to ensure the fairness and reliability of multi-agent systems.", "authors": ["Jiayi Li", "Xiao Liu", "Yansong Feng"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-14", "url": "https://arxiv.org/abs/2511.11789", "pdf_url": "https://arxiv.org/pdf/2511.11789v1", "arxiv_id": "2511.11789", "doi": "10.48550/arXiv.2511.11789", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3735}
{"id": "bda2957865ee41356a6e0e83bcbabb54ea153dc4eda970e8e3b79d6029c060e7", "sources": ["arxiv", "semantic_scholar"], "title": "HPCAgentTester: A Multi-Agent LLM Approach for Enhanced HPC Unit Test Generation", "abstract": "Unit testing in High-Performance Computing (HPC) is critical but challenged by parallelism, complex algorithms, and diverse hardware. Traditional methods often fail to address non-deterministic behavior and synchronization issues in HPC applications. This paper introduces HPCAgentTester, a novel multi-agent Large Language Model (LLM) framework designed to automate and enhance unit test generation for HPC software utilizing OpenMP and MPI. HPCAgentTester employs a unique collaborative workflow where specialized LLM agents (Recipe Agent and Test Agent) iteratively generate and refine test cases through a critique loop. This architecture enables the generation of context-aware unit tests that specifically target parallel execution constructs, complex communication patterns, and hierarchical parallelism. We demonstrate HPCAgentTester's ability to produce compilable and functionally correct tests for OpenMP and MPI primitives, effectively identifying subtle bugs that are often missed by conventional techniques. Our evaluation shows that HPCAgentTester significantly improves test compilation rates and correctness compared to standalone LLMs, offering a more robust and scalable solution for ensuring the reliability of parallel software systems.", "authors": ["Rabimba Karanjai", "Lei Xu", "Weidong Shi"], "categories": ["cs.DC", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-13", "url": "https://arxiv.org/abs/2511.10860", "pdf_url": "https://arxiv.org/pdf/2511.10860v1", "arxiv_id": "2511.10860", "doi": "10.1109/AIware69974.2025.00031", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.237}
{"id": "a1ccc730d39aaea77a788bd288017d1d0a4a949691116fcd368909dfb4c5025c", "sources": ["arxiv", "semantic_scholar"], "title": "Behavior Modeling for Training-free Building of Private Domain Multi Agent System", "abstract": "The rise of agentic systems that combine orchestration, tool use, and conversational capabilities, has been more visible by the recent advent of large language models (LLMs). While open-domain frameworks exist, applying them in private domains remains difficult due to heterogeneous tool formats, domain-specific jargon, restricted accessibility of APIs, and complex governance. Conventional solutions, such as fine-tuning on synthetic dialogue data, are burdensome and brittle under domain shifts, and risk degrading general performance. In this light, we introduce a framework for private-domain multi-agent conversational systems that avoids training and data generation by adopting behavior modeling and documentation. Our design simply assumes an orchestrator, a tool-calling agent, and a general chat agent, with tool integration defined through structured specifications and domain-informed instructions. This approach enables scalable adaptation to private tools and evolving contexts without continual retraining. The framework supports practical use cases, including lightweight deployment of multi-agent systems, leveraging API specifications as retrieval resources, and generating synthetic dialogue for evaluation -- providing a sustainable method for aligning agent behavior with domain expertise in private conversational ecosystems.", "authors": ["Won Ik Cho", "Woonghee Han", "Kyung Seo Ki", "Young Min Kim"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-13", "url": "https://arxiv.org/abs/2511.10283", "pdf_url": "https://arxiv.org/pdf/2511.10283v1", "arxiv_id": "2511.10283", "doi": "10.48550/arXiv.2511.10283", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3724}
{"id": "024623795d63c4138d33d3a1f631c1ae8412e2950779da14847837181ce415c2", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond ReAct: A Planner-Centric Framework for Complex Tool-Augmented LLM Reasoning", "abstract": "Existing tool-augmented large language models (LLMs) encounter significant challenges when processing complex queries. Current frameworks such as ReAct are prone to local optimization traps due to their reliance on incremental decision-making processes. To address these limitations, we propose a novel Planner-centric Plan-Execute paradigm that fundamentally resolves local optimization bottlenecks through architectural innovation. Central to our approach is a novel Planner model that performs global Directed Acyclic Graph (DAG) planning for complex queries, enabling optimized execution beyond conventional tool coordination. We also introduce ComplexTool-Plan, a large-scale benchmark dataset featuring complex queries that demand sophisticated multi-tool composition and coordination capabilities. Additionally, we develop a two-stage training methodology that integrates Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO), systematically enhancing the Planner's tool selection accuracy and global planning awareness through structured DAG-based planning. When integrated with a capable executor, our framework achieves state-of-the-art performance on the StableToolBench benchmark for complex user queries, demonstrating superior end-to-end execution capabilities and robust handling of intricate multi-tool workflows.", "authors": ["Xiaolong Wei", "Yuehu Dong", "Xingliang Wang", "Xingyu Zhang", "Zhejun Zhao", "Dongdong Shen", "Long Xia", "Dawei Yin"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-13", "url": "https://arxiv.org/abs/2511.10037", "pdf_url": "https://arxiv.org/pdf/2511.10037v2", "arxiv_id": "2511.10037", "doi": "10.48550/arXiv.2511.10037", "citation_count": 14, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3724}
{"id": "488b33606ad8615d140f511f0c5a2ce36d40cf07367b99e6fad871ba6d354895", "sources": ["arxiv", "semantic_scholar"], "title": "Echoing: Identity Failures when LLM Agents Talk to Each Other", "abstract": "As large language model (LLM) based agents interact autonomously with one another, a new class of failures emerges that cannot be predicted from single agent performance: behavioral drifts in agent-agent conversations (AxA). Unlike human-agent interactions, where humans ground and steer conversations, AxA lacks such stabilizing signals, making these failures unique. We investigate one such failure, echoing, where agents abandon their assigned roles and instead mirror their conversational partners, undermining their intended objectives. Through experiments across $66$ AxA configurations, $4$ domains (3 transactional, 1 advisory), and $2500+$ conversations (over $250000$ LLM inferences), we show that echoing occurs across major LLM providers, with echoing rates as high as $70\\%$ depending on the model and domain. Moreover, we find that echoing is persistent even in advanced reasoning models with substantial rates ($32.8\\%$) that are not reduced by reasoning efforts. We analyze prompt, conversation dynamics, showing that echoing arises as interaction grows longer ($7+$ agent turns) and is not merely an artifact of sub-optimal experiment design. Finally, we introduce a protocol-level mitigation where targeted use of structured response reduces echoing to $9\\%$.", "authors": ["Sarath Shekkizhar", "Romain Cosentino", "Adam Earle", "Silvio Savarese"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-12", "url": "https://arxiv.org/abs/2511.09710", "pdf_url": "https://arxiv.org/pdf/2511.09710v3", "arxiv_id": "2511.09710", "doi": "10.48550/arXiv.2511.09710", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3713}
{"id": "0119a0e4d47774db288e490e39d8090a15e930cae2bf71d2fccf3857eecf3b3f", "sources": ["arxiv", "semantic_scholar"], "title": "Tele-LLM-Hub: Building Context-Aware Multi-Agent LLM Systems for Telecom Networks", "abstract": "This paper introduces Tele-LLM-Hub, a user friendly low-code solution for rapid prototyping and deployment of context aware multi-agent (MA) Large Language Model (LLM) systems tailored for 5G and beyond. As telecom wireless networks become increasingly complex, intelligent LLM applications must share a domainspecific understanding of network state. We propose TeleMCP, the Telecom Model Context Protocol, to enable structured and context-rich communication between agents in telecom environments. Tele-LLM-Hub actualizes TeleMCP through a low-code interface that supports agent creation, workflow composition, and interaction with software stacks such as srsRAN. Key components include a direct chat interface, a repository of pre-built systems, an Agent Maker leveraging finetuning with our RANSTRUCT framework, and an MA-Maker for composing MA workflows. The goal of Tele-LLM-Hub is to democratize the design of contextaware MA systems and accelerate innovation in next-generation wireless networks.", "authors": ["Pranshav Gajjar", "Cong Shen", "Vijay K Shah"], "categories": ["cs.NI", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-12", "url": "https://arxiv.org/abs/2511.09087", "pdf_url": "https://arxiv.org/pdf/2511.09087v2", "arxiv_id": "2511.09087", "doi": "10.48550/arXiv.2511.09087", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3713}
{"id": "7db19d854dffb007d933785aa3460bf5108ec42971d0f6a6c8ac344696e147d3", "sources": ["arxiv", "semantic_scholar"], "title": "Achieving Equilibrium under Utility Heterogeneity: An Agent-Attention Framework for Multi-Agent Multi-Objective Reinforcement Learning", "abstract": "Multi-agent multi-objective systems (MAMOS) have emerged as powerful frameworks for modelling complex decision-making problems across various real-world domains, such as robotic exploration, autonomous traffic management, and sensor network optimisation. MAMOS offers enhanced scalability and robustness through decentralised control and more accurately reflects inherent trade-offs between conflicting objectives. In MAMOS, each agent uses utility functions that map return vectors to scalar values. Existing MAMOS optimisation methods face challenges in handling heterogeneous objective and utility function settings, where training non-stationarity is intensified due to private utility functions and the associated policies. In this paper, we first theoretically prove that direct access to, or structured modeling of, global utility functions is necessary for the Bayesian Nash Equilibrium under decentralised execution constraints. To access the global utility functions while preserving the decentralised execution, we propose an Agent-Attention Multi-Agent Multi-Objective Reinforcement Learning (AA-MAMORL) framework. Our approach implicitly learns a joint belief over other agents' utility functions and their associated policies during centralised training, effectively mapping global states and utilities to each agent's policy. In execution, each agent independently selects actions based on local observations and its private utility function to approximate a BNE, without relying on inter-agent communication. We conduct comprehensive experiments in both a custom-designed MAMO Particle environment and the standard MOMALand benchmark. The results demonstrate that access to global preferences and our proposed AA-MAMORL significantly improve performance and consistently outperform state-of-the-art methods.", "authors": ["Zhuhui Li", "Chunbo Luo", "Liming Huang", "Luyu Qi", "Geyong Min"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-12", "url": "https://arxiv.org/abs/2511.08926", "pdf_url": "https://arxiv.org/pdf/2511.08926v1", "arxiv_id": "2511.08926", "doi": "10.48550/arXiv.2511.08926", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3713}
{"id": "f6723d53317c4abce049ae651fe35fe34814718d094a305e0bef606b306a587c", "sources": ["arxiv", "semantic_scholar"], "title": "Can LLM Agents Really Debate? A Controlled Study of Multi-Agent Debate in Logical Reasoning", "abstract": "Multi-agent debate (MAD) has recently emerged as a promising framework for improving the reasoning performance of large language models (LLMs). Yet, whether LLM agents can genuinely engage in deliberative reasoning, beyond simple ensembling or majority voting, remains unclear. We address this question through a controlled study using the Knight--Knave--Spy logic puzzle, which enables precise, step-wise evaluation of debate outcomes and processes under verifiable ground truth. We systematically set up six structural and cognitive factors, including agent team size, composition, confidence visibility, debate order, debate depth, and task difficulty, to disentangle their respective effects on collective reasoning. Our results show that intrinsic reasoning strength and group diversity are the dominant drivers of debate success, while structural parameters such as order or confidence visibility offer limited gains. Beyond outcomes, process-level analyses identify key behavioral patterns: majority pressure suppresses independent correction, effective teams overturn incorrect consensus, and rational, validity-aligned reasoning most strongly predicts improvement. These findings provide valuable insights into how and why LLM debates succeed or fail, offering guidance for designing interpretable and truth-seeking multi-agent reasoning systems.", "authors": ["Haolun Wu", "Zhenkun Li", "Lingyao Li"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-11", "url": "https://arxiv.org/abs/2511.07784", "pdf_url": "https://arxiv.org/pdf/2511.07784v1", "arxiv_id": "2511.07784", "doi": "10.48550/arXiv.2511.07784", "citation_count": 10, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3701}
{"id": "13903ea2407ba4f553be2c0e6da69986728bb40c45974efdcd80440ecf74dabe", "sources": ["arxiv", "semantic_scholar"], "title": "Who Gets the Reward, Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents", "abstract": "Large Language Models (LLMs) in multi-agent systems (MAS) have shown promise for complex tasks, yet current training methods lack principled ways to connect system-level evaluation with agent-level and message-level learning. We propose a theoretical framework that unifies cooperative game-theoretic attribution with process reward modeling to transform system evaluation into agent credit and then into response-level signals. Unlike prior approaches that rely only on attribution (e.g., Shapley) or step-level labels (e.g., PRM), our method produces local, signed, and credit-conserving signals. In success cases, Shapley-based credit assignment fairly allocates outcomes across agents and is refined into per-message rewards that promote cooperation while discouraging redundancy or sabotage. In failure cases, first-error localization yields repair-aware preferences that penalize harmful steps while rewarding corrective attempts. The resulting signals are bounded, cooperative, and directly compatible with reinforcement-based or preference-based post-training, providing a unified and auditable pathway from global evaluation to local supervision in LLM multi-agent training. Our contribution is conceptual: we present a theoretical foundation and training signals, leaving empirical validation for future work.", "authors": ["Chih-Hsuan Yang", "Tanwi Mallick", "Le Chen", "Krishnan Raghavan", "Azton Wells", "Amal Gueroudji", "Ian T. Foster", "Rajeev Thakur"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-11", "url": "https://arxiv.org/abs/2511.10687", "pdf_url": "https://arxiv.org/pdf/2511.10687v2", "arxiv_id": "2511.10687", "doi": "10.48550/arXiv.2511.10687", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3701}
{"id": "73e5e2732ada7646b9c9b3fe7e3117a41cc0105b22286d0a2db6195793e75b8f", "sources": ["arxiv", "semantic_scholar"], "title": "Structured Uncertainty guided Clarification for LLM Agents", "abstract": "LLM agents with tool-calling capabilities often fail when user instructions are ambiguous or incomplete, leading to incorrect invocations and task failures. Existing approaches operate in unstructured language spaces, generating clarifying questions through prompting strategies that lack principled criteria for determining which questions to ask and when to stop. We introduce a principled formulation of structured uncertainty that operates directly over tool parameters and their domains, cleanly separating specification uncertainty (what the user wants) from model uncertainty (what the LLM predicts). Our formulation uses Expected Value of Perfect Information (EVPI) to quantify the disambiguation value of each potential question, balanced against aspect-based cost modeling that prevents redundant questioning. We demonstrate the versatility of this formulation through two applications. First, SAGE-Agent uses structured uncertainty for inference-time question selection, achieving 7-39% higher coverage on ambiguous tasks while reducing clarification questions by 1.5-2.7x compared to strong prompting and uncertainty-based baselines. Second, we show that structured uncertainty provides effective training signals: uncertainty-guided reward modeling boosts When2Call accuracy from 36.5% to 65.2% (3B model) and 36.7% to 62.9% (7B model) through uncertainty-weighted GRPO training, demonstrating more sample-efficient reinforcement learning for tool-calling agents. To enable evaluation, we present ClarifyBench, the first multi-turn dynamic tool-calling disambiguation benchmark. Our results establish structured uncertainty as a principled framework that improves both inference-time interaction efficiency and training-time sample efficiency in tool-augmented agents.", "authors": ["Manan Suri", "Puneet Mathur", "Nedim Lipka", "Franck Dernoncourt", "Ryan A. Rossi", "Dinesh Manocha"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-11", "url": "https://arxiv.org/abs/2511.08798", "pdf_url": "https://arxiv.org/pdf/2511.08798v2", "arxiv_id": "2511.08798", "doi": "10.48550/arXiv.2511.08798", "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3701}
{"id": "3fb0c91ef680bff7ed45d741686e286220776494436b62669a3fd60a0fe0a534", "sources": ["arxiv", "semantic_scholar"], "title": "Efficient LLM Safety Evaluation through Multi-Agent Debate", "abstract": "Safety evaluation of large language models (LLMs) increasingly relies on LLM-as-a-judge pipelines, but strong judges can still be expensive to use at scale. We study whether structured multi-agent debate can improve judge reliability while keeping backbone size and cost modest. To do so, we introduce HAJailBench, a human-annotated jailbreak benchmark with 11,100 labeled interactions spanning diverse attack methods and target models, and we pair it with a Multi-Agent Judge framework in which critic, defender, and judge agents debate under a shared safety rubric. On HAJailBench, the framework improves over matched small-model prompt baselines and prior multi-agent judges, while remaining more economical than GPT-4o under the evaluated pricing snapshot. Ablation results further show that a small number of debate rounds is sufficient to capture most of the gain. Together, these results support structured, value-aligned debate as a practical design for scalable LLM safety evaluation.", "authors": ["Dachuan Lin", "Guobin Shen", "Zihao Yang", "Tianrong Liu", "Dongcheng Zhao", "Yi Zeng"], "categories": ["cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-09", "url": "https://arxiv.org/abs/2511.06396", "pdf_url": "https://arxiv.org/pdf/2511.06396v3", "arxiv_id": "2511.06396", "doi": "10.48550/arXiv.2511.06396", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3678}
{"id": "cf6532bbc7d3dac94c541c3885865466c16e2e5d07b0d6019b5ffdf7c9186a7f", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Guided Reinforcement Learning with Representative Agents for Traffic Modeling", "abstract": "Large language models (LLMs) are increasingly used as behavioral proxies for self-interested travelers in agent-based traffic models. Although more flexible and generalizable than conventional models, the practical use of these approaches remains limited by scalability due to the cost of calling one LLM for every traveler. Moreover, it has been found that LLM agents often make opaque choices and produce unstable day-to-day dynamics. To address these challenges, we propose to model each homogeneous traveler group facing the same decision context with a single representative LLM agent who behaves like the population's average, maintaining and updating a mixed strategy over routes that coincides with the group's aggregate flow proportions. Each day, the LLM reviews the travel experience and flags routes with positive reinforcement that they hope to use more often, and an interpretable update rule then converts this judgment into strategy adjustments using a tunable (progressively decaying) step size. The representative-agent design improves scalability, while the separation of reasoning from updating clarifies the decision logic while stabilizing learning. In classic traffic assignment settings, we find that the proposed approach converges rapidly to the user equilibrium. In richer settings with income heterogeneity, multi-criteria costs, and multi-modal choices, the generated dynamics remain stable and interpretable, reproducing plausible behavioral patterns well-documented in psychology and economics, for example, the decoy effect in toll versus non-toll road selection, and higher willingness-to-pay for convenience among higher-income travelers when choosing between driving, transit, and park-and-ride options.", "authors": ["Hanlin Sun", "Jiayang Li"], "categories": ["cs.GT", "cs.AI", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-11-09", "url": "https://arxiv.org/abs/2511.06260", "pdf_url": "https://arxiv.org/pdf/2511.06260v1", "arxiv_id": "2511.06260", "doi": "10.48550/arXiv.2511.06260", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3678}
{"id": "2f9513394eac782d355c264c659ab18cf10e663d64df918873beba969efdf8b4", "sources": ["arxiv", "semantic_scholar"], "title": "Maestro: Learning to Collaborate via Conditional Listwise Policy Optimization for Multi-Agent LLMs", "abstract": "Multi-agent systems (MAS) built on Large Language Models (LLMs) are being used to approach complex problems and can surpass single model inference. However, their success hinges on navigating a fundamental cognitive tension: the need to balance broad, divergent exploration of the solution space with a principled, convergent synthesis to the optimal solution. Existing paradigms often struggle to manage this duality, leading to premature consensus, error propagation, and a critical credit assignment problem that fails to distinguish between genuine reasoning and superficially plausible arguments. To resolve this core challenge, we propose the Multi-Agent Exploration-Synthesis framework Through Role Orchestration (Maestro), a principled paradigm for collaboration that structurally decouples these cognitive modes. Maestro uses a collective of parallel Execution Agents for diverse exploration and a specialized Central Agent for convergent, evaluative synthesis. To operationalize this critical synthesis phase, we introduce Conditional Listwise Policy Optimization (CLPO), a reinforcement learning objective that disentangles signals for strategic decisions and tactical rationales. By combining decision-focused policy gradients with a list-wise ranking loss over justifications, CLPO achieves clean credit assignment and stronger comparative supervision. Experiments on mathematical reasoning and general problem-solving benchmarks demonstrate that Maestro, coupled with CLPO, consistently outperforms existing state-of-the-art multi-agent approaches, delivering absolute accuracy gains of 6% on average and up to 10% at best.", "authors": ["Wei Yang", "Jiacheng Pang", "Shixuan Li", "Paul Bogdan", "Stephen Tu", "Jesse Thomason"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-08", "url": "https://arxiv.org/abs/2511.06134", "pdf_url": "https://arxiv.org/pdf/2511.06134v1", "arxiv_id": "2511.06134", "doi": "10.48550/arXiv.2511.06134", "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3667}
{"id": "4d43f1d87a3bae8d415896a41c9f7b71a6087efc4e4ea38554315435d8e68bf4", "sources": ["arxiv", "semantic_scholar"], "title": "TAMAS: Benchmarking Adversarial Risks in Multi-Agent LLM Systems", "abstract": "Large Language Models (LLMs) have demonstrated strong capabilities as autonomous agents through tool use, planning, and decision-making abilities, leading to their widespread adoption across diverse tasks. As task complexity grows, multi-agent LLM systems are increasingly used to solve problems collaboratively. However, safety and security of these systems remains largely under-explored. Existing benchmarks and datasets predominantly focus on single-agent settings, failing to capture the unique vulnerabilities of multi-agent dynamics and co-ordination. To address this gap, we introduce $\\textbf{T}$hreats and $\\textbf{A}$ttacks in $\\textbf{M}$ulti-$\\textbf{A}$gent $\\textbf{S}$ystems ($\\textbf{TAMAS}$), a benchmark designed to evaluate the robustness and safety of multi-agent LLM systems. TAMAS includes five distinct scenarios comprising 300 adversarial instances across six attack types and 211 tools, along with 100 harmless tasks. We assess system performance across ten backbone LLMs and three agent interaction configurations from Autogen and CrewAI frameworks, highlighting critical challenges and failure modes in current multi-agent deployments. Furthermore, we introduce Effective Robustness Score (ERS) to assess the tradeoff between safety and task effectiveness of these frameworks. Our findings show that multi-agent systems are highly vulnerable to adversarial attacks, underscoring the urgent need for stronger defenses. TAMAS provides a foundation for systematically studying and improving the safety of multi-agent LLM systems.", "authors": ["Ishan Kavathekar", "Hemang Jain", "Ameya Rathod", "Ponnurangam Kumaraguru", "Tanuja Ganu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-07", "url": "https://arxiv.org/abs/2511.05269", "pdf_url": "https://arxiv.org/pdf/2511.05269v1", "arxiv_id": "2511.05269", "doi": "10.48550/arXiv.2511.05269", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3655}
{"id": "0d7c13b341fba298cd8819b52e5699a47ceb250803b8e04cd3d81906c8ab0e3f", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaborative Framework For Math Problem Generation", "abstract": "Automatic question generation (AQG) for mathematics education remains an elusive goal for Intelligent Tutoring Systems and educators. While pre-trained transformer-based language models have significantly advanced natural language generation, they often struggle to precisely control problem complexity and cognitive demands. In this paper, we introduce a collaborative multi-agent framework as a novel method of incorporating inference-time computation into AQG. This approach leverages multiple agents that iteratively refine generated question-answer pairs to better balance complexity and cognitive demand. We evaluate the generated questions on five meta-evaluation criteria: relevance, importance, clarity, difficulty matching, answerability, to assess the system's ability to control the required complexity and quality of the questions. Preliminary evaluations show that this collaborative multi-agent framework elevates the quality of generated educational content by fostering a more nuanced balance between cognitive challenge and clarity. These promising outcomes suggest that integrating collaborative multi-agent workflows can yield more controlled, pedagogically valuable content that can help advance automated educational content generation and adaptive learning environments.", "authors": ["Kia Karbasi", "Kevin Hong", "Mohammad Amin Samadi", "Gregory Pottie"], "categories": ["cs.MA", "cs.CL", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-06", "url": "https://arxiv.org/abs/2511.03958", "pdf_url": "https://arxiv.org/pdf/2511.03958v1", "arxiv_id": "2511.03958", "doi": "10.5281/zenodo.15870246", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Educational Data Mining", "quality_score": 0.3644}
{"id": "62afe6c480fc7a66e77a63bc35d99ed6c469905ddfffe07dd10293b949d065b5", "sources": ["arxiv", "semantic_scholar"], "title": "ArchPilot: A Proxy-Guided Multi-Agent Approach for Machine Learning Engineering", "abstract": "Recent LLM-based agents have demonstrated strong capabilities in automated ML engineering. However, they heavily rely on repeated full training runs to evaluate candidate solutions, resulting in significant computational overhead, limited scalability to large search spaces, and slow iteration cycles. To address these challenges, we introduce ArchPilot, a multi-agent system that integrates architecture generation, proxy-based evaluation, and adaptive search into a unified framework. ArchPilot consists of three specialized agents: an orchestration agent that coordinates the search process using a Monte Carlo Tree Search (MCTS)-inspired novel algorithm with a restart mechanism and manages memory of previous candidates; a generation agent that iteratively generates, improves, and debugs candidate architectures; and an evaluation agent that executes proxy training runs, generates and optimizes proxy functions, and aggregates the proxy scores into a fidelity-aware performance metric. This multi-agent collaboration allows ArchPilot to prioritize high-potential candidates with minimal reliance on expensive full training runs, facilitating efficient ML engineering under limited budgets. Experiments on MLE-Bench demonstrate that ArchPilot outperforms SOTA baselines such as AIDE and ML-Master, validating the effectiveness of our multi-agent system.", "authors": ["Zhuowen Yuan", "Tao Liu", "Yang Yang", "Yang Wang", "Feng Qi", "Kaushik Rangadurai", "Bo Li", "Shuang Yang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-06", "url": "https://arxiv.org/abs/2511.03985", "pdf_url": "https://arxiv.org/pdf/2511.03985v1", "arxiv_id": "2511.03985", "doi": "10.48550/arXiv.2511.03985", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3644}
{"id": "677e6d83ead037bb1e6dff4ae1d52a8653978c50ccee7413afa04040df64465f", "sources": ["arxiv", "semantic_scholar"], "title": "RefAgent: A Multi-agent LLM-based Framework for Automatic Software Refactoring", "abstract": "Large Language Models (LLMs) have substantially influenced various software engineering tasks. Indeed, in the case of software refactoring, traditional LLMs have shown the ability to reduce development time and enhance code quality. However, these LLMs often rely on static, detailed instructions for specific tasks. In contrast, LLM-based agents can dynamically adapt to evolving contexts and autonomously make decisions by interacting with software tools and executing workflows. In this paper, we explore the potential of LLM-based agents in supporting refactoring activities. Specifically, we introduce RefAgent, a multi-agent LLM-based framework for end-to-end software refactoring. RefAgent consists of specialized agents responsible for planning, executing, testing, and iteratively refining refactorings using self-reflection and tool-calling capabilities. We evaluate RefAgent on eight open-source Java projects, comparing its effectiveness against a single-agent approach, a search-based refactoring tool, and historical developer refactorings. Our assessment focuses on: (1) the impact of generated refactorings on software quality, (2) the ability to identify refactoring opportunities, and (3) the contribution of each LLM agent through an ablation study. Our results show that RefAgent achieves a median unit test pass rate of 90%, reduces code smells by a median of 52.5%, and improves key quality attributes (e.g., reusability) by a median of 8.6%. Additionally, it closely aligns with developer refactorings and the search-based tool in identifying refactoring opportunities, attaining a median F1-score of 79.15% and 72.7%, respectively. Compared to single-agent approaches, RefAgent improves the median unit test pass rate by 64.7% and the median compilation success rate by 40.1%. These findings highlight the promise of multi-agent architectures in advancing automated software refactoring.", "authors": ["Khouloud Oueslati", "Maxime Lamothe", "Foutse Khomh"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-05", "url": "https://arxiv.org/abs/2511.03153", "pdf_url": "https://arxiv.org/pdf/2511.03153v2", "arxiv_id": "2511.03153", "doi": "10.48550/arXiv.2511.03153", "citation_count": 5, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5614}
{"id": "183060fb98190a593870306ddcec8235c53c929c262e2507b4fa513e3bc104c3", "sources": ["arxiv", "semantic_scholar"], "title": "ALAS: Transactional and Dynamic Multi-Agent LLM Planning", "abstract": "Large language models enable flexible multi-agent planning but remain fragile in practice: verification is often circular, state changes are not tracked for repair, and small faults trigger costly global recomputation. We present ALAS, a stateful, disruption-aware framework that separates planning from non-circular validation, records a versioned execution log for grounded checks and restore points, and performs localized repair that preserves work in progress. The validator operates independently of the planning LLM with fresh, bounded context, avoiding self-check loops and mid-context attrition. The repair protocol edits only the minimal affected region under explicit policies (retry, catch, timeout, backoff, idempotency keys, compensation, loop guards) defined in a canonical workflow IR that maps to Amazon States Language and Argo Workflows. On job-shop scheduling suites (DMU, TA) across five classical benchmarks, ALAS matches or exceeds strong single-LLM and multi-agent baselines, achieving 83.7% success, reducing token usage by 60%, and running 1.82times faster under comparable settings. A minimal reliability study shows that the validator detects injected structural faults with low overhead, and that localized repair contains runtime perturbations with a bounded edit radius and less makespan degradation than global recompute. Results indicate that the combination of validator isolation, versioned execution logs, and localized repair provides measurable efficiency, feasibility, and scalability for multi-agent LLM planning. Code and seeds will be released.", "authors": ["Longling Geng", "Edward Y. Chang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-05", "url": "https://arxiv.org/abs/2511.03094", "pdf_url": "https://arxiv.org/pdf/2511.03094v1", "arxiv_id": "2511.03094", "doi": "10.48550/arXiv.2511.03094", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3632}
{"id": "05ab352d16cc13f8040c9f5f058f1883bdc88874f0e84052835d4a8c59e61631", "sources": ["arxiv", "semantic_scholar"], "title": "Unlocking the Power of Multi-Agent LLM for Reasoning: From Lazy Agents to Deliberation", "abstract": "Large Language Models (LLMs) trained with reinforcement learning and verifiable rewards have achieved strong results on complex reasoning tasks. Recent work extends this paradigm to a multi-agent setting, where a meta-thinking agent proposes plans and monitors progress while a reasoning agent executes subtasks through sequential conversational turns. Despite promising performance, we identify a critical limitation: lazy agent behavior, in which one agent dominates while the other contributes little, undermining collaboration and collapsing the setup to an ineffective single agent. In this paper, we first provide a theoretical analysis showing why lazy behavior naturally arises in multi-agent reasoning. We then introduce a stable and efficient method for measuring causal influence, helping mitigate this issue. Finally, as collaboration intensifies, the reasoning agent risks getting lost in multi-turn interactions and trapped by previous noisy responses. To counter this, we propose a verifiable reward mechanism that encourages deliberation by allowing the reasoning agent to discard noisy outputs, consolidate instructions, and restart its reasoning process when necessary. Extensive experiments demonstrate that our framework alleviates lazy agent behavior and unlocks the full potential of multi-agent framework for complex reasoning tasks.", "authors": ["Zhiwei Zhang", "Xiaomin Li", "Yudi Lin", "Hui Liu", "Ramraj Chandradevan", "Linlin Wu", "Minhua Lin", "Fali Wang", "Xianfeng Tang", "Qi He", "Suhang Wang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.02303", "pdf_url": "https://arxiv.org/pdf/2511.02303v1", "arxiv_id": "2511.02303", "doi": "10.48550/arXiv.2511.02303", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3621}
{"id": "a5de215da041e423a77f018fc906faf4116d181292f2d28958a45e6e0ae00b79", "sources": ["arxiv", "semantic_scholar"], "title": "Controlling Performance and Budget of a Centralized Multi-agent LLM System with Reinforcement Learning", "abstract": "Large language models (LLMs) exhibit complementary strengths across domains and come with varying inference costs, motivating the design of multi-agent LLM systems where specialized models collaborate efficiently. Existing approaches predominantly rely on decentralized frameworks, which invoke multiple LLMs for every input and thus lead to substantial and uncontrolled inference costs. In this work, we introduce a centralized multi-LLM framework, where a controller LLM selectively coordinates a pool of expert models in a cost-efficient and cost-controllable manner. We formulate this coordination problem as reinforcement learning with dual objectives: maximizing task performance while minimizing the overall inference cost. In addition, we expect the multi-agent system to have adapted behavior with different budget conditions during inference. To this end, we propose CoRL, a reinforcement learning framework that optimizes the performance cost trade-off in a controllable multi-budget setting. Experiments on four diverse benchmarks demonstrate that CoRL enables a single system to surpass the best expert LLM under high-budget settings, while maintaining strong performance in more economical low-budget modes, highlighting the effectiveness of centralized coordination for scalable and cost-efficient multi-agent LLM systems.", "authors": ["Bowen Jin", "TJ Collins", "Donghan Yu", "Mert Cemri", "Shenao Zhang", "Mengyu Li", "Jay Tang", "Tian Qin", "Zhiyang Xu", "Jiarui Lu", "Guoli Yin", "Jiawei Han", "Zirui Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.02755", "pdf_url": "https://arxiv.org/pdf/2511.02755v1", "arxiv_id": "2511.02755", "doi": "10.48550/arXiv.2511.02755", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3621}
{"id": "4ebdb60cfd6770c46b9a521364985b51930e17c48743c42a19f72a3a1acf0c8f", "sources": ["arxiv", "semantic_scholar"], "title": "Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Live", "abstract": "KV cache management is essential for efficient LLM inference. To maximize utilization, existing inference engines evict finished requests' KV cache if new requests are waiting. This policy breaks for agentic workloads, which interleave LLM calls with tools, introducing pauses that prevent effective KV reuse across turns. Since many tool calls have much shorter durations than human response multi-turn chatbot, it would be promising to retain the KV cache in during these tools. However, many challenges remain. First, we need to consider both the potential cost of recomputation or reloading (if offloading enabled) as well as the increasing queueing delays after eviction from GPU. Second, due to the internal variance of tool call durations, the method needs to remain robust under limited predictability of tool call durations. We present Continuum, a serving system to optimize job completion time for multi-turn agent workloads by introducing time-to-live mechanism for KV cache retention. For requests that generate tool calls, Continuum selectively pins the KV cache in GPU memory with a time-to-live value determined by the reload cost and potential queueing delay induced by eviction. When the TTL expires, the KV cache can be automatically evicted to free up GPU memory, providing robust performance under edge cases. When combined with program-level first-come-first-serve, Continuum preserves multi-turn continuity, and reduces delay for agentic workflows. Evaluations on real-world agents (SWE-Bench, BFCL, OpenHand) with Llama-3.1 8B/70B, Gemma-3 12B, and GLM-4.5 355B shows that Continuum improves the average job completion times by over 8x while improving throughput.", "authors": ["Hanchen Li", "Runyuan He", "Qiuyang Mang", "Qizheng Zhang", "Huanzhi Mao", "Xiaokun Chen", "Hangrui Zhou", "Alvin Cheung", "Joseph Gonzalez", "Ion Stoica"], "categories": ["cs.OS", "cs.AI", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.02230", "pdf_url": "https://arxiv.org/pdf/2511.02230v6", "arxiv_id": "2511.02230", "doi": "10.48550/arXiv.2511.02230", "citation_count": 31, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "161e4d1803763fa088ab1bf61b5b87cd6eaf5669ac4ba7a047614e9091e5fe0d", "sources": ["arxiv", "semantic_scholar"], "title": "PublicAgent: Multi-Agent Design Principles From an LLM-Based Open Data Analysis Framework", "abstract": "Open data repositories hold potential for evidence-based decision-making, yet are inaccessible to non-experts lacking expertise in dataset discovery, schema mapping, and statistical analysis. Large language models show promise for individual tasks, but end-to-end analytical workflows expose fundamental limitations: attention dilutes across growing contexts, specialized reasoning patterns interfere, and errors propagate undetected. We present PublicAgent, a multi-agent framework that addresses these limitations through decomposition into specialized agents for intent clarification, dataset discovery, analysis, and reporting. This architecture maintains focused attention within agent contexts and enables validation at each stage. Evaluation across five models and 50 queries derives five design principles for multi-agent LLM systems. First, specialization provides value independent of model strength--even the strongest model shows 97.5% agent win rates, with benefits orthogonal to model scale. Second, agents divide into universal (discovery, analysis) and conditional (report, intent) categories. Universal agents show consistent effectiveness (std dev 12.4%) while conditional agents vary by model (std dev 20.5%). Third, agents mitigate distinct failure modes--removing discovery or analysis causes catastrophic failures (243-280 instances), while removing report or intent causes quality degradation. Fourth, architectural benefits persist across task complexity with stable win rates (86-92% analysis, 84-94% discovery), indicating workflow management value rather than reasoning enhancement. Fifth, wide variance in agent effectiveness across models (42-96% for analysis) requires model-aware architecture design. These principles guide when and why specialization is necessary for complex analytical workflows while enabling broader access to public data through natural language interfaces.", "authors": ["Sina Montazeri", "Yunhe Feng", "Kewei Sha"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.03023", "pdf_url": "https://arxiv.org/pdf/2511.03023v1", "arxiv_id": "2511.03023", "doi": "10.48550/arXiv.2511.03023", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3621}
{"id": "96eebb26bccfd44c9546bec3822d2d47da0e9191acc1f3cfacde662b88161bc8", "sources": ["arxiv", "semantic_scholar"], "title": "Optimal-Agent-Selection: State-Aware Routing Framework for Efficient Multi-Agent Collaboration", "abstract": "The emergence of multi-agent systems powered by large language models (LLMs) has unlocked new frontiers in complex task-solving, enabling diverse agents to integrate unique expertise, collaborate flexibly, and address challenges unattainable for individual models. However, the full potential of such systems is hindered by rigid agent scheduling and inefficient coordination strategies that fail to adapt to evolving task requirements. In this paper, we propose STRMAC, a state-aware routing framework designed for efficient collaboration in multi-agent systems. Our method separately encodes interaction history and agent knowledge to power the router, which adaptively selects the most suitable single agent at each step for efficient and effective collaboration. Furthermore, we introduce a self-evolving data generation approach that accelerates the collection of high-quality execution paths for efficient system training. Experiments on challenging collaborative reasoning benchmarks demonstrate that our method achieves state-of-the-art performance, achieving up to 23.8% improvement over baselines and reducing data collection overhead by up to 90.1% compared to exhaustive search.", "authors": ["Jingbo Wang", "Sendong Zhao", "Haochun Wang", "Yuzheng Fan", "Lizhe Zhang", "Yan Liu", "Ting Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.02200", "pdf_url": "https://arxiv.org/pdf/2511.02200v1", "arxiv_id": "2511.02200", "doi": "10.48550/arXiv.2511.02200", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3621}
{"id": "a5cb92cdc2f8caec1d0e5eebb49d58e5781dd7cab481f8ec63131d0555b6ddd0", "sources": ["arxiv", "semantic_scholar"], "title": "CostBench: Evaluating Multi-Turn Cost-Optimal Planning and Adaptation in Dynamic Environments for LLM Tool-Use Agents", "abstract": "Current evaluations of Large Language Model (LLM) agents primarily emphasize task completion, often overlooking resource efficiency and adaptability. This neglects a crucial capability: agents' ability to devise and adjust cost-optimal plans in response to changing environments. To bridge this gap, we introduce CostBench, a scalable, cost-centric benchmark designed to evaluate agents' economic reasoning and replanning abilities. Situated in the travel-planning domain, CostBench comprises tasks solvable via multiple sequences of atomic and composite tools with diverse, customizable costs. It also supports four types of dynamic blocking events, such as tool failures and cost changes, to simulate real-world unpredictability and necessitate agents to adapt in real time. Evaluating leading open-sourced and proprietary models on CostBench reveals a substantial gap in cost-aware planning: agents frequently fail to identify cost-optimal solutions in static settings, with even GPT-5 achieving less than 75% exact match rate on the hardest tasks, and performance further dropping by around 40% under dynamic conditions. By diagnosing these weaknesses, CostBench lays the groundwork for developing future agents that are both economically rational and robust.", "authors": ["Jiayu Liu", "Cheng Qian", "Zhaochen Su", "Qing Zong", "Shijue Huang", "Bingxiang He", "Yi R. Fung"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-04", "url": "https://arxiv.org/abs/2511.02734", "pdf_url": "https://arxiv.org/pdf/2511.02734v2", "arxiv_id": "2511.02734", "doi": "10.48550/arXiv.2511.02734", "citation_count": 18, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5596}
{"id": "7a1ff4a2edf696474b5c86d2e73d0de51bea22655ab0054b11000e6d41e71684", "sources": ["arxiv", "semantic_scholar"], "title": "Tool-to-Agent Retrieval: Bridging Tools and Agents for Scalable LLM Multi-Agent Systems", "abstract": "Recent advances in LLM Multi-Agent Systems enable scalable orchestration of sub-agents, each coordinating hundreds or thousands of tools or Model Context Protocol (MCP) servers. However, existing retrieval methods typically match queries against coarse agent-level descriptions before routing, which obscures fine-grained tool functionality and often results in suboptimal agent selection. We introduce Tool-to-Agent Retrieval, a unified framework that embeds both tools and their parent agents in a shared vector space and connects them through metadata relationships. By explicitly representing tool capabilities and traversing metadata to the agent level, Tool-to-Agent Retrieval enables granular tool-level or agent-level retrieval, ensuring that agents and their underlying tools or MCP servers are equally represented without the context dilution that arises from chunking many tools together. Evaluating Tool-to-Agent Retrieval across eight embedding models, our approach achieves consistent improvements of 19.4% in Recall@5 and 17.7% in nDCG@5 over previous state-of-the-art agent retrievers on the LiveMCPBench benchmark.", "authors": ["Elias Lumer", "Faheem Nizar", "Anmol Gulati", "Pradeep Honaganahalli Basavaraju", "Vamse Kumar Subbiah"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-03", "url": "https://arxiv.org/abs/2511.01854", "pdf_url": "https://arxiv.org/pdf/2511.01854v2", "arxiv_id": "2511.01854", "doi": "10.48550/arXiv.2511.01854", "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3609}
{"id": "15e1a44ac5d4b43b21148c396ea9285fe825ac9ca10d8986214b5a3d2a222a5e", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling Graph Chain-of-Thought Reasoning: A Multi-Agent Framework with Efficient LLM Serving", "abstract": "Graph Chain-of-Thought (Graph-CoT) enables large language models (LLMs) to perform step-by-step reasoning over graph-structured knowledge, but existing pipelines suffer from low accuracy, excessive token usage, high latency, and low throughput due to single-agent monolithic prompts, repeated context re-encoding, and inefficient serving execution. We present GLM, the first multi-agent Graph-CoT system co-designed with an optimized LLM serving architecture. GLM decomposes reasoning into specialized agents for classification, reasoning, action generation, and graph retrieval, enabling branching and selective context sharing to reduce prompt length and reasoning iterations while preserving reasoning quality, thereby improving accuracy and reducing overall token consumption. To scale inference, we introduce a Graph-CoT-aware LLM inference mechanism with graph-specific KV-cache management, priority-based eviction, and pipelined execution to improve serving efficiency. Experiments demonstrate that GLM improves answer accuracy by up to 38%, reduces token cost by up to 95.7%, lowers inference latency by 90.3%, and achieves up to 15.1x higher throughput compared to state-of-the-art Graph-CoT baselines, enabling efficient adoption for complex real-world reasoning at scale.", "authors": ["Chengying Huan", "Ziheng Meng", "Yongchao Liu", "Zhengyi Yang", "Yun Zhu", "Yue Yun", "Shipeng Li", "Rong Gu", "Xiabao Wu", "Haitao Zhang", "Chuntao Hong", "Shaonan Ma", "Guihai Chen", "Chen Tian"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-03", "url": "https://arxiv.org/abs/2511.01633", "pdf_url": "https://arxiv.org/pdf/2511.01633v1", "arxiv_id": "2511.01633", "doi": "10.48550/arXiv.2511.01633", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3609}
{"id": "5807781456cf5a5837d4c8cb27981d8f3d103ddfd397f2dbca279549a4f2a99f", "sources": ["arxiv", "semantic_scholar"], "title": "HAFixAgent: History-Aware Program Repair Agent", "abstract": "Automated program repair (APR) has recently shifted toward large language models and agent-based systems, yet most systems rely on local snapshot context, overlooking repository history. Prior work shows that repository history helps repair single-line bugs, since the last commit touching the buggy line is often the bug-introducing one. In this paper, we investigate whether repository history can also improve agentic APR systems at scale, especially for complex multi-hunk bugs. We present HAFixAgent, a History-Aware Bug-Fixing Agent that injects blame-derived repository heuristics into its repair loop. A preliminary study on 854 Defects4J (Java) and 501 BugsInPy (Python) bugs motivates our design, showing that bug-relevant history is widely available across both benchmarks. Using the same LLM (DeepSeek-V3.2-Exp) for all experiments, including replicated baselines, we show: (1) Effectiveness: HAFixAgent outperforms RepairAgent (+56.6\\%) and BIRCH-feedback (+47.1\\%) on Defects4J. Historical context further improves repair by +4.4\\% on Defects4J and +38.6\\% on BugsInPy, especially on single-file multi-hunk (SFMH) bugs. (2) Robustness: under noisy fault localization (+1/+3/+5 line shifts), history provides increasing resilience, maintaining 40 to 56\\% success on SFMH bugs where the non-history baseline collapses to 0\\%. (3) Efficiency: history does not significantly increase agent steps or token costs on either benchmark.", "authors": ["Yu Shi", "Hao Li", "Bram Adams", "Ahmed E. Hassan"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-02", "url": "https://arxiv.org/abs/2511.01047", "pdf_url": "https://arxiv.org/pdf/2511.01047v3", "arxiv_id": "2511.01047", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.229}
{"id": "cd5f241d8308f294031b9566d0160e47115a8d3e7579eac52a15b1e167537279", "sources": ["arxiv", "semantic_scholar"], "title": "AgentGit: A Version Control Framework for Reliable and Scalable LLM-Powered Multi-Agent Systems", "abstract": "With the rapid progress of large language models (LLMs), LLM-powered multi-agent systems (MAS) are drawing increasing interest across academia and industry. However, many current MAS frameworks struggle with reliability and scalability, especially on complex tasks. We present AgentGit, a framework that brings Git-like rollback and branching to MAS workflows. Built as an infrastructure layer on top of LangGraph, AgentGit supports state commit, revert, and branching, allowing agents to traverse, compare, and explore multiple trajectories efficiently. To evaluate AgentGit, we designed an experiment that optimizes target agents by selecting better prompts. We ran a multi-step A/B test against three baselines -- LangGraph, AutoGen, and Agno -- on a real-world task: retrieving and analyzing paper abstracts. Results show that AgentGit significantly reduces redundant computation, lowers runtime and token usage, and supports parallel exploration across multiple branches, enhancing both reliability and scalability in MAS development. This work offers a practical path to more robust MAS design and enables error recovery, safe exploration, iterative debugging, and A/B testing in collaborative AI systems.", "authors": ["Yang Li", "Siqi Ping", "Xiyu Chen", "Xiaojian Qi", "Zigan Wang", "Ye Luo", "Xiaowei Zhang"], "categories": ["cs.MA", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-11-01", "url": "https://arxiv.org/abs/2511.00628", "pdf_url": "https://arxiv.org/pdf/2511.00628v1", "arxiv_id": "2511.00628", "doi": "10.48550/arXiv.2511.00628", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3586}
{"id": "660133b96485c7bbe7bece3e7c6250d368c8453f7d1b8d81356b8f412a877aea", "sources": ["arxiv", "semantic_scholar"], "title": "FinPos: A Position-Aware Trading Agent System for Real Financial Markets", "abstract": "The exceptional potential of large language models (LLMs) in handling text information has garnered significant attention in the field of financial trading. However, most existing trading agents operate under intraday, independent unit-based trading tasks, where decisions are made as isolated directional actions, and thus lack awareness of continuous position management. Therefore, we propose a position-aware trading task designed to simulate a more realistic market. To address this task, we propose FinPos, a position-aware trading agent system designed to explicitly model and manage continuous positions. FinPos enhances position awareness through three key mechanisms: (1) professional-level interpretation of heterogeneous market information; (2) a dual-agent decision structure that separates directional reasoning from risk-aware position adjustment; and (3) multi-timescale reward signals, allowing the agent to internalize position awareness through experiential feedback rather than static instructions alone. Extensive experiments demonstrate that FinPos surpasses state-of-the-art trading agents in the position-aware trading task, which closely mirrors real market conditions. More importantly, our findings reveal that LLM-centered agent systems exhibit a vast, largely unexplored potential in long-term market decision-making.", "authors": ["Bijia Liu", "Ronghao Dang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-31", "url": "https://arxiv.org/abs/2510.27251", "pdf_url": "https://arxiv.org/pdf/2510.27251v2", "arxiv_id": "2510.27251", "doi": "10.48550/arXiv.2510.27251", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3575}
{"id": "4b2425fd121c3195fae4210974e978b17558b087668d12285a542fde53f43ecd", "sources": ["arxiv", "semantic_scholar"], "title": "Challenges in Credit Assignment for Multi-Agent Reinforcement Learning in Open Agent Systems", "abstract": "In the rapidly evolving field of multi-agent reinforcement learning (MARL), understanding the dynamics of open systems is crucial. Openness in MARL refers to the dynam-ic nature of agent populations, tasks, and agent types with-in a system. Specifically, there are three types of openness as reported in (Eck et al. 2023) [2]: agent openness, where agents can enter or leave the system at any time; task openness, where new tasks emerge, and existing ones evolve or disappear; and type openness, where the capabil-ities and behaviors of agents change over time. This report provides a conceptual and empirical review, focusing on the interplay between openness and the credit assignment problem (CAP). CAP involves determining the contribution of individual agents to the overall system performance, a task that becomes increasingly complex in open environ-ments. Traditional credit assignment (CA) methods often assume static agent populations, fixed and pre-defined tasks, and stationary types, making them inadequate for open systems. We first conduct a conceptual analysis, in-troducing new sub-categories of openness to detail how events like agent turnover or task cancellation break the assumptions of environmental stationarity and fixed team composition that underpin existing CAP methods. We then present an empirical study using representative temporal and structural algorithms in an open environment. The results demonstrate that openness directly causes credit misattribution, evidenced by unstable loss functions and significant performance degradation.", "authors": ["Alireza Saleh Abadi", "Leen-Kiat Soh"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-31", "url": "https://arxiv.org/abs/2510.27659", "pdf_url": "https://arxiv.org/pdf/2510.27659v1", "arxiv_id": "2510.27659", "doi": "10.48550/arXiv.2510.27659", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3575}
{"id": "b156cb651662019b4136bd5c94192277c7a4d3c45d0e2d61fb80a88b34a8789c", "sources": ["arxiv", "semantic_scholar"], "title": "The Geometry of Dialogue: Graphing Language Models to Reveal Synergistic Teams for Multi-Agent Collaboration", "abstract": "While a multi-agent approach based on large language models (LLMs) represents a promising strategy to surpass the capabilities of single models, its success is critically dependent on synergistic team composition. However, forming optimal teams is a significant challenge, as the inherent opacity of most models obscures the internal characteristics necessary for effective collaboration. In this paper, we propose an interaction-centric framework for automatic team composition that does not require any prior knowledge including their internal architectures, training data, or task performances. Our method constructs a \"language model graph\" that maps relationships between models from the semantic coherence of pairwise conversations, and then applies community detection to identify synergistic model clusters. Our experiments with diverse LLMs demonstrate that the proposed method discovers functionally coherent groups that reflect their latent specializations. Priming conversations with specific topics identified synergistic teams which outperform random baselines on downstream benchmarks and achieve comparable accuracy to that of manually-curated teams based on known model specializations. Our findings provide a new basis for the automated design of collaborative multi-agent LLM teams.", "authors": ["Kotaro Furuya", "Yuichi Kitagawa"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-30", "url": "https://arxiv.org/abs/2510.26352", "pdf_url": "https://arxiv.org/pdf/2510.26352v2", "arxiv_id": "2510.26352", "doi": "10.48550/arXiv.2510.26352", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3564}
{"id": "91a733601fba1d6f8d8de4e882fe57e0d84e7cea77d234cfba4d931bc772d43f", "sources": ["arxiv", "semantic_scholar"], "title": "Using Copilot Agent Mode to Automate Library Migration: A Quantitative Assessment", "abstract": "Keeping software systems up to date is essential to avoid technical debt, security vulnerabilities, and the rigidity typical of legacy systems. However, updating libraries and frameworks remains a time consuming and error-prone process. Recent advances in Large Language Models (LLMs) and agentic coding systems offer new opportunities for automating such maintenance tasks. In this paper, we evaluate the update of a well-known Python library, SQLAlchemy, across a dataset of ten client applications. For this task, we use the Github's Copilot Agent Mode, an autonomous AI systema capable of planning and executing multi-step migration workflows. To assess the effectiveness of the automated migration, we also introduce Migration Coverage, a metric that quantifies the proportion of API usage points correctly migrated. The results of our study show that the LLM agent was capable of migrating functionalities and API usages between SQLAlchemy versions (migration coverage: 100%, median), but failed to maintain the application functionality, leading to a low test-pass rate (39.75%, median).", "authors": ["Aylton Almeida", "Laerte Xavier", "Marco Tulio Valente"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-30", "url": "https://arxiv.org/abs/2510.26699", "pdf_url": "https://arxiv.org/pdf/2510.26699v3", "arxiv_id": "2510.26699", "doi": "10.1145/3786167.3788411", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2268}
{"id": "beda1c796795c89a84c55b80a4a2856713d1ef6a1652aad6ddd4265a578a6b85", "sources": ["arxiv", "semantic_scholar"], "title": "Urban-MAS: Human-Centered Urban Prediction with LLM-Based Multi-Agent System", "abstract": "Urban Artificial Intelligence (Urban AI) has advanced human-centered urban tasks such as perception prediction and human dynamics. Large Language Models (LLMs) can integrate multimodal inputs to address heterogeneous data in complex urban systems but often underperform on domain-specific tasks. Urban-MAS, an LLM-based Multi-Agent System (MAS) framework, is introduced for human-centered urban prediction under zero-shot settings. It includes three agent types: Predictive Factor Guidance Agents, which prioritize key predictive factors to guide knowledge extraction and enhance the effectiveness of compressed urban knowledge in LLMs; Reliable UrbanInfo Extraction Agents, which improve robustness by comparing multiple outputs, validating consistency, and re-extracting when conflicts occur; and Multi-UrbanInfo Inference Agents, which integrate extracted multi-source information across dimensions for prediction. Experiments on running-amount prediction and urban perception across Tokyo, Milan, and Seattle demonstrate that Urban-MAS substantially reduces errors compared to single-LLM baselines. Ablation studies indicate that Predictive Factor Guidance Agents are most critical for enhancing predictive performance, positioning Urban-MAS as a scalable paradigm for human-centered urban AI prediction. Code is available on the project website:https://github.com/THETUREHOOHA/UrbanMAS", "authors": ["Shangyu Lou"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-30", "url": "https://arxiv.org/abs/2511.00096", "pdf_url": "https://arxiv.org/pdf/2511.00096v1", "arxiv_id": "2511.00096", "doi": "10.1145/3764926.3771951", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/THETUREHOOHA/UrbanMAS", "venue": null, "quality_score": 0.4211}
{"id": "a123fa8c0045549c2da965c235c0b180cf421a6fec6e0f14b0f13e469a3afbea", "sources": ["arxiv", "semantic_scholar"], "title": "Communication and Verification in LLM Agents towards Collaboration under Information Asymmetry", "abstract": "While Large Language Model (LLM) agents are often approached from the angle of action planning/generation to accomplish a goal (e.g., given by language descriptions), their abilities to collaborate with each other to achieve a joint goal are not well explored. To address this limitation, this paper studies LLM agents in task collaboration, particularly under the condition of information asymmetry, where agents have disparities in their knowledge and skills and need to work together to complete a shared task. We extend Einstein Puzzles, a classical symbolic puzzle, to a table-top game. In this game, two LLM agents must reason, communicate, and act to satisfy spatial and relational constraints required to solve the puzzle. We apply a fine-tuning-plus-verifier framework in which LLM agents are equipped with various communication strategies and verification signals from the environment. Empirical results highlight the critical importance of aligned communication, especially when agents possess both information-seeking and -providing capabilities. Interestingly, agents without communication can still achieve high task performance; however, further analysis reveals a lack of true rule understanding and lower trust from human evaluators. Instead, by integrating an environment-based verifier, we enhance agents' ability to comprehend task rules and complete tasks, promoting both safer and more interpretable collaboration in AI systems. https://github.com/Roihn/EinsteinPuzzles", "authors": ["Run Peng", "Ziqiao Ma", "Amy Pang", "Sikai Li", "Zhang Xi-Jia", "Yingzhuo Yu", "Cristian-Paul Bara", "Joyce Chai"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-29", "url": "https://arxiv.org/abs/2510.25595", "pdf_url": "https://arxiv.org/pdf/2510.25595v1", "arxiv_id": "2510.25595", "doi": "10.48550/arXiv.2510.25595", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Roihn/EinsteinPuzzles", "venue": "arXiv.org", "quality_score": 0.549}
{"id": "cbfb8523cf0519836b01c9ff312d4fd9a4760199136a58ccc2873c7329321a80", "sources": ["arxiv", "semantic_scholar"], "title": "SMAGDi: Socratic Multi Agent Interaction Graph Distillation for Efficient High Accuracy Reasoning", "abstract": "Multi-agent systems (MAS) often achieve higher reasoning accuracy than single models, but their reliance on repeated debates across agents makes them computationally expensive. We introduce SMAGDi, a distillation framework that transfers the debate dynamics of a five-agent Llama-based MAS into a compact Socratic decomposer-solver student. SMAGDi represents debate traces as directed interaction graphs, where nodes encode intermediate reasoning steps with correctness labels and edges capture continuity and cross-agent influence. The student is trained with a composite objective combining language modeling, graph-based supervision, contrastive reasoning, and embedding alignment to preserve both fluency and structured reasoning. On StrategyQA and MMLU, SMAGDi compresses a 40B multi-agent system into a 6B student while retaining 88% of its accuracy, substantially outperforming prior distillation methods such as MAGDi, standard KD, and fine-tuned baselines. These results highlight that explicitly modeling interaction graphs and Socratic decomposition enable small models to inherit the accuracy benefits of multi-agent debate while remaining efficient enough for real-world deployment.", "authors": ["Aayush Aluru", "Myra Malik", "Samarth Patankar", "Spencer Kim", "Kevin Zhu", "Sean O'Brien", "Vasu Sharma"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-29", "url": "https://arxiv.org/abs/2511.05528", "pdf_url": "https://arxiv.org/pdf/2511.05528v1", "arxiv_id": "2511.05528", "doi": "10.48550/arXiv.2511.05528", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3552}
{"id": "09d37776277f867a1faa79bc3b1edfe28769c8e056c66b297d49f50994e242e7", "sources": ["arxiv", "semantic_scholar"], "title": "Debate2Create: Robot Co-design via Multi-Agent LLM Debate", "abstract": "We introduce Debate2Create (D2C), a multi-agent LLM framework that formulates robot co-design as structured, iterative debate grounded in physics-based evaluation. A design agent and control agent engage in a thesis-antithesis-synthesis loop, while pluralistic LLM judges provide multi-objective feedback to steer exploration. Across five MuJoCo locomotion benchmarks, D2C achieves up to $3.2\\times$ the default Ant score and $\\sim9\\times$ on Swimmer, outperforming prior LLM-based methods and black-box optimization. Iterative debate yields 18--35% gains over compute-matched zero-shot generation, and D2C-generated rewards transfer to default morphologies in 4/5 tasks. Our results demonstrate that structured multi-agent debate offers an effective alternative to hand-designed objectives for joint morphology-reward optimization.", "authors": ["Kevin Qiu", "Marek Cygan"], "categories": ["cs.RO", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-29", "url": "https://arxiv.org/abs/2510.25850", "pdf_url": "https://arxiv.org/pdf/2510.25850v2", "arxiv_id": "2510.25850", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.226}
{"id": "0d3c2597176e08353a221f7319dc8dfc352c818399062cfeb022aab9875c1a6c", "sources": ["arxiv", "semantic_scholar"], "title": "GAP: Graph-Based Agent Planning with Parallel Tool Use and Reinforcement Learning", "abstract": "Autonomous agents powered by large language models (LLMs) have shown impressive capabilities in tool manipulation for complex task-solving. However, existing paradigms such as ReAct rely on sequential reasoning and execution, failing to exploit the inherent parallelism among independent sub-tasks. This sequential bottleneck leads to inefficient tool utilization and suboptimal performance in multi-step reasoning scenarios. We introduce Graph-based Agent Planning (GAP), a novel framework that explicitly models inter-task dependencies through graph-based planning to enable adaptive parallel and serial tool execution. Our approach trains agent foundation models to decompose complex tasks into dependency-aware sub-task graphs, autonomously determining which tools can be executed in parallel and which must follow sequential dependencies. This dependency-aware orchestration achieves substantial improvements in both execution efficiency and task accuracy. To train GAP, we construct a high-quality dataset of graph-based planning traces derived from the Multi-Hop Question Answering (MHQA) benchmark. We employ a two-stage training strategy: supervised fine-tuning (SFT) on the curated dataset, followed by reinforcement learning (RL) with a correctness-based reward function on strategically sampled queries where tool-based reasoning provides maximum value. Experimental results on MHQA datasets demonstrate that GAP significantly outperforms traditional ReAct baselines, particularly on multi-step retrieval tasks, while achieving dramatic improvements in tool invocation efficiency through intelligent parallelization. The project page is available at: https://github.com/WJQ7777/Graph-Agent-Planning.", "authors": ["Jiaqi Wu", "Qinlao Zhao", "Zefeng Chen", "Kai Qin", "Yifei Zhao", "Xueqian Wang", "Yuhang Yao"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-29", "url": "https://arxiv.org/abs/2510.25320", "pdf_url": "https://arxiv.org/pdf/2510.25320v1", "arxiv_id": "2510.25320", "doi": "10.48550/arXiv.2510.25320", "citation_count": 6, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/WJQ7777/Graph-Agent-Planning", "venue": "arXiv.org", "quality_score": 0.549}
{"id": "b8998b4bedff3fc5c32eee389d8339a4179f6488099229183c8055ca034b87f2", "sources": ["arxiv", "semantic_scholar"], "title": "Affordance Representation and Recognition for Autonomous Agents", "abstract": "The autonomy of software agents is fundamentally dependent on their ability to construct an actionable internal world model from the structured data that defines their digital environment, such as the Document Object Model (DOM) of web pages and the semantic descriptions of web services. However, constructing this world model from raw structured data presents two critical challenges: the verbosity of raw HTML makes it computationally intractable for direct use by foundation models, while the static nature of hardcoded API integrations prevents agents from adapting to evolving services. This paper introduces a pattern language for world modeling from structured data, presenting two complementary architectural patterns. The DOM Transduction Pattern addresses the challenge of web page complexity by distilling} a verbose, raw DOM into a compact, task-relevant representation or world model optimized for an agent's reasoning core. Concurrently, the Hypermedia Affordances Recognition Pattern enables the agent to dynamically enrich its world model by parsing standardized semantic descriptions to discover and integrate the capabilities of unknown web services at runtime. Together, these patterns provide a robust framework for engineering agents that can efficiently construct and maintain an accurate world model, enabling scalable, adaptive, and interoperable automation across the web and its extended resources.", "authors": ["Habtom Kahsay Gidey", "Niklas Huber", "Alexander Lenz", "Alois Knoll"], "categories": ["cs.AI", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-28", "url": "https://arxiv.org/abs/2510.24459", "pdf_url": "https://arxiv.org/pdf/2510.24459v1", "arxiv_id": "2510.24459", "doi": "10.48550/arXiv.2510.24459", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3541}
{"id": "5aa4d7356eaea0e12aa8f3efef1c0b957849d024a60283c8652f5c4d071e3787", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcement Learning for Long-Horizon Multi-Turn Search Agents", "abstract": "Large Language Model (LLM) agents can leverage multiple turns and tools to solve complex tasks, with prompt-based approaches achieving strong performance. This work demonstrates that Reinforcement Learning (RL) can push capabilities significantly further by learning from experience. Through experiments on a legal document search benchmark, we show that our RL-trained 14 Billion parameter model outperforms frontier class models (85% vs 78% accuracy). In addition, we explore turn-restricted regimes, during training and at test-time, that show these agents achieve better results if allowed to operate over longer multi-turn horizons.", "authors": ["Vivek Kalyan", "Martin Andrews"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-28", "url": "https://arxiv.org/abs/2510.24126", "pdf_url": "https://arxiv.org/pdf/2510.24126v1", "arxiv_id": "2510.24126", "doi": "10.48550/arXiv.2510.24126", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3541}
{"id": "491312c512a302761ed3bb09b14a0fc7e6071827f282fc58988986c15d064097", "sources": ["arxiv", "semantic_scholar"], "title": "Retrieval- and Argumentation-Enhanced Multi-Agent LLMs for Judgmental Forecasting (Extended Version with Supplementary Material)", "abstract": "Judgmental forecasting is the task of making predictions about future events based on human judgment. This task can be seen as a form of claim verification, where the claim corresponds to a future event and the task is to assess the plausibility of that event. In this paper, we propose a novel multi-agent framework for claim verification, whereby different agents may disagree on claim veracity and bring specific evidence for and against the claims, represented as quantitative bipolar argumentation frameworks (QBAFs). We then instantiate the framework for supporting claim verification, with a variety of agents realised with Large Language Models (LLMs): (1) ArgLLM agents, an existing approach for claim verification that generates and evaluates QBAFs; (2) RbAM agents, whereby LLM-empowered Relation-based Argument Mining (RbAM) from external sources is used to generate QBAFs; (3) RAG-ArgLLM agents, extending ArgLLM agents with a form of Retrieval-Augmented Generation (RAG) of arguments from external sources. Finally, we conduct experiments with two standard judgmental forecasting datasets, with instances of our framework with two or three agents, empowered by six different base LLMs. We observe that combining evidence from agents can improve forecasting accuracy, especially in the case of three agents, while providing an explainable combination of evidence for claim verification.", "authors": ["Deniz Gorur", "Antonio Rago", "Francesca Toni"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-28", "url": "https://arxiv.org/abs/2510.24303", "pdf_url": "https://arxiv.org/pdf/2510.24303v3", "arxiv_id": "2510.24303", "doi": "10.48550/arXiv.2510.24303", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2253}
{"id": "cfdaa2d642ad9c4dee8462f2ca5567632cf009fcc661ba4972a64ca169e6ea2f", "sources": ["arxiv", "semantic_scholar"], "title": "Your LLM Agents are Temporally Blind: The Misalignment Between Tool Use Decisions and Human Time Perception", "abstract": "Large language model (LLM) agents are increasingly used to interact with and execute tasks in dynamic environments. However, a critical yet overlooked limitation of these agents is that they, by default, assume a stationary context, failing to account for the real-world time elapsed between messages. We refer to this as \"temporal blindness\". This limitation hinders decisions about when to invoke tools, leading agents to either over-rely on stale context and skip needed tool calls, or under-rely on it and redundantly repeat tool calls. To study this challenge, we constructed TicToc, a diverse dataset of multi-turn user-agent message trajectories across 76 scenarios, spanning dynamic environments with high, medium, and low time sensitivity. We collected human preferences between \"calling a tool\" and \"directly answering\" on each sample, and evaluated how well LLM tool-calling decisions align with human preferences under varying amounts of elapsed time. Our analysis reveals that existing models display poor alignment with human temporal perception, with no model achieving a normalized alignment rate better than 65% when given time stamp information. We also show that naive, prompt-based alignment techniques have limited effectiveness for most models, but specific post-training alignment can be a viable way to align multi-turn LLM tool use with human temporal perception. Our data and findings provide a first step toward understanding and mitigating temporal blindness, offering insights to foster the development of more time-aware and human-aligned agents.", "authors": ["Yize Cheng", "Arshia Soltani Moakhar", "Chenrui Fan", "Parsa Hosseini", "Kazem Faghih", "Zahra Sodagar", "Wenxiao Wang", "Soheil Feizi"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-27", "url": "https://arxiv.org/abs/2510.23853", "pdf_url": "https://arxiv.org/pdf/2510.23853v3", "arxiv_id": "2510.23853", "doi": null, "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2246}
{"id": "16536c4867f681652acc80ab2b7a3106e36e661af3e2f94f2a9419f2e66e07c7", "sources": ["arxiv", "semantic_scholar"], "title": "P1GPT: a multi-agent LLM workflow module for multi-modal financial information analysis", "abstract": "Recent advances in large language models (LLMs) have enabled multi-agent reasoning systems capable of collaborative decision-making. However, in financial analysis, most frameworks remain narrowly focused on either isolated single-agent predictors or loosely connected analyst ensembles, and they lack a coherent reasoning workflow that unifies diverse data modalities. We introduce P1GPT, a layered multi-agent LLM framework for multi-modal financial information analysis and interpretable trading decision support. Unlike prior systems that emulate trading teams through role simulation, P1GPT implements a structured reasoning pipeline that systematically fuses technical, fundamental, and news-based insights through coordinated agent communication and integration-time synthesis. Backtesting on multi-modal datasets across major U.S. equities demonstrates that P1GPT achieves superior cumulative and risk-adjusted returns, maintains low drawdowns, and provides transparent causal rationales. These findings suggest that structured reasoning workflows, rather than agent role imitation, offer a scalable path toward explainable and trustworthy financial AI systems.", "authors": ["Chen-Che Lu", "Yun-Cheng Chou", "Teng-Ruei Chen"], "categories": ["cs.CE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-27", "url": "https://arxiv.org/abs/2510.23032", "pdf_url": "https://arxiv.org/pdf/2510.23032v1", "arxiv_id": "2510.23032", "doi": "10.48550/arXiv.2510.23032", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3529}
{"id": "ffd3240afb918fa369fb038d184bfbbbe449a361be0a886f6424fa67139860a3", "sources": ["arxiv", "semantic_scholar"], "title": "On Generalization in Agentic Tool Calling: CoreThink Agentic Reasoner and MAVEN Dataset", "abstract": "Generalization across Agentic tool-calling environments remains a key unsolved challenge in developing reliable agentic reasoning systems. While large language models (LLMs) demonstrate strong performance on isolated benchmarks, their ability to transfer reasoning strategies and co-ordinate tools across diverse domains is poorly understood. In this work, we conduct a large-scale evaluation of state-of-the-art LLMs on multiple tool-calling benchmarksBFCL v3, TauBench, Tau2Bench, and AceBenchand introduce MAVEN (Math & Physics Adversarial Verification & Evaluation Network), a new out of distribution (OOD) benchmark designed to stress-test multi-step reasoning through explicit verification and adversarial task composition. Our results show that most current models achieve below 50% accuracy on MAVEN, revealing a significant generalization gap across tool-use settings. To address this, we present the CoreThink Agentic Reasoner, a framework that augments LLMs with a lightweight symbolic reasoning layer for structured decomposition and adaptive tool orchestration. Without additional training, it generalizes across all benchmarks, achieving state-of-the-art performance with 530% improvements over existing baselines at roughly one-tenth the computational cost.", "authors": ["Vishvesh Bhat", "Omkar Ghugarkar", "Julian McAuley"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-27", "url": "https://arxiv.org/abs/2510.22898", "pdf_url": "https://arxiv.org/pdf/2510.22898v1", "arxiv_id": "2510.22898", "doi": "10.48550/arXiv.2510.22898", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3529}
{"id": "3c8034daf4daa1cbb0b58e90dea086d52cdb5d217367978a1fdbd0e93cc4f7aa", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Evolve: LLM Self-Improve through Co-evolution", "abstract": "Reinforcement Learning (RL) has demonstrated significant potential in enhancing the reasoning capabilities of large language models (LLMs). However, the success of RL for LLMs heavily relies on human-curated datasets and verifiable rewards, which limit their scalability and generality. Recent Self-Play RL methods, inspired by the success of the paradigm in games and Go, aim to enhance LLM reasoning capabilities without human-annotated data. However, their methods primarily depend on a grounded environment for feedback (e.g., a Python interpreter or a game engine); extending them to general domains remains challenging. To address these challenges, we propose Multi-Agent Evolve (MAE), a framework that enables LLMs to self-evolve in solving diverse tasks, including mathematics, reasoning, and general knowledge Q&A. The core design of MAE is based on a triplet of interacting agents (Proposer, Solver, Judge) that are instantiated from a single LLM, and applies reinforcement learning to optimize their behaviors. The Proposer generates questions, the Solver attempts solutions, and the Judge evaluates both while co-evolving. Experiments on Qwen2.5-3B-Instruct demonstrate that MAE achieves an average improvement of 4.54% on multiple benchmarks. These results highlight MAE as a scalable, data-efficient method for enhancing the general reasoning abilities of LLMs with minimal reliance on human-curated supervision.", "authors": ["Yixing Chen", "Yiding Wang", "Siqi Zhu", "Haofei Yu", "Tao Feng", "Muhan Zhang", "Mostofa Patwary", "Jiaxuan You"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-27", "url": "https://arxiv.org/abs/2510.23595", "pdf_url": "https://arxiv.org/pdf/2510.23595v3", "arxiv_id": "2510.23595", "doi": "10.48550/arXiv.2510.23595", "citation_count": 37, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4515}
{"id": "a995079e380f5cc31616290b3002881177e9ecf27c0ea37a69db436787a61e06", "sources": ["arxiv", "semantic_scholar"], "title": "MAD-Fact: A Multi-Agent Debate Framework for Long-Form Factuality Evaluation in LLMs", "abstract": "The widespread adoption of Large Language Models (LLMs) raises critical concerns about the factual accuracy of their outputs, especially in high-risk domains such as biomedicine, law, and education. Existing evaluation methods for short texts often fail on long-form content due to complex reasoning chains, intertwined perspectives, and cumulative information. To address this, we propose a systematic approach integrating large-scale long-form datasets, multi-agent verification mechanisms, and weighted evaluation metrics. We construct LongHalluQA, a Chinese long-form factuality dataset; and develop MAD-Fact, a debate-based multi-agent verification system. We introduce a fact importance hierarchy to capture the varying significance of claims in long-form texts. Experiments on two benchmarks show that larger LLMs generally maintain higher factual consistency, while domestic models excel on Chinese content. Our work provides a structured framework for evaluating and enhancing factual reliability in long-form LLM outputs, guiding their safe deployment in sensitive domains.", "authors": ["Yucheng Ning", "Xixun Lin", "Fang Fang", "Yanan Cao"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-27", "url": "https://arxiv.org/abs/2510.22967", "pdf_url": "https://arxiv.org/pdf/2510.22967v2", "arxiv_id": "2510.22967", "doi": "10.48550/arXiv.2510.22967", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3529}
{"id": "7aee1e4be42765fbfccbe9d17c18fc061ca30f4e947edd73171372777d08322c", "sources": ["arxiv", "semantic_scholar"], "title": "Policies over Poses: Reinforcement Learning based Distributed Pose-Graph Optimization for Multi-Robot SLAM", "abstract": "We consider the distributed pose-graph optimization (PGO) problem, which is fundamental in accurate trajectory estimation in multi-robot simultaneous localization and mapping (SLAM). Conventional iterative approaches linearize a highly non-convex optimization objective, requiring repeated solving of normal equations, which often converge to local minima and thus produce suboptimal estimates. We propose a scalable, outlier-robust distributed planar PGO framework using Multi-Agent Reinforcement Learning (MARL). We cast distributed PGO as a partially observable Markov game defined on local pose-graphs, where each action refines a single edge's pose estimate. A graph partitioner decomposes the global pose graph, and each robot runs a recurrent edge-conditioned Graph Neural Network (GNN) encoder with adaptive edge-gating to denoise noisy edges. Robots sequentially refine poses through a hybrid policy that utilizes prior action memory and graph embeddings. After local graph correction, a consensus scheme reconciles inter-robot disagreements to produce a globally consistent estimate. Our extensive evaluations on a comprehensive suite of synthetic and real-world datasets demonstrate that our learned MARL-based actors reduce the global objective by an average of 37.5% more than the state-of-the-art distributed PGO framework, while enhancing inference efficiency by at least 6X. We also demonstrate that actor replication allows a single learned policy to scale effortlessly to substantially larger robot teams without any retraining. Code is publicly available at https://github.com/herolab-uga/policies-over-poses.", "authors": ["Sai Krishna Ghanta", "Ramviyas Parasuraman"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-26", "url": "https://arxiv.org/abs/2510.22740", "pdf_url": "https://arxiv.org/pdf/2510.22740v1", "arxiv_id": "2510.22740", "doi": "10.1109/MRS66243.2025.11357260", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/herolab-uga/policies-over-poses", "venue": "International Symposium on Multi-Robot and Multi-Agent Systems", "quality_score": 0.5436}
{"id": "571f73655a57187fbbe11ed8032e672c7cff3d9275196da44aaa52603ff95326", "sources": ["arxiv", "semantic_scholar"], "title": "Group size effects and collective misalignment in LLM multi-agent systems", "abstract": "Multi-agent systems of large language models (LLMs) are rapidly expanding across domains, introducing dynamics not captured by single-agent evaluations. Yet, existing work has mostly contrasted the behavior of a single agent with that of a collective of fixed size, leaving open a central question: how does group size shape dynamics? Here, we move beyond this dichotomy and systematically explore outcomes across the full range of group sizes. We focus on multi-agent misalignment, building on recent evidence that interacting LLMs playing a simple coordination game can generate collective biases absent in individual models. First, we show that collective bias is a deeper phenomenon than previously assessed: interaction can amplify individual biases, introduce new ones, or override model-level preferences. Second, we demonstrate that group size affects the dynamics in a non-linear way, revealing model-dependent dynamical regimes. Finally, we develop a mean-field analytical approach and show that, above a critical population size, simulations converge to deterministic predictions that expose the basins of attraction of competing equilibria. These findings establish group size as a key driver of multi-agent dynamics and highlight the need to consider population-level effects when deploying LLM-based systems at scale.", "authors": ["Ariel Flint", "Luca Maria Aiello", "Romualdo Pastor-Satorras", "Andrea Baronchelli"], "categories": ["cs.MA", "cs.AI", "cs.CY", "physics.soc-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-10-25", "url": "https://arxiv.org/abs/2510.22422", "pdf_url": "https://arxiv.org/pdf/2510.22422v1", "arxiv_id": "2510.22422", "doi": "10.48550/arXiv.2510.22422", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3506}
{"id": "c8b86b60d22610b32f6fb3bdbb6d19fde6d245b453ac301cd41a23055b6e9873", "sources": ["arxiv", "semantic_scholar"], "title": "EU-Agent-Bench: Measuring Illegal Behavior of LLM Agents Under EU Law", "abstract": "Large language models (LLMs) are increasingly deployed as agents in various contexts by providing tools at their disposal. However, LLM agents can exhibit unpredictable behaviors, including taking undesirable and/or unsafe actions. In order to measure the latent propensity of LLM agents for taking illegal actions under an EU legislative context, we introduce EU-Agent-Bench, a verifiable human-curated benchmark that evaluates an agent's alignment with EU legal norms in situations where benign user inputs could lead to unlawful actions. Our benchmark spans scenarios across several categories, including data protection, bias/discrimination, and scientific integrity, with each user request allowing for both compliant and non-compliant execution of the requested actions. Comparing the model's function calls against a rubric exhaustively supported by citations of the relevant legislature, we evaluate the legal compliance of frontier LLMs, and furthermore investigate the compliance effect of providing the relevant legislative excerpts in the agent's system prompt along with explicit instructions to comply. We release a public preview set for the research community, while holding out a private test set to prevent data contamination in evaluating upcoming models. We encourage future work extending agentic safety benchmarks to different legal jurisdictions and to multi-turn and multilingual interactions. We release our code on \\href{https://github.com/ilijalichkovski/eu-agent-bench}{this URL}.", "authors": ["Ilija Lichkovski", "Alexander Müller", "Mariam Ibrahim", "Tiwai Mhundwa"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-24", "url": "https://arxiv.org/abs/2510.21524", "pdf_url": "https://arxiv.org/pdf/2510.21524v1", "arxiv_id": "2510.21524", "doi": "10.48550/arXiv.2510.21524", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ilijalichkovski/eu-agent-bench}{this", "venue": "arXiv.org", "quality_score": 0.5401}
{"id": "42e7a5f4873b0bed032d895dd243d5038e6403cec4cdcd728a7aa0b9f0ba13c9", "sources": ["arxiv", "semantic_scholar"], "title": "When Users Are Happy but Agents Are Wrong: Multi-Dimensional Evaluation of Tool-Augmented Dialogue", "abstract": "Evaluating conversational AI systems that use external tools is challenging, as errors can arise from complex interactions among user, agent, and tools. While existing evaluation methods assess either user satisfaction or agents' tool-calling capabilities, they fail to capture critical errors in multi-turn tool-augmented dialogues-such as when agents misinterpret tool results yet appear satisfactory to users. We introduce TRACE, a benchmark of systematically synthesized tool-augmented conversations covering diverse error cases. Evaluation with state-of-the-art conversation evaluation frameworks reveals that all approaches remain far from ideal performance, demonstrating the fundamental difficulty of this benchmark.", "authors": ["Tanya Shourya", "Yingfan Wang", "Zhaoyi Joey Hou", "Shamik Roy", "Vinayshekhar Bannihatti Kumar", "Rashmi Gangadharaiah"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-22", "url": "https://arxiv.org/abs/2510.19186", "pdf_url": "https://arxiv.org/pdf/2510.19186v2", "arxiv_id": "2510.19186", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2209}
{"id": "f84d1cdee0150233521748dc5541f8966a39d49e36ad8219e0729306834c79d2", "sources": ["arxiv", "semantic_scholar"], "title": "ToolScope: Enhancing LLM Agent Tool Use through Tool Merging and Context-Aware Filtering", "abstract": "Large language model (LLM) agents rely on external tools to solve complex tasks, but real-world toolsets often contain redundant tools with overlapping names and descriptions, introducing ambiguity and reducing selection accuracy. LLMs also face strict input context limits, preventing efficient consideration of large toolsets. To address these challenges, we propose ToolScope, which includes: (1) ToolScopeMerger with Auto-Correction to automatically audit and fix tool merges, reducing redundancy, and (2) ToolScopeRetriever to rank and select only the most relevant tools for each query, compressing toolsets to fit within context limits without sacrificing accuracy. Evaluations on three state-of-the-art LLMs and three open-source tool-use benchmarks show gains of 8.38% to 38.6% in tool selection accuracy, demonstrating ToolScope's effectiveness in enhancing LLM tool use.", "authors": ["Marianne Menglin Liu", "Daniel Garcia", "Fjona Parllaku", "Vikas Upadhyay", "Syed Fahad Allam Shah", "Dan Roth"], "categories": ["cs.CL", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-22", "url": "https://arxiv.org/abs/2510.20036", "pdf_url": "https://arxiv.org/pdf/2510.20036v2", "arxiv_id": "2510.20036", "doi": "10.48550/arXiv.2510.20036", "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5366}
{"id": "37120133d01c49f01ee8d626a2ef4a26e1a0a8cf3d170b6f7206a478c7e0c8b6", "sources": ["arxiv", "semantic_scholar"], "title": "AutoMT: A Multi-Agent LLM Framework for Automated Metamorphic Testing of Autonomous Driving Systems", "abstract": "Autonomous Driving Systems (ADS) are safety-critical, where failures can be severe. While Metamorphic Testing (MT) is effective for fault detection in ADS, existing methods rely heavily on manual effort and lack automation. We present AutoMT, a multi-agent MT framework powered by Large Language Models (LLMs) that automates the extraction of Metamorphic Relations (MRs) from local traffic rules and the generation of valid follow-up test cases. AutoMT leverages LLMs to extract MRs from traffic rules in Gherkin syntax using a predefined ontology. A vision-language agent analyzes scenarios, and a search agent retrieves suitable MRs from a RAG-based database to generate follow-up cases via computer vision. Experiments show that AutoMT achieves up to 5 x higher test diversity in follow-up case generation compared to the best baseline (manual expert-defined MRs) in terms of validation rate, and detects up to 20.55% more behavioral violations. While manual MT relies on a fixed set of predefined rules, AutoMT automatically extracts diverse metamorphic relations that augment real-world datasets and help uncover corner cases often missed during in-field testing and data collection. Its modular architecture separating MR extraction, filtering, and test generation supports integration into industrial pipelines and potentially enables simulation-based testing to systematically cover underrepresented or safety-critical scenarios.", "authors": ["Linfeng Liang", "Chenkai Tan", "Yao Deng", "Yingfeng Cai", "T. Y Chen", "Xi Zheng"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-22", "url": "https://arxiv.org/abs/2510.19438", "pdf_url": "https://arxiv.org/pdf/2510.19438v1", "arxiv_id": "2510.19438", "doi": "10.48550/arXiv.2510.19438", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3472}
{"id": "b669bb7d468d6d4daa31c70997e88302fe28bd74a37338fb9b8da9b3a5f25913", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Make Friends: Coaching LLM Agents toward Emergent Social Ties", "abstract": "Can large language model (LLM) agents reproduce the complex social dynamics that characterize human online behavior -- shaped by homophily, reciprocity, and social validation -- and what memory and learning mechanisms enable such dynamics to emerge? We present a multi-agent LLM simulation framework in which agents repeatedly interact, evaluate one another, and adapt their behavior through in-context learning accelerated by a coaching signal. To model human social behavior, we design behavioral reward functions that capture core drivers of online engagement, including social interaction, information seeking, self-presentation, coordination, and emotional support. These rewards align agent objectives with empirically observed user motivations, enabling the study of how network structures and group formations emerge from individual decision-making. Our experiments show that coached LLM agents develop stable interaction patterns and form emergent social ties, yielding network structures that mirror properties of real online communities. By combining behavioral rewards with in-context adaptation, our framework establishes a principled testbed for investigating collective dynamics in LLM populations and reveals how artificial agents may approximate or diverge from human-like social behavior.", "authors": ["Philipp J. Schneider", "Lin Tian", "Marian-Andrei Rizoiu"], "categories": ["cs.AI", "cs.MA", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-22", "url": "https://arxiv.org/abs/2510.19299", "pdf_url": "https://arxiv.org/pdf/2510.19299v1", "arxiv_id": "2510.19299", "doi": "10.48550/arXiv.2510.19299", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3472}
{"id": "c013ec99b78db4fdb186e2eb25c8b4f598555c578fd41c847b2311a974129df9", "sources": ["arxiv", "semantic_scholar"], "title": "WebGraphEval: Multi-Turn Trajectory Evaluation for Web Agents using Graph Representation", "abstract": "Current evaluation of web agents largely reduces to binary success metrics or conformity to a single reference trajectory, ignoring the structural diversity present in benchmark datasets. We present WebGraphEval, a framework that abstracts trajectories from multiple agents into a unified, weighted action graph. This representation is directly compatible with benchmarks such as WebArena, leveraging leaderboard runs and newly collected trajectories without modifying environments. The framework canonically encodes actions, merges recurring behaviors, and applies structural analyses including reward propagation and success-weighted edge statistics. Evaluations across thousands of trajectories from six web agents show that the graph abstraction captures cross-model regularities, highlights redundancy and inefficiency, and identifies critical decision points overlooked by outcome-based metrics. By framing web interaction as graph-structured data, WebGraphEval establishes a general methodology for multi-path, cross-agent, and efficiency-aware evaluation of web agents.", "authors": ["Yaoyao Qian", "Yuanli Wang", "Jinda Zhang", "Yun Zong", "Meixu Chen", "Hanhan Zhou", "Jindan Huang", "Yifan Zeng", "Xinyu Hu", "Chan Hee Song", "Danqing Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-22", "url": "https://arxiv.org/abs/2510.19205", "pdf_url": "https://arxiv.org/pdf/2510.19205v1", "arxiv_id": "2510.19205", "doi": "10.48550/arXiv.2510.19205", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3472}
{"id": "09f8410f8f890a2a1f81e3a5867a42fc4c942b9762da99f3659575c959139065", "sources": ["arxiv", "semantic_scholar"], "title": "Fetch.ai: An Architecture for Modern Multi-Agent Systems", "abstract": "Recent surges in LLM-driven intelligent systems largely overlook decades of foundational multi-agent systems (MAS) research, resulting in frameworks with critical limitations such as centralization and inadequate trust and communication protocols. This paper introduces the Fetch.ai architecture, an industrial-strength platform designed to bridge this gap by facilitating the integration of classical MAS principles with modern AI capabilities. We present a novel, multi-layered solution built on a decentralized foundation of on-chain blockchain services for verifiable identity, discovery, and transactions. This is complemented by a comprehensive development framework for creating secure, interoperable agents, a cloud-based platform for deployment, and an intelligent orchestration layer where an agent-native LLM translates high-level human goals into complex, multi-agent workflows. We demonstrate the deployed nature of this system through a decentralized logistics use case where autonomous agents dynamically discover, negotiate, and transact with one another securely. Ultimately, the Fetch.ai stack provides a principled architecture for moving beyond current agent implementations towards open, collaborative, and economically sustainable multi-agent ecosystems.", "authors": ["Michael J. Wooldridge", "Attila Bagoly", "Jonathan J. Ward", "Emanuele La Malfa", "Gabriel Paludo Licks"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-21", "url": "https://arxiv.org/abs/2510.18699", "pdf_url": "https://arxiv.org/pdf/2510.18699v1", "arxiv_id": "2510.18699", "doi": "10.48550/arXiv.2510.18699", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.346}
{"id": "d39c380c75d5b392ed4e0fb9e7519ac7be4840d5b6a964e3a3e26627d7e00933", "sources": ["arxiv", "semantic_scholar"], "title": "TokenCake: A KV-Cache-centric Serving Framework for LLM-based Multi-Agent Applications", "abstract": "Large Language Models (LLMs) are increasingly deployed in complex multi-agent applications that rely on external function calls. This workload creates severe performance challenges for the KV Cache: spatial contention leads to the eviction of critical agents' caches and temporal underutilization leaves the cache of agents stalled on long-running function calls idling in GPU memory. We present TokenCake, a KV-Cache-centric serving framework that bridges this gap by co-optimizing scheduling and memory management through an agent-aware design. TokenCake's Temporal Scheduler employs an event-driven, opportunistic policy to proactively offload idle KV Caches during function calls and uses predictive uploading to hide data transfer latency. TokenCake's Spatial Scheduler uses dynamic memory partitioning, guided by a hybrid priority metric combining graph structure and runtime state, to reserve GPU memory for critical-path agents. Our evaluation on representative multi-agent benchmarks shows that TokenCake reduces end-to-end latency by over 47.06% and improves effective GPU memory utilization by up to 16.9% compared to vLLM.", "authors": ["Zhuohang Bian", "Feiyang Wu", "Zhuoran Li", "Teng Ma", "Youwei Zhuo"], "categories": ["cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-21", "url": "https://arxiv.org/abs/2510.18586", "pdf_url": "https://arxiv.org/pdf/2510.18586v3", "arxiv_id": "2510.18586", "doi": "10.48550/arXiv.2510.18586", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.346}
{"id": "f8cf0a11f88e8a27cf318ed7548d55f67077006510ec50a6e6738afea4b8fb68", "sources": ["arxiv", "semantic_scholar"], "title": "Adaptive Coopetition: Leveraging Coarse Verifier Signals for Resilient Multi-Agent LLM Reasoning", "abstract": "Inference-time computation is a critical yet challenging paradigm for enhancing the reasoning performance of large language models (LLMs). While existing strategies improve reasoning stability and consistency, they suffer from notable limitations: self-correction often reinforces the model's initial biases, and Multi-Agent Collaboration (MAC) often fails due to the lack of efficient coordination mechanisms, leading to collective errors. Although high-performing verifiers can detect reasoning errors, making them reliable requires substantial training. To address these challenges, we introduce a novel inference-time framework, Adaptive Coopetition (AdCo), in which LLM agents utilize an adaptive, UCB-based \"coopetition\" mechanism. At each round, agents leverage coarse verifier signals to determine whether to collaborate or compete, and iteratively refine their reasoning based on peer feedback. Without relying on high-performance verifiers, our adaptive strategy achieves significant performance gains on mathematical reasoning benchmarks, yielding a 20% relative improvement over baselines on the more challenging dataset. Our approach remains robust and consistent in terms of accuracy under different sample sizes and configurations. This adaptive, signal-guided \"coopetition\" framework enhances reasoning robustness by leveraging both model knowledge diversity and reasoning trace measures, while also promoting uncertainty-driven exploration, especially when participants have comparable capabilities. From this perspective, our work offers a fresh lens on inference-time computation and paves the way for more resilient multi-agent LLM systems. Our code is available at: https://github.com/AdCo-Research/adaptive-coopetition.", "authors": ["Rui Jerry Huang", "Wendy Liu", "Anastasia Miin", "Lei Ding"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-21", "url": "https://arxiv.org/abs/2510.18179", "pdf_url": "https://arxiv.org/pdf/2510.18179v2", "arxiv_id": "2510.18179", "doi": "10.48550/arXiv.2510.18179", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/AdCo-Research/adaptive-coopetition", "venue": null, "quality_score": 0.409}
{"id": "f3b55418093ac8ebc1acccc4db3fb2ac304cf9b9b3c3ca763f6b12a9d7802772", "sources": ["arxiv", "semantic_scholar"], "title": "Probabilistic Modeling of Intentions in Socially Intelligent LLM Agents", "abstract": "We present a probabilistic intent modeling framework for large language model (LLM) agents in multi-turn social dialogue. The framework maintains a belief distribution over a partner's latent intentions, initialized from contextual priors and dynamically updated through likelihood estimation after each utterance. The evolving distribution provides additional contextual grounding for the policy, enabling adaptive dialogue strategies under uncertainty. Preliminary experiments in the SOTOPIA environment show consistent improvements: the proposed framework increases the Overall score by 9.0% on SOTOPIA-All and 4.1% on SOTOPIA-Hard compared with the Qwen2.5-7B baseline, and slightly surpasses an oracle agent that directly observes partner intentions. These early results suggest that probabilistic intent modeling can contribute to the development of socially intelligent LLM agents.", "authors": ["Feifan Xia", "Yuyang Fang", "Defang Li", "Yantong Xie", "Weikang Li", "Yang Li", "Deguo Xia", "Jizhou Huang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-21", "url": "https://arxiv.org/abs/2510.18476", "pdf_url": "https://arxiv.org/pdf/2510.18476v1", "arxiv_id": "2510.18476", "doi": "10.48550/arXiv.2510.18476", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.346}
{"id": "d0bd373a7655b8d8077c6dd80fc5b82d53c5a6eeccd39dcb0ff0aca644049128", "sources": ["arxiv", "semantic_scholar"], "title": "OPTAGENT: Optimizing Multi-Agent LLM Interactions Through Verbal Reinforcement Learning for Enhanced Reasoning", "abstract": "Large Language Models (LLMs) have shown remarkable reasoning capabilities in mathematical and scientific tasks. To enhance complex reasoning, multi-agent systems have been proposed to harness the collective intelligence of LLM agents. However, existing collaboration structures are either predefined or rely on majority voting or round-table debates, which can suppress correct but less dominant agent contributions. Recent approaches model multi-agent systems as graph networks but optimize purely for agent performance, neglecting the quality of interactions. We hypothesize that effective agent communication is crucial for multi-agent reasoning and that debating quality plays a significant role. To address this, we propose $\\ours$, a multi-agent verbal reinforcement learning algorithm that dynamically constructs and refines multi-agent collaboration structures. Our method defines action spaces and a feedback mechanism that evaluates communication robustness and coherence throughout the debate. The final decision is achieved through a majority vote over all the agents. We assess $\\ours$ on various reasoning tasks, including mathematical reasoning, creative writing, scientific reasoning, and numerical sorting. Results demonstrate that our approach significantly outperforms single-agent prompting methods and state-of-the-art multi-agent frameworks on diverse tasks.", "authors": ["Zhenyu Bi", "Meng Lu", "Yang Li", "Swastik Roy", "Weijie Guan", "Morteza Ziyadi", "Xuan Wang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-20", "url": "https://arxiv.org/abs/2510.18032", "pdf_url": "https://arxiv.org/pdf/2510.18032v1", "arxiv_id": "2510.18032", "doi": "10.48550/arXiv.2510.18032", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2195}
{"id": "a7f8fc83305a87a6768cd07ce961923765cc337c0de6f094a3848679c3103fd3", "sources": ["arxiv", "semantic_scholar"], "title": "R2BC: Multi-Agent Imitation Learning from Single-Agent Demonstrations", "abstract": "Imitation Learning (IL) is a natural way for humans to teach robots, particularly when high-quality demonstrations are easy to obtain. While IL has been widely applied to single-robot settings, relatively few studies have addressed the extension of these methods to multi-agent systems, especially in settings where a single human must provide demonstrations to a team of collaborating robots. In this paper, we introduce and study Round-Robin Behavior Cloning (R2BC), a method that enables a single human operator to effectively train multi-robot systems through sequential, single-agent demonstrations. Our approach allows the human to teleoperate one agent at a time and incrementally teach multi-agent behavior to the entire system, without requiring demonstrations in the joint multi-agent action space. We show that R2BC methods match, and in some cases surpass, the performance of an oracle behavior cloning approach trained on privileged synchronized demonstrations across four multi-agent simulated tasks. Finally, we deploy R2BC on two physical robot tasks trained using real human demonstrations.", "authors": ["Connor Mattson", "Varun Raveendra", "Ellen Novoseller", "Nicholas Waytowich", "Vernon J. Lawhern", "Daniel S. Brown"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-20", "url": "https://arxiv.org/abs/2510.18085", "pdf_url": "https://arxiv.org/pdf/2510.18085v1", "arxiv_id": "2510.18085", "doi": "10.48550/arXiv.2510.18085", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3449}
{"id": "154b232200b138d61ea0a8d3da9edccad983eeba68a7c51238b32acebeca363d", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Based Multi-Agent System for Simulating and Analyzing Marketing and Consumer Behavior", "abstract": "Simulating consumer decision-making is vital for designing and evaluating marketing strategies before costly real-world deployment. However, post-event analyses and rule-based agent-based models (ABMs) struggle to capture the complexity of human behavior and social interaction. We introduce an LLM-powered multi-agent simulation framework that models consumer decisions and social dynamics. Building on recent advances in large language model simulation in a sandbox environment, our framework enables generative agents to interact, express internal reasoning, form habits, and make purchasing decisions without predefined rules. In a price-discount marketing scenario, the system delivers actionable strategy-testing outcomes and reveals emergent social patterns beyond the reach of conventional methods. This approach offers marketers a scalable, low-risk tool for pre-implementation testing, reducing reliance on time-intensive post-event evaluations and lowering the risk of underperforming campaigns.", "authors": ["Man-Lin Chu", "Lucian Terhorst", "Kadin Reed", "Tom Ni", "Weiwei Chen", "Rongyu Lin"], "categories": ["cs.AI", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-20", "url": "https://arxiv.org/abs/2510.18155", "pdf_url": "https://arxiv.org/pdf/2510.18155v1", "arxiv_id": "2510.18155", "doi": "10.1109/ICEBE68123.2025.00018", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on e-Business Engineering", "quality_score": 0.3449}
{"id": "0bb4862a23520c09f2a581550eb61b7cefc43f7fc041aaed76dc213d04e7a3ef", "sources": ["arxiv", "semantic_scholar"], "title": "Verification-Aware Planning for Multi-Agent Systems", "abstract": "Large language model (LLM) agents are increasingly deployed to tackle complex tasks, often necessitating collaboration among multiple specialized agents. However, multi-agent collaboration introduces new challenges in planning, coordination, and verification. Execution failures frequently arise not from flawed reasoning alone, but from subtle misalignments in task interpretation, output format, or inter-agent handoffs. To address these challenges, we present VeriMAP, a framework for multi-agent collaboration with verification-aware planning. The VeriMAP planner decomposes tasks, models subtask dependencies, and encodes planner-defined passing criteria as subtask verification functions (VFs) in Python and natural language. We evaluate VeriMAP on diverse datasets, demonstrating that it outperforms both single- and multi-agent baselines while enhancing system robustness and interpretability. Our analysis highlights how verification-aware planning enables reliable coordination and iterative refinement in multi-agent systems, without relying on external labels or annotations.", "authors": ["Tianyang Xu", "Dan Zhang", "Kushan Mitra", "Estevam Hruschka"], "categories": ["cs.CL", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-20", "url": "https://arxiv.org/abs/2510.17109", "pdf_url": "https://arxiv.org/pdf/2510.17109v1", "arxiv_id": "2510.17109", "doi": "10.48550/arXiv.2510.17109", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.3449}
{"id": "5e209cd1c6334c4e0cbf4f404359c6ade1000a9c8f6af8728f85cd59983dbe15", "sources": ["arxiv", "semantic_scholar"], "title": "TACLA: An LLM-Based Multi-Agent Tool for Transactional Analysis Training in Education", "abstract": "Simulating nuanced human social dynamics with Large Language Models (LLMs) remains a significant challenge, particularly in achieving psychological depth and consistent persona behavior crucial for high-fidelity training tools. This paper introduces TACLA (Transactional Analysis Contextual LLM-based Agents), a novel Multi-Agent architecture designed to overcome these limitations. TACLA integrates core principles of Transactional Analysis (TA) by modeling agents as an orchestrated system of distinct Parent, Adult, and Child ego states, each with its own pattern memory. An Orchestrator Agent prioritizes ego state activation based on contextual triggers and an agent's life script, ensuring psychologically authentic responses. Validated in an educational scenario, TACLA demonstrates realistic ego state shifts in Student Agents, effectively modeling conflict de-escalation and escalation based on different teacher intervention strategies. Evaluation shows high conversational credibility and confirms TACLA's capacity to create dynamic, psychologically-grounded social simulations, advancing the development of effective AI tools for education and beyond.", "authors": ["Monika Zamojska", "Jarosław A. Chudziak"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-19", "url": "https://arxiv.org/abs/2510.17913", "pdf_url": "https://arxiv.org/pdf/2510.17913v1", "arxiv_id": "2510.17913", "doi": "10.1109/ICTAI66417.2025.00048", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Tools with Artificial Intelligence", "quality_score": 0.3438}
{"id": "6714679a48385e78249b0e2b5d322e52131cd0e0b99d70751595644f12b82d02", "sources": ["arxiv", "semantic_scholar"], "title": "Unleashing Diverse Thinking Modes in LLMs through Multi-Agent Collaboration", "abstract": "Large Language Models (LLMs) demonstrate strong performance but often lack interpretable reasoning. This paper introduces the Multi-Agent Collaboration Framework for Diverse Thinking Modes (DiMo), which enhances both performance and interpretability by simulating a structured debate among four specialized LLM agents. Each agent embodies a distinct reasoning paradigm, allowing the framework to collaboratively explore diverse cognitive approaches. Through iterative debate, agents challenge and refine initial responses, yielding more robust conclusions and an explicit, auditable reasoning chain. Across six benchmarks and under a unified open-source setup, DiMo improves accuracy over widely used single-model and debate baselines, with the largest gains on math. We position DiMo as a semantics-aware, Web-native multi-agent framework: it models human-machine intelligence with LLM agents that produce semantically typed, URL-annotated evidence chains for explanations and user-friendly interactions. Although our experiments use standard reasoning benchmarks, the framework is designed to be instantiated over Web corpora and knowledge graphs, combining retrieval-augmented reasoning with structured justifications that downstream systems can inspect and reuse.", "authors": ["Zhixuan He", "Yue Feng"], "categories": ["cs.CL", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-18", "url": "https://arxiv.org/abs/2510.16645", "pdf_url": "https://arxiv.org/pdf/2510.16645v1", "arxiv_id": "2510.16645", "doi": "10.48550/arXiv.2510.16645", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5295}
{"id": "dab985d41866f0179e83907604f87711afab5b0b25caec510fc4fb00710eedd3", "sources": ["arxiv", "semantic_scholar"], "title": "CodeCRDT: Observation-Driven Coordination for Multi-Agent LLM Code Generation", "abstract": "Multi-agent LLM systems fail to realize parallel speedups due to costly coordination. We present CodeCRDT, an observation-driven coordination pattern where agents coordinate by monitoring a shared state with observable updates and deterministic convergence, rather than explicit message passing. Using Conflict-Free Replicated Data Types (CRDTs), CodeCRDT enables lock-free, conflict-free concurrent code generation with strong eventual consistency. Evaluation across 600 trials (6 tasks, 50 runs per mode) shows both benefits and trade-offs: up to 21.1% speedup on some tasks, up to 39.4% slowdown on others, and 100% convergence with zero merge failures. The study formalizes observation-driven coordination for stochastic LLM agents, revealing semantic conflict rates (5-10%) and quality-performance tradeoffs, and provides empirical characterization of when parallel coordination succeeds versus fails based on task structure.", "authors": ["Sergey Pugachev"], "categories": ["cs.DC", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-18", "url": "https://arxiv.org/abs/2510.18893", "pdf_url": "https://arxiv.org/pdf/2510.18893v1", "arxiv_id": "2510.18893", "doi": "10.48550/arXiv.2510.18893", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3426}
{"id": "47518a84b2f59bf1f3867af1881d0708ef7bd1ef4e84f08497da2fbf7564d5a0", "sources": ["arxiv", "semantic_scholar"], "title": "Check Yourself Before You Wreck Yourself: Selectively Quitting Improves LLM Agent Safety", "abstract": "As Large Language Model (LLM) agents increasingly operate in complex environments with real-world consequences, their safety becomes critical. While uncertainty quantification is well-studied for single-turn tasks, multi-turn agentic scenarios with real-world tool access present unique challenges where uncertainties and ambiguities compound, leading to severe or catastrophic risks beyond traditional text generation failures. We propose using \"quitting\" as a simple yet effective behavioral mechanism for LLM agents to recognize and withdraw from situations where they lack confidence. Leveraging the ToolEmu framework, we conduct a systematic evaluation of quitting behavior across 12 state-of-the-art LLMs. Our results demonstrate a highly favorable safety-helpfulness trade-off: agents prompted to quit with explicit instructions improve safety by an average of +0.39 on a 0-3 scale across all models (+0.64 for proprietary models), while maintaining a negligible average decrease of -0.03 in helpfulness. Our analysis demonstrates that simply adding explicit quit instructions proves to be a highly effective safety mechanism that can immediately be deployed in existing agent systems, and establishes quitting as an effective first-line defense mechanism for autonomous agents in high-stakes applications.", "authors": ["Vamshi Krishna Bonagiri", "Ponnurangam Kumaragurum", "Khanh Nguyen", "Benjamin Plaut"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-18", "url": "https://arxiv.org/abs/2510.16492", "pdf_url": "https://arxiv.org/pdf/2510.16492v3", "arxiv_id": "2510.16492", "doi": "10.48550/arXiv.2510.16492", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3426}
{"id": "19ac28bdf5c2ba668c1844bd9a57805c97057df2292c56b6e8a949c74c1b1811", "sources": ["arxiv", "semantic_scholar"], "title": "MARSHAL: Incentivizing Multi-Agent Reasoning via Self-Play with Strategic LLMs", "abstract": "Developing Large Language Models (LLMs) to cooperate and compete effectively within multi-agent systems (MASs) is a critical step towards more advanced intelligence. While reinforcement learning (RL) has proven effective for enhancing reasoning in single-agent tasks, its extension to multi-turn, multi-agent scenarios remains underexplored due to the challenges of long-horizon credit assignment and agent-specific advantage estimation. To address these challenges, we introduce MARSHAL, an end-to-end RL framework that incentivizes Multi-Agent Reasoning through Self-play witH strAtegic LLMs in both cooperative and competitive games. MARSHAL features a turn-level advantage estimator that aligns learning signals with each interaction for credit assignment, and an agent-specific advantage normalization to stabilize multi-agent training. By learning with self-play across cooperative and competitive games, MARSHAL agents trained from Qwen3-4B develop strong strategic abilities, with up to 28.7% performance improvements in held-out games. More importantly, the capability acquired through self-play generalizes beyond games, yielding consistent performance gains of MASs in reasoning benchmarks. When integrated into leading MASs, our MARSHAL agent achieves significant zero-shot performance gains of up to 10.0% on AIME, 7.6% on GPQA-Diamond, and 3.5% on average across all benchmarks. These results establish self-play in strategic games as a powerful approach for developing generalizable multi-agent reasoning capabilities in LLMs.", "authors": ["Huining Yuan", "Zelai Xu", "Zheyue Tan", "Xiangmin Yi", "Mo Guang", "Kaiwen Long", "Haojia Hui", "Boxun Li", "Xinlei Chen", "Bo Zhao", "Xiao-Ping Zhang", "Chao Yu", "Yu Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-17", "url": "https://arxiv.org/abs/2510.15414", "pdf_url": "https://arxiv.org/pdf/2510.15414v3", "arxiv_id": "2510.15414", "doi": null, "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2785}
{"id": "bee5c9f1ae55fb069727625261561b4efd6143b800bf5dcd7c9386d52c72ff92", "sources": ["arxiv", "semantic_scholar"], "title": "Repairing Tool Calls Using Post-tool Execution Reflection and RAG", "abstract": "Agentic systems interact with external systems by calling tools such as Python functions, REST API endpoints, or command line tools such as kubectl in Kubernetes. These tool calls often fail for various syntactic and semantic reasons. Some less obvious semantic errors can only be identified and resolved after analyzing the tool's response. To repair these errors, we develop a post-tool execution reflection component that combines large language model (LLM)-based reflection with domain-specific retrieval-augmented generation (RAG) using documents describing both the specific tool being called and troubleshooting documents related to the tool. For this paper, we focus on the use case of the kubectl command line tool to manage Kubernetes, a platform for orchestrating cluster applications. Through a larger empirical study and a smaller manual evaluation, we find that our RAG-based reflection will repair kubectl commands such that they are both more likely to successfully execute (pass rate) for 55% of our models evaluated and 36% more likely to correctly answer the user query on average. We find that troubleshooting documents improve pass rate compared to official documentation by an average of 10%.", "authors": ["Jason Tsay", "Zidane Wright", "Gaodan Fang", "Kiran Kate", "Saurabh Jha", "Yara Rizk"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-17", "url": "https://arxiv.org/abs/2510.17874", "pdf_url": "https://arxiv.org/pdf/2510.17874v1", "arxiv_id": "2510.17874", "doi": "10.48550/arXiv.2510.17874", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3415}
{"id": "5a848c407dd53e342dd51a26fc3c114c49b8f998a980dddf7355ac641ee30857", "sources": ["arxiv", "semantic_scholar"], "title": "TriAgent: Automated Biomarker Discovery with Deep Research Grounding for Triage in Acute Care by LLM-Based Multi-Agent Collaboration", "abstract": "Emergency departments worldwide face rising patient volumes, workforce shortages, and variability in triage decisions that threaten the delivery of timely and accurate care. Current triage methods rely primarily on vital signs, routine laboratory values, and clinicians' judgment, which, while effective, often miss emerging biological signals that could improve risk prediction for infection typing or antibiotic administration in acute conditions. To address this challenge, we introduce TriAgent, a large language model (LLM)-based multi-agent framework that couples automated biomarker discovery with deep research for literature-grounded validation and novelty assessment. TriAgent employs a supervisor research agent to generate research topics and delegate targeted queries to specialized sub-agents for evidence retrieval from various data sources. Findings are synthesized to classify biomarkers as either grounded in existing knowledge or flagged as novel candidates, offering transparent justification and highlighting unexplored pathways in acute care risk stratification. Unlike prior frameworks limited to existing routine clinical biomarkers, TriAgent aims to deliver an end-to-end framework from data analysis to literature grounding to improve transparency, explainability and expand the frontier of potentially actionable clinical biomarkers. Given a user's clinical query and quantitative triage data, TriAgent achieved a topic adherence F1 score of 55.7 +/- 5.0%, surpassing the CoT-ReAct agent by over 10%, and a faithfulness score of 0.42 +/- 0.39, exceeding all baselines by more than 50%. Across experiments, TriAgent consistently outperformed state-of-the-art LLM-based agentic frameworks in biomarker justification and literature-grounded novelty assessment. We share our repo: https://github.com/CellFace/TriAgent.", "authors": ["Kerem Delikoyun", "Qianyu Chen", "Win Sen Kuan", "John Tshon Yit Soong", "Matthew Edward Cove", "Oliver Hayden"], "categories": ["q-bio.QM", "cs.AI"], "fields_of_study": ["Computer Science", "Biology"], "published_date": "2025-10-17", "url": "https://arxiv.org/abs/2510.16080", "pdf_url": "https://arxiv.org/pdf/2510.16080v1", "arxiv_id": "2510.16080", "doi": "10.48550/arXiv.2510.16080", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/CellFace/TriAgent", "venue": "arXiv.org", "quality_score": 0.5277}
{"id": "1d2cf5ef6ee1ac927bf5d6f36fcaadc3246759cf45b04cd40751a60f69448883", "sources": ["arxiv", "semantic_scholar"], "title": "The Spark Effect: On Engineering Creative Diversity in Multi-Agent AI Systems", "abstract": "Creative services teams increasingly rely on large language models (LLMs) to accelerate ideation, yet production systems often converge on homogeneous outputs that fail to meet brand or artistic expectations. Art of X developed persona-conditioned LLM agents -- internally branded as \"Sparks\" and instantiated through a library of role-inspired system prompts -- to intentionally diversify agent behaviour within a multi-agent workflow. This white paper documents the problem framing, experimental design, and quantitative evidence behind the Spark agent programme. Using an LLM-as-a-judge protocol calibrated against human gold standards, we observe a mean diversity gain of +4.1 points (on a 1-10 scale) when persona-conditioned Spark agents replace a uniform system prompt, narrowing the gap to human experts to 1.0 point. We also surface evaluator bias and procedural considerations for future deployments.", "authors": ["Alexander Doudkin", "Anton Voelker", "Friedrich von Borries"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-17", "url": "https://arxiv.org/abs/2510.15568", "pdf_url": "https://arxiv.org/pdf/2510.15568v1", "arxiv_id": "2510.15568", "doi": "10.48550/arXiv.2510.15568", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3415}
{"id": "46528d126ced67d4667fa412fdd86d3433911259db8f397a28fdaed4ccc1acb4", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-dimensional Data Analysis and Applications Basing on LLM Agents and Knowledge Graph Interactions", "abstract": "In the current era of big data, extracting deep insights from massive, heterogeneous, and complexly associated multi-dimensional data has become a significant challenge. Large Language Models (LLMs) perform well in natural language understanding and generation, but still suffer from \"hallucination\" issues when processing structured knowledge and are difficult to update in real-time. Although Knowledge Graphs (KGs) can explicitly store structured knowledge, their static nature limits dynamic interaction and analytical capabilities. Therefore, this paper proposes a multi-dimensional data analysis method based on the interactions between LLM agents and KGs, constructing a dynamic, collaborative analytical ecosystem. This method utilizes LLM agents to automatically extract product data from unstructured data, constructs and visualizes the KG in real-time, and supports users in deep exploration and analysis of graph nodes through an interactive platform. Experimental results show that this method has significant advantages in product ecosystem analysis, relationship mining, and user-driven exploratory analysis, providing new ideas and tools for multi-dimensional data analysis.", "authors": ["Xi Wang", "Xianyao Ling", "Kun Li", "Gang Yin", "Liang Zhang", "Jiang Wu", "Jun Xu", "Fu Zhang", "Wenbo Lei", "Annie Wang", "Peng Gong"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-17", "url": "https://arxiv.org/abs/2510.15258", "pdf_url": "https://arxiv.org/pdf/2510.15258v2", "arxiv_id": "2510.15258", "doi": "10.48550/arXiv.2510.15258", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3415}
{"id": "2f7b3ff07ea1fb0d347c65fbd0ed81ee501421baf55a9d8ed2b923df1e18779d", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agents Beyond Utility: An Open-Ended Perspective", "abstract": "Recent LLM agents have made great use of chain of thought reasoning and function calling. As their capabilities grow, an important question arises: can this software represent not only a smart problem-solving tool, but an entity in its own right, that can plan, design immediate tasks, and reason toward broader, more ambiguous goals? To study this question, we adopt an open-ended experimental setting where we augment a pretrained LLM agent with the ability to generate its own tasks, accumulate knowledge, and interact extensively with its environment. We study the resulting open-ended agent qualitatively. It can reliably follow complex multi-step instructions, store and reuse information across runs, and propose and solve its own tasks, though it remains sensitive to prompt design, prone to repetitive task generation, and unable to form self-representations. These findings illustrate both the promise and current limits of adapting pretrained LLMs toward open-endedness, and point to future directions for training agents to manage memory, explore productively, and pursue abstract long-term goals.", "authors": ["Asen Nachkov", "Xi Wang", "Luc Van Gool"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-16", "url": "https://arxiv.org/abs/2510.14548", "pdf_url": "https://arxiv.org/pdf/2510.14548v1", "arxiv_id": "2510.14548", "doi": "10.48550/arXiv.2510.14548", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3403}
{"id": "bd611174a0e7c8ce25571f6fe2a76a4c92ad21a1048865ec8ee1bf60e885bb35", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Agentic Self-Learning LLMs in Search Environment", "abstract": "We study whether self-learning can scale LLM-based agents without relying on human-curated datasets or predefined rule-based rewards. Through controlled experiments in a search-agent setting, we identify two key determinants of scalable agent training: the source of reward signals and the scale of agent task data. We find that rewards from a Generative Reward Model (GRM) outperform rigid rule-based signals for open-domain learning, and that co-evolving the GRM with the policy further boosts performance. Increasing the volume of agent task data-even when synthetically generated-substantially enhances agentic capabilities. Building on these insights, we propose \\textbf{Agentic Self-Learning} (ASL), a fully closed-loop, multi-role reinforcement learning framework that unifies task generation, policy execution, and evaluation within a shared tool environment and LLM backbone. ASL coordinates a Prompt Generator, a Policy Model, and a Generative Reward Model to form a virtuous cycle of harder task setting, sharper verification, and stronger solving. Empirically, ASL delivers steady, round-over-round gains, surpasses strong RLVR baselines (e.g., Search-R1) that plateau or degrade, and continues improving under zero-labeled-data conditions, indicating superior sample efficiency and robustness. We further show that GRM verification capacity is the main bottleneck: if frozen, it induces reward hacking and stalls progress; continual GRM training on the evolving data distribution mitigates this, and a small late-stage injection of real verification data raises the performance ceiling. This work establishes reward source and data scale as critical levers for open-domain agent learning and demonstrates the efficacy of multi-role co-evolution for scalable, self-improving agents. The data and code of this paper are released at https://github.com/forangel2014/Towards-Agentic-Self-Learning", "authors": ["Wangtao Sun", "Xiang Cheng", "Jialin Fan", "Yao Xu", "Xing Yu", "Shizhu He", "Jun Zhao", "Kang Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-16", "url": "https://arxiv.org/abs/2510.14253", "pdf_url": "https://arxiv.org/pdf/2510.14253v2", "arxiv_id": "2510.14253", "doi": "10.48550/arXiv.2510.14253", "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/forangel2014/Towards-Agentic-Self-Learning", "venue": "arXiv.org", "quality_score": 0.5259}
{"id": "95b185c89522cb8052edff32d4863e80cb3e7b54bd69a94c2b661d6d8bcba337", "sources": ["arxiv", "semantic_scholar"], "title": "The Role of Social Learning and Collective Norm Formation in Fostering Cooperation in LLM Multi-Agent Systems", "abstract": "A growing body of multi-agent studies with LLMs explores how norms and cooperation emerge in mixed-motive scenarios, where pursuing individual gain can undermine the collective good. While prior work has explored these dynamics in both richly contextualized simulations and simplified game-theoretic environments, most LLM systems featuring common-pool resource (CPR) games provide agents with explicit reward functions directly tied to their actions. In contrast, human cooperation often emerges without explicit knowledge of the payoff structure or how individual actions translate into long-run outcomes, relying instead on heuristics, communication, and enforcement. We introduce a CPR simulation framework that removes explicit reward signals and embeds cultural-evolutionary mechanisms: social learning (adopting strategies and beliefs from successful peers) and norm-based punishment, grounded in Ostrom's principles of resource governance. Agents also individually learn from the consequences of harvesting, monitoring, and punishing via environmental feedback, enabling norms to emerge endogenously. We establish the validity of our simulation by reproducing key findings from existing studies on human behavior. Building on this, we examine norm evolution across a $2\\times2$ grid of environmental and social initialisations (resource-rich vs. resource-scarce; altruistic vs. selfish) and benchmark how agentic societies comprised of different LLMs perform under these conditions. Our results reveal systematic model differences in sustaining cooperation and norm formation, positioning the framework as a rigorous testbed for studying emergent norms in mixed-motive LLM societies. Such analysis can inform the design of AI systems deployed in social and organizational contexts, where alignment with cooperative norms is critical for stability, fairness, and effective governance of AI-mediated environments.", "authors": ["Prateek Gupta", "Qiankun Zhong", "Hiromu Yakura", "Thomas Eisenmann", "Iyad Rahwan"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-16", "url": "https://arxiv.org/abs/2510.14401", "pdf_url": "https://arxiv.org/pdf/2510.14401v2", "arxiv_id": "2510.14401", "doi": "10.65109/CZDC3237", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2166}
{"id": "9874931efb8e10e03c2bce8af19bba8f1c6d7d2caa27dbba7a5f56066bf9bbb4", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agents for Automated Web Vulnerability Reproduction: Are We There Yet?", "abstract": "Large language model (LLM) agents have demonstrated remarkable capabilities in software engineering and cybersecurity tasks, including code generation, vulnerability discovery, and automated testing. One critical but underexplored application is automated web vulnerability reproduction, which transforms vulnerability reports into working exploits. Although recent advances suggest promising potential, challenges remain in applying LLM agents to real-world web vulnerability reproduction scenarios. In this paper, we present the first comprehensive evaluation of state-of-the-art LLM agents for automated web vulnerability reproduction. We systematically assess 20 agents from software engineering, cybersecurity, and general domains across 16 dimensions, including technical capabilities, environment adaptability, and user experience factors, on 3 representative web vulnerabilities. Based on the results, we select three top-performing agents (OpenHands, SWE-agent, and CAI) for in-depth evaluation on our benchmark dataset of 80 real-world CVEs spanning 7 vulnerability types and 6 web technologies. Our results reveal that while LLM agents achieve reasonable success on simple library-based vulnerabilities, they consistently fail on complex service-based vulnerabilities requiring multi-component environments. Complex environment configurations and authentication barriers create a gap where agents can execute exploit code but fail to trigger actual vulnerabilities. We observe high sensitivity to input guidance, with performance degrading by over 33% under incomplete authentication information. Our findings highlight the significant gap between current LLM agent capabilities and the demands of reliable automated vulnerability reproduction, emphasizing the need for advances in environmental adaptation and autonomous problem-solving capabilities.", "authors": ["Bin Liu", "Yanjie Zhao", "Guoai Xu", "Haoyu Wang"], "categories": ["cs.SE", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-16", "url": "https://arxiv.org/abs/2510.14700", "pdf_url": "https://arxiv.org/pdf/2510.14700v1", "arxiv_id": "2510.14700", "doi": "10.48550/arXiv.2510.14700", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3403}
{"id": "d30a5c165eca4bcbe09531f2ccea7912004c0582a37f607ecb1432cf32cc1078", "sources": ["arxiv", "semantic_scholar"], "title": "AOAD-MAT: Transformer-based multi-agent deep reinforcement learning model considering agents' order of action decisions", "abstract": "Multi-agent reinforcement learning focuses on training the behaviors of multiple learning agents that coexist in a shared environment. Recently, MARL models, such as the Multi-Agent Transformer (MAT) and ACtion dEpendent deep Q-learning (ACE), have significantly improved performance by leveraging sequential decision-making processes. Although these models can enhance performance, they do not explicitly consider the importance of the order in which agents make decisions. In this paper, we propose an Agent Order of Action Decisions-MAT (AOAD-MAT), a novel MAT model that considers the order in which agents make decisions. The proposed model explicitly incorporates the sequence of action decisions into the learning process, allowing the model to learn and predict the optimal order of agent actions. The AOAD-MAT model leverages a Transformer-based actor-critic architecture that dynamically adjusts the sequence of agent actions. To achieve this, we introduce a novel MARL architecture that cooperates with a subtask focused on predicting the next agent to act, integrated into a Proximal Policy Optimization based loss function to synergistically maximize the advantage of the sequential decision-making. The proposed method was validated through extensive experiments on the StarCraft Multi-Agent Challenge and Multi-Agent MuJoCo benchmarks. The experimental results show that the proposed AOAD-MAT model outperforms existing MAT and other baseline models, demonstrating the effectiveness of adjusting the AOAD order in MARL.", "authors": ["Shota Takayama", "Katsuhide Fujita"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-15", "url": "https://arxiv.org/abs/2510.13343", "pdf_url": "https://arxiv.org/pdf/2510.13343v1", "arxiv_id": "2510.13343", "doi": "10.1007/978-3-032-13562-9_23", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Prima", "quality_score": 0.3392}
{"id": "86ea6ae84ba08003ce6fbcc821388de30456de76e18dc62658e3f111c1615810", "sources": ["arxiv", "semantic_scholar"], "title": "MADREC: A Multi-Aspect Driven LLM Agent for Explainable and Adaptive Recommendation", "abstract": "Recent attempts to integrate large language models (LLMs) into recommender systems have gained momentum, but most remain limited to simple text generation or static prompt-based inference, failing to capture the complexity of user preferences and real-world interactions. This study proposes the Multi-Aspect Driven LLM Agent MADRec, an autonomous LLM-based recommender that constructs user and item profiles by unsupervised extraction of multi-aspect information from reviews and performs direct recommendation, sequential recommendation, and explanation generation. MADRec generates structured profiles via aspect-category-based summarization and applies Re-Ranking to construct high-density inputs. When the ground-truth item is missing from the output, the Self-Feedback mechanism dynamically adjusts the inference criteria. Experiments across multiple domains show that MADRec outperforms traditional and LLM-based baselines in both precision and explainability, with human evaluation further confirming the persuasiveness of the generated explanations.", "authors": ["Jiin Park", "Misuk Kim"], "categories": ["cs.IR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-15", "url": "https://arxiv.org/abs/2510.13371", "pdf_url": "https://arxiv.org/pdf/2510.13371v1", "arxiv_id": "2510.13371", "doi": "10.48550/arXiv.2510.13371", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3392}
{"id": "fa01984f3401ed7e969fd6466b46652b0b39b4de0799f32afa8332d525966069", "sources": ["arxiv", "semantic_scholar"], "title": "Static Sandboxes Are Inadequate: Modeling Societal Complexity Requires Open-Ended Co-Evolution in LLM-Based Multi-Agent Simulations", "abstract": "What if artificial agents could not just communicate, but also evolve, adapt, and reshape their worlds in ways we cannot fully predict? With llm now powering multi-agent systems and social simulations, we are witnessing new possibilities for modeling open-ended, ever-changing environments. Yet, most current simulations remain constrained within static sandboxes, characterized by predefined tasks, limited dynamics, and rigid evaluation criteria. These limitations prevent them from capturing the complexity of real-world societies. In this paper, we argue that static, task-specific benchmarks are fundamentally inadequate and must be rethought. We critically review emerging architectures that blend llm with multi-agent dynamics, highlight key hurdles such as balancing stability and diversity, evaluating unexpected behaviors, and scaling to greater complexity, and introduce a fresh taxonomy for this rapidly evolving field. Finally, we present a research roadmap centered on open-endedness, continuous co-evolution, and the development of resilient, socially aligned AI ecosystems. We call on the community to move beyond static paradigms and help shape the next generation of adaptive, socially-aware multi-agent simulations.", "authors": ["Jinkun Chen", "Sher Badshah", "Xuemin Yu", "Sijia Han"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-15", "url": "https://arxiv.org/abs/2510.13982", "pdf_url": "https://arxiv.org/pdf/2510.13982v3", "arxiv_id": "2510.13982", "doi": "10.48550/arXiv.2510.13982", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3392}
{"id": "31942d6f9ae58ad03c510663b90865f5e6a7c555afcb82aeefe1f70680fb391f", "sources": ["arxiv", "semantic_scholar"], "title": "Stop Reducing Responsibility in LLM-Powered Multi-Agent Systems to Local Alignment", "abstract": "LLM-powered Multi-Agent Systems (LLM-MAS) unlock new potentials in distributed reasoning, collaboration, and task generalization but also introduce additional risks due to unguaranteed agreement, cascading uncertainty, and adversarial vulnerabilities. We argue that ensuring responsible behavior in such systems requires a paradigm shift: from local, superficial agent-level alignment to global, systemic agreement. We conceptualize responsibility not as a static constraint but as a lifecycle-wide property encompassing agreement, uncertainty, and security, each requiring the complementary integration of subjective human-centered values and objective verifiability. Furthermore, a dual-perspective governance framework that combines interdisciplinary design with human-AI collaborative oversight is essential for tracing and ensuring responsibility throughout the lifecycle of LLM-MAS. Our position views LLM-MAS not as loose collections of agents, but as unified, dynamic socio-technical systems that demand principled mechanisms to support each dimension of responsibility and enable ethically aligned, verifiably coherent, and resilient behavior for sustained, system-wide agreement.", "authors": ["Jinwei Hu", "Yi Dong", "Shuang Ao", "Zhuoyun Li", "Boxuan Wang", "Lokesh Singh", "Guangliang Cheng", "Sarvapali D. Ramchurn", "Xiaowei Huang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-15", "url": "https://arxiv.org/abs/2510.14008", "pdf_url": "https://arxiv.org/pdf/2510.14008v2", "arxiv_id": "2510.14008", "doi": "10.48550/arXiv.2510.14008", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3392}
{"id": "02baaa1569e7745290ba6b5c333f08b2e98f3742f73ea30069fadc3c87ba7942", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Engineering Multi-Agent LLMs: A Protocol-Driven Approach", "abstract": "The increasing demand for software development has driven interest in automating software engineering (SE) tasks using Large Language Models (LLMs). Recent efforts extend LLMs into multi-agent systems (MAS) that emulate collaborative development workflows, but these systems often fail due to three core deficiencies: under-specification, coordination misalignment, and inappropriate verification, arising from the absence of foundational SE structuring principles. This paper introduces Software Engineering Multi-Agent Protocol (SEMAP), a protocol-layer methodology that instantiates three core SE design principles for multi-agent LLMs: (1) explicit behavioral contract modeling, (2) structured messaging, and (3) lifecycle-guided execution with verification, and is implemented atop Google's Agent-to-Agent (A2A) infrastructure. Empirical evaluation using the Multi-Agent System Failure Taxonomy (MAST) framework demonstrates that SEMAP effectively reduces failures across different SE tasks. In code development, it achieves up to a 69.6% reduction in total failures for function-level development and 56.7% for deployment-level development. For vulnerability detection, SEMAP reduces failure counts by up to 47.4% on Python tasks and 28.2% on C/C++ tasks.", "authors": ["Zhenyu Mao", "Jacky Keung", "Fengji Zhang", "Shuo Liu", "Yifei Wang", "Jialong Li"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-14", "url": "https://arxiv.org/abs/2510.12120", "pdf_url": "https://arxiv.org/pdf/2510.12120v1", "arxiv_id": "2510.12120", "doi": "10.1109/APSEC66846.2025.00100", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Asia-Pacific Software Engineering Conference", "quality_score": 0.338}
{"id": "c9f6abcf5ecd0e9e12c752a1047c1eebbb44f95a4df27b2ab82b8c488210c5f2", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Debate for LLM Judges with Adaptive Stability Detection", "abstract": "With advancements in reasoning capabilities, Large Language Models (LLMs) are increasingly employed for automated judgment tasks. While LLMs-as-Judges offer promise in automating evaluations, current approaches often rely on simplistic aggregation methods (e.g., majority voting), which can fail even when individual agents provide correct answers. To address this, we propose a multi-agent debate judge framework where agents collaboratively reason and iteratively refine their responses. We formalize the debate process mathematically, analyzing agent interactions and proving that debate amplifies correctness compared to static ensembles. To enhance efficiency, we introduce a stability detection mechanism that models judge consensus dynamics via a time-varying Beta-Binomial mixture, with adaptive stopping based on distributional similarity (Kolmogorov-Smirnov test). This mechanism models the judges' collective correct rate dynamics using a time-varying mixture of Beta-Binomial distributions and employs an adaptive stopping criterion based on distributional similarity (Kolmogorov-Smirnov statistic). Experiments across multiple benchmarks and models demonstrate that our framework improves judgment accuracy over majority voting while maintaining computational efficiency.", "authors": ["Tianyu Hu", "Zhen Tan", "Song Wang", "Huaizhi Qu", "Tianlong Chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-14", "url": "https://arxiv.org/abs/2510.12697", "pdf_url": "https://arxiv.org/pdf/2510.12697v1", "arxiv_id": "2510.12697", "doi": "10.48550/arXiv.2510.12697", "citation_count": 9, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.338}
{"id": "1892cba2bc59a8908d9b4737731a7175aba855e8ff49283ebdc0ef8f6379f691", "sources": ["arxiv", "semantic_scholar"], "title": "Benefits and Limitations of Communication in Multi-Agent Reasoning", "abstract": "Chain-of-thought prompting has popularized step-by-step reasoning in large language models, yet model performance still degrades as problem complexity and context length grow. By decomposing difficult tasks with long contexts into shorter, manageable ones, recent multi-agent paradigms offer a promising near-term solution to this problem. However, the fundamental capacities of such systems are poorly understood. In this work, we propose a theoretical framework to analyze the expressivity of multi-agent systems. We apply our framework to three algorithmic families: state tracking, recall, and $k$-hop reasoning. We derive bounds on (i) the number of agents required to solve the task exactly, (ii) the quantity and structure of inter-agent communication, and (iii) the achievable speedups as problem size and context scale. Our results identify regimes where communication is provably beneficial, delineate tradeoffs between agent count and bandwidth, and expose intrinsic limitations when either resource is constrained. We complement our theoretical analysis with a set of experiments on pretrained LLMs using controlled synthetic benchmarks. Empirical outcomes confirm the tradeoffs between key quantities predicted by our theory. Collectively, our analysis offers principled guidance for designing scalable multi-agent reasoning systems.", "authors": ["Michael Rizvi-Martel", "Satwik Bhattamishra", "Neil Rathi", "Guillaume Rabusseau", "Michael Hahn"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-14", "url": "https://arxiv.org/abs/2510.13903", "pdf_url": "https://arxiv.org/pdf/2510.13903v1", "arxiv_id": "2510.13903", "doi": "10.48550/arXiv.2510.13903", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.338}
{"id": "eda9635eac189526881811363d3080c3866d76ee844afe339bee914a8a6126ff", "sources": ["arxiv", "semantic_scholar"], "title": "KVCOMM: Online Cross-context KV-cache Communication for Efficient LLM-based Multi-agent Systems", "abstract": "Multi-agent large language model (LLM) systems are increasingly adopted for complex language processing tasks that require communication and coordination among agents. However, these systems often suffer substantial overhead from repeated reprocessing of overlapping contexts across agents. In typical pipelines, once an agent receives a message from its predecessor, the full context-including prior turns-must be reprocessed from scratch, leading to inefficient processing. While key-value (KV) caching is an effective solution for avoiding redundant computation in single-agent settings where prefixes remain unchanged, it cannot be directly reused in multi-agent scenarios due to diverging prefixes introduced by agent-specific context extensions. We identify that the core challenge lies in the offset variance of KV-caches across agents. To address this, we propose KVCOMM, a training-free framework that enables efficient prefilling in multi-agent inference by reusing KV-caches and aligning cache offsets of overlapping contexts under diverse prefix contexts. KVCOMM estimates and adjusts KV-caches for shared content by referencing a pool of cached examples-termed anchors-that store observed cache deviations under varying prefixes. The anchor pool is maintained and updated online, allowing dynamic adaptation to distinct user requests and context structures. KVCOMM achieves over 70% reuse rate across diverse multi-agent workloads, including retrieval-augmented generation, math reasoning, and collaborative coding tasks, all without quality degradation. Particularly, when each fully-connected agent receives 1K input tokens with 512 prefix tokens and 512 output tokens under a five-agent setting, KVCOMM achieves up to 7.8x speedup compared to the standard prefill pipeline, reducing TTFT from ~430 ms to ~55 ms.", "authors": ["Hancheng Ye", "Zhengqi Gao", "Mingyuan Ma", "Qinsi Wang", "Yuzhe Fu", "Ming-Yu Chung", "Yueqian Lin", "Zhijian Liu", "Jianyi Zhang", "Danyang Zhuo", "Yiran Chen"], "categories": ["cs.MA", "cs.AI", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2025-10-14", "url": "https://arxiv.org/abs/2510.12872", "pdf_url": "https://arxiv.org/pdf/2510.12872v2", "arxiv_id": "2510.12872", "doi": "10.48550/arXiv.2510.12872", "citation_count": 17, "influential_citation_count": 7, "has_code": true, "code_url": "https://github.com/FastMAS/KVCOMM}", "venue": "arXiv.org", "quality_score": 0.5224}
{"id": "c0e4bbd9c08a53dc012c82d406752a401c2ad1e8eea5d17bdc863db250b659d0", "sources": ["arxiv", "semantic_scholar"], "title": "Proof-of-Use: Mitigating Tool-Call Hacking in Deep Research Agents", "abstract": "While reinforcement learning (RL) enhances their ability to plan and reason across retrieval steps, we identify a critical failure mode in this setting: Tool-Call Hacking. Unlike execution-based tools (e.g., code or math), whose effects are directly observable, the weak observability of causal dependencies between retrieved evidence and reasoning under format- and outcome-level supervision enables agents to maximize surface-level reward signals without genuinely grounding their reasoning in the returned evidence. This leads to distinctive pathologies, including mode collapse via tool overuse and hallucinated tool usage where tool calls are largely decorative. To address this issue, we propose Proof-of-Use (PoU), an evidence grounded RL framework that explicitly optimizes the causal dependency from retrieval to reasoning and final answers. PoU re-fomulate a fine-grained stepwise interaction protocol in which agents must auditably cite normalized evidence identifiers. We operationalize this via a multi-objective reward design consisting of: (1) two progressive process rewards that constrain citation validity at intermediate steps; (2) a global Answer--Support Alignment reward that enforces consistency between final answers and retrieved evidence; and (3) a curriculum-style adaptive reward mixing mechanism that smoothly transitions agents from dense process supervision to sparse outcome-based objectives. Extensive experiments show the strong performance of PoU and demonstrate the effectiveness in mitigating tool-call hacking. Beyond this, PoU exhibits a notable emergent property: adaptive and robust tool-usage patterns naturally arise under domain and tool shifts, even though PoU does not explicitly optimize for tool adaptation.", "authors": ["SHengjie Ma", "Chenlong Deng", "Jiaxin Mao", "Jiadeng Huang", "Teng Wang", "Junjie Wu", "Changwang Zhang", "Jun wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.10931", "pdf_url": "https://arxiv.org/pdf/2510.10931v2", "arxiv_id": "2510.10931", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2144}
{"id": "6b79ebd4b9a05f85703e3ffce6b21ea96fce76db2de28d53ac8431189801e011", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Shadows: Distributed Backdoor Attacks in LLM-Based Multi-Agent Systems", "abstract": "LLM-based multi-agent systems (MAS) demonstrate increasing integration into next-generation applications, but their safety in backdoor attacks remains largely underexplored. However, existing research has focused exclusively on single-agent backdoor attacks, overlooking the novel attack surfaces introduced by agent collaboration in MAS. To bridge this gap, we present the first Distributed Backdoor Attack tailored to MAS. We decompose the backdoor into multiple distributed attack primitives that are embedded within MAS tools. These primitives remain dormant individually but collectively activate only when agents collaborate in a specific sequence, thereby assembling the full backdoor to execute targeted attacks such as data exfiltration. To fully assess this threat, we introduce a benchmark for multi-role collaborative tasks and a sandboxed framework to evaluate. Extensive experiments demonstrate that our attack achieves an attack success rate exceeding 95% without degrading performance on benign tasks. This work exposes novel backdoor attack surfaces that exploit agent collaboration, underscoring the need to move beyond single-agent protection. Code and benchmark are available at https://github.com/whfeLingYu/Distributed-Backdoor-Attacks-in-MAS.", "authors": ["Pengyu Zhu", "Lijun Li", "Yaxing Lyu", "Li Sun", "Sen Su", "Jing Shao"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.11246", "pdf_url": "https://arxiv.org/pdf/2510.11246v1", "arxiv_id": "2510.11246", "doi": "10.48550/arXiv.2510.11246", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/whfeLingYu/Distributed-Backdoor-Attacks-in-MAS", "venue": "arXiv.org", "quality_score": 0.5206}
{"id": "776aa8d4952894d97212f686398a4915d279ddfd50b23225928e5115ef3ef253", "sources": ["arxiv", "semantic_scholar"], "title": "PaperArena: An Evaluation Benchmark for Tool-Augmented Agentic Reasoning on Scientific Literature", "abstract": "Understanding and reasoning on the large-scale scientific literature is a crucial touchstone for large language model (LLM) based agents. However, existing works are mainly restricted to tool-free tasks within single papers, largely due to the lack of a benchmark that evaluates cross-paper reasoning and multi-tool orchestration in authentic research scenarios. In this work, we propose PaperArena, a benchmark to evaluate LLM-based agents on questions that require integrating information across multiple papers with the assistance of external tools. Given a research question, agents should formulate a reasoning plan, interact with multiple papers, and invoke appropriate tools to produce a well-grounded answer. To support standardized evaluation, we provide a platform for agent execution, offering a modular tool environment including multimodal parsing, context retrieval, and programmatic computation. Experiments reveal that even the leading LLM powering a well-established agentic workflow achieves merely 38.78% average accuracy, while on the hard subset, accuracy drops to only 18.47%. We also analyze reasoning traces and diagnose agent behavior, providing the community with insights to develop and evaluate more capable scientific agents.", "authors": ["Daoyu Wang", "Mingyue Cheng", "Shuo Yu", "Zirui Liu", "Ze Guo", "Xin Li", "Qi Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.10909", "pdf_url": "https://arxiv.org/pdf/2510.10909v4", "arxiv_id": "2510.10909", "doi": "10.48550/arXiv.2510.10909", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3369}
{"id": "4ccd2db6d640958c8b6de498c8ebbb96ca6bec0c5cdc55c21bb0e489e6c6662e", "sources": ["arxiv", "semantic_scholar"], "title": "When Agents Trade: Live Multi-Market Trading Benchmark for LLM Agents", "abstract": "Although Large Language Model (LLM)-based agents are increasingly used in financial trading, it remains unclear whether they can reason and adapt in live markets, as most studies test models instead of agents, cover limited periods and assets, and rely on unverified data. To address these gaps, we introduce Agent Market Arena (AMA), the first lifelong, real-time benchmark for evaluating LLM-based trading agents across multiple markets. AMA integrates verified trading data, expert-checked news, and diverse agent architectures within a unified trading framework, enabling fair and continuous comparison under real conditions. It implements four agents, including InvestorAgent as a single-agent baseline, TradeAgent and HedgeFundAgent with different risk styles, and DeepFundAgent with memory-based reasoning, and evaluates them across GPT-4o, GPT-4.1, Claude-3.5-haiku, Claude-sonnet-4, and Gemini-2.0-flash. Live experiments on both cryptocurrency and stock markets demonstrate that agent frameworks display markedly distinct behavioral patterns, spanning from aggressive risk-taking to conservative decision-making, whereas model backbones contribute less to outcome variation. AMA thus establishes a foundation for rigorous, reproducible, and continuously evolving evaluation of financial reasoning and trading intelligence in LLM-based agents.", "authors": ["Lingfei Qian", "Xueqing Peng", "Yan Wang", "Vincent Jim Zhang", "Huan He", "Hanley Smith", "Yi Han", "Yueru He", "Haohang Li", "Yupeng Cao", "Yangyang Yu", "Alejandro Lopez-Lira", "Peng Lu", "Jian-Yun Nie", "Guojun Xiong", "Jimin Huang", "Sophia Ananiadou"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.11695", "pdf_url": "https://arxiv.org/pdf/2510.11695v2", "arxiv_id": "2510.11695", "doi": "10.48550/arXiv.2510.11695", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3369}
{"id": "135d19f221b409a58f7772428ed95eddc2bb94854f356fd6e8a5ca439130bc5d", "sources": ["arxiv", "semantic_scholar"], "title": "Stronger-MAS: Multi-Agent Reinforcement Learning for Collaborative LLMs", "abstract": "Multi-agent systems (MAS) and reinforcement learning (RL) are widely used to enhance the agentic capabilities of large language models (LLMs). MAS improves task performance through role-based orchestration, while RL uses environmental rewards to learn stronger policies, such as GRPO-style optimization. However, applying on-policy RL to MAS remains underexplored and presents unique challenges. Algorithmically, standard GRPO grouping assumptions break down because prompts vary by role and by turn. System-wise, the training stack must support MAS-workflow rollouts and on-policy updates for both single-policy and multi-policy models. We propose AT-GRPO, which includes (i) an agent- and turn-wise grouped RL algorithm tailored to MAS and (ii) a training system that supports both single- and multi-policy regimes. Across game, planning, coding, and math tasks, AT-GRPO delivers substantial gains. On long-horizon planning, it increases accuracy from a 14.0 to 47.0 percent single-agent RL baseline to 96.0 to 99.5 percent. It also improves reasoning performance, with average gains of 3.87 to 7.62 percent on coding tasks and 9.0 to 17.93 percent on math. Code and environments are available at: https://github.com/pettingllms-ai/PettingLLMs.", "authors": ["Yujie Zhao", "Lanxiang Hu", "Yang Wang", "Minmin Hou", "Hao Zhang", "Ke Ding", "Jishen Zhao"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.11062", "pdf_url": "https://arxiv.org/pdf/2510.11062v5", "arxiv_id": "2510.11062", "doi": null, "citation_count": 14, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/pettingllms-ai/PettingLLMs", "venue": null, "quality_score": 0.3981}
{"id": "41945b0242b8f8285ac4eaa3af28f9adb52df87dc47734e77fbfd0f0c2be0ccb", "sources": ["arxiv", "semantic_scholar"], "title": "LLM$\\times$MapReduce-V3: Enabling Interactive In-Depth Survey Generation through a MCP-Driven Hierarchically Modular Agent System", "abstract": "We introduce LLM x MapReduce-V3, a hierarchically modular agent system designed for long-form survey generation. Building on the prior work, LLM x MapReduce-V2, this version incorporates a multi-agent architecture where individual functional components, such as skeleton initialization, digest construction, and skeleton refinement, are implemented as independent model-context-protocol (MCP) servers. These atomic servers can be aggregated into higher-level servers, creating a hierarchically structured system. A high-level planner agent dynamically orchestrates the workflow by selecting appropriate modules based on their MCP tool descriptions and the execution history. This modular decomposition facilitates human-in-the-loop intervention, affording users greater control and customization over the research process. Through a multi-turn interaction, the system precisely captures the intended research perspectives to generate a comprehensive skeleton, which is then developed into an in-depth survey. Human evaluations demonstrate that our system surpasses representative baselines in both content depth and length, highlighting the strength of MCP-based modular planning.", "authors": ["Yu Chao", "Siyu Lin", "xiaorong wang", "Zhu Zhang", "Zihan Zhou", "Haoyu Wang", "Shuo Wang", "Jie Zhou", "Zhiyuan Liu", "Maosong Sun"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-13", "url": "https://arxiv.org/abs/2510.10890", "pdf_url": "https://arxiv.org/pdf/2510.10890v2", "arxiv_id": "2510.10890", "doi": "10.18653/v1/2025.emnlp-demos.51", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3369}
{"id": "80271edfdf983a976de338a7845358df23237c72a52511532a03f2907a18ef0c", "sources": ["arxiv", "semantic_scholar"], "title": "ATLAS: Adaptive Trading with LLM AgentS Through Dynamic Prompt Optimization and Multi-Agent Coordination", "abstract": "Large language models show promise for financial decision-making, yet deploying them as autonomous trading agents raises fundamental challenges: how to adapt instructions when rewards arrive late and obscured by market noise, how to synthesize heterogeneous information streams into coherent decisions, and how to bridge the gap between model outputs and executable market actions. We present ATLAS (Adaptive Trading with LLM AgentS), a unified multi-agent framework that integrates structured information from markets, news, and corporate fundamentals to support robust trading decisions. Within ATLAS, the central trading agent operates in an order-aware action space, ensuring that outputs correspond to executable market orders rather than abstract signals. The agent can incorporate feedback while trading using Adaptive-OPRO, a novel prompt-optimization technique that dynamically adapts the prompt by incorporating real-time, stochastic feedback, leading to increasing performance over time. Across regime-specific equity studies and multiple LLM families, Adaptive-OPRO consistently outperforms fixed prompts, while reflection-based feedback fails to provide systematic gains.", "authors": ["Charidimos Papadakis", "Angeliki Dimitriou", "Giorgos Filandrianos", "Maria Lymperaiou", "Konstantinos Thomas", "Giorgos Stamou"], "categories": ["q-fin.TR", "cs.AI"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2025-10-10", "url": "https://arxiv.org/abs/2510.15949", "pdf_url": "https://arxiv.org/pdf/2510.15949v5", "arxiv_id": "2510.15949", "doi": "10.48550/arXiv.2510.15949", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3334}
{"id": "b2205e39743f06cb5521894a199703f902974ce59b1cfcc15aea76c3ac849f34", "sources": ["arxiv", "semantic_scholar"], "title": "MASA: LLM-Driven Multi-Agent Systems for Autoformalization", "abstract": "Autoformalization serves a crucial role in connecting natural language and formal reasoning. This paper presents MASA, a novel framework for building multi-agent systems for autoformalization driven by Large Language Models (LLMs). MASA leverages collaborative agents to convert natural language statements into their formal representations. The architecture of MASA is designed with a strong emphasis on modularity, flexibility, and extensibility, allowing seamless integration of new agents and tools to adapt to a fast-evolving field. We showcase the effectiveness of MASA through use cases on real-world mathematical definitions and experiments on formal mathematics datasets. This work highlights the potential of multi-agent systems powered by the interaction of LLMs and theorem provers in enhancing the efficiency and reliability of autoformalization, providing valuable insights and support for researchers and practitioners in the field.", "authors": ["Lan Zhang", "Marco Valentino", "André Freitas"], "categories": ["cs.CL", "cs.FL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-10", "url": "https://arxiv.org/abs/2510.08988", "pdf_url": "https://arxiv.org/pdf/2510.08988v1", "arxiv_id": "2510.08988", "doi": "10.18653/v1/2025.emnlp-demos.44", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/lanzhang128/multi_agent_autoformalization", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.5153}
{"id": "66e0618cee72f2424fa884e4b735f815ca63e2c0fa56ead2fc51c747c91169f7", "sources": ["arxiv", "semantic_scholar"], "title": "Modeling Layered Consciousness with Multi-Agent Large Language Models", "abstract": "We propose a multi-agent framework for modeling artificial consciousness in large language models (LLMs), grounded in psychoanalytic theory. Our \\textbf{Psychodynamic Model} simulates self-awareness, preconsciousness, and unconsciousness through agent interaction, guided by a Personalization Module combining fixed traits and dynamic needs. Using parameter-efficient fine-tuning on emotionally rich dialogues, the system was evaluated across eight personalized conditions. An LLM as a judge approach showed a 71.2\\% preference for the fine-tuned model, with improved emotional depth and reduced output variance, demonstrating its potential for adaptive, personalized cognition.", "authors": ["Sang Hun Kim", "Jongmin Lee", "Dongkyu Park", "So Young Lee", "Yosep Chong"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-10", "url": "https://arxiv.org/abs/2510.17844", "pdf_url": "https://arxiv.org/pdf/2510.17844v1", "arxiv_id": "2510.17844", "doi": "10.48550/arXiv.2510.17844", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3334}
{"id": "6396f1fc5a7018df5f677b1714351b09b603f3dc907373d2edae16b515695228", "sources": ["arxiv", "semantic_scholar"], "title": "RA-Gen: A Controllable Code Generation Framework Using ReAct for Multi-Agent Task Execution", "abstract": "Code generation models based on large language models (LLMs) have gained wide adoption, but challenges remain in ensuring safety, accuracy, and controllability, especially for complex tasks. Existing methods often lack dynamic integration of external tools, transparent reasoning, and user control over safety. To address these issues, we propose a controllable code generation framework utilizing the ReAct paradigm for multi-agent task execution. This framework is a multi-agent system designed to enable efficient, precise, and interpretable code generation through dynamic interactions between LLMs and external resources. The framework adopts a collaborative architecture comprising four specialized agents: a Planner for task decomposition, a Searcher that leverages the ReAct framework for reasoning and tool integration, a CodeGen agent for accurate code generation, and an Extractor for structured data retrieval. The ReAct-based Searcher alternates between generating reasoning traces and executing actions, facilitating seamless integration of internal knowledge with external tools (such as search engines) to enhance accuracy and user control. Experimental results show the framework's effectiveness across multiple languages, achieving a 94.8% security rate on the SVEN dataset with CodeQL, outperforming existing approaches. Its transparent reasoning process fosters user trust and improves controllability.", "authors": ["Aofan Liu", "Haoxuan Li", "Bin Wang", "Ao Yang", "Hui Li"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-09", "url": "https://arxiv.org/abs/2510.08665", "pdf_url": "https://arxiv.org/pdf/2510.08665v1", "arxiv_id": "2510.08665", "doi": "10.1109/IJCNN64981.2025.11228706", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Joint Conference on Neural Network", "quality_score": 0.3323}
{"id": "44891a24fa2dde3e27494e09e512f084fc8b9e9d95893e8bcd2d75bffe822e5e", "sources": ["arxiv", "semantic_scholar"], "title": "Opponent Shaping in LLM Agents", "abstract": "Large Language Models (LLMs) are increasingly being deployed as autonomous agents in real-world environments. As these deployments scale, multi-agent interactions become inevitable, making it essential to understand strategic behavior in such systems. A central open question is whether LLM agents, like reinforcement learning agents, can shape the learning dynamics and influence the behavior of others through interaction alone. In this paper, we present the first investigation of opponent shaping (OS) with LLM-based agents. Existing OS algorithms cannot be directly applied to LLMs, as they require higher-order derivatives, face scalability constraints, or depend on architectural components that are absent in transformers. To address this gap, we introduce ShapeLLM, an adaptation of model-free OS methods tailored for transformer-based agents. Using ShapeLLM, we examine whether LLM agents can influence co-players' learning dynamics across diverse game-theoretic environments. We demonstrate that LLM agents can successfully guide opponents toward exploitable equilibria in competitive games (Iterated Prisoner's Dilemma, Matching Pennies, and Chicken) and promote coordination and improve collective welfare in cooperative games (Iterated Stag Hunt and a cooperative version of the Prisoner's Dilemma). Our findings show that LLM agents can both shape and be shaped through interaction, establishing opponent shaping as a key dimension of multi-agent LLM research.", "authors": ["Marta Emili Garcia Segura", "Stephen Hailes", "Mirco Musolesi"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-09", "url": "https://arxiv.org/abs/2510.08255", "pdf_url": "https://arxiv.org/pdf/2510.08255v1", "arxiv_id": "2510.08255", "doi": "10.48550/arXiv.2510.08255", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3323}
{"id": "353a1a4c6f7e7ec713d8938604c3d6efb588faa70516298e5f7db8b4c4c2f5fd", "sources": ["arxiv", "semantic_scholar"], "title": "SurveyG: A Multi-Agent LLM Framework with Hierarchical Citation Graph for Automated Survey Generation", "abstract": "Large language models (LLMs) are increasingly adopted for automating survey paper generation \\cite{wang2406autosurvey, liang2025surveyx, yan2025surveyforge,su2025benchmarking,wen2025interactivesurvey}. Existing approaches typically extract content from a large collection of related papers and prompt LLMs to summarize them directly. However, such methods often overlook the structural relationships among papers, resulting in generated surveys that lack a coherent taxonomy and a deeper contextual understanding of research progress. To address these shortcomings, we propose \\textbf{SurveyG}, an LLM-based agent framework that integrates \\textit{hierarchical citation graph}, where nodes denote research papers and edges capture both citation dependencies and semantic relatedness between their contents, thereby embedding structural and contextual knowledge into the survey generation process. The graph is organized into three layers: \\textbf{Foundation}, \\textbf{Development}, and \\textbf{Frontier}, to capture the evolution of research from seminal works to incremental advances and emerging directions. By combining horizontal search within layers and vertical depth traversal across layers, the agent produces multi-level summaries, which are consolidated into a structured survey outline. A multi-agent validation stage then ensures consistency, coverage, and factual accuracy in generating the final survey. Experiments, including evaluations by human experts and LLM-as-a-judge, demonstrate that SurveyG outperforms state-of-the-art frameworks, producing surveys that are more comprehensive and better structured to the underlying knowledge taxonomy of a field.", "authors": ["Minh-Anh Nguye", "Minh-Duc Nguyen", "Ha Lan N. T.", "Kieu Hai Dang", "Nguyen Tien Dong", "Dung D. Le"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-09", "url": "https://arxiv.org/abs/2510.07733", "pdf_url": "https://arxiv.org/pdf/2510.07733v3", "arxiv_id": "2510.07733", "doi": "10.48550/arXiv.2510.07733", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3323}
{"id": "2d988c3d947e7b99fa084f5dc1d5ab7817ba1da07e9cc11ff5e73af0c5339e7a", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic Generation of Multi-LLM Agents Communication Topologies with Graph Diffusion Models", "abstract": "The efficiency of multi-agent systems driven by large language models (LLMs) largely hinges on their communication topology. However, designing an optimal topology is a non-trivial challenge, as it requires balancing competing objectives such as task performance, communication cost, and robustness. Existing frameworks often rely on static or hand-crafted topologies, which inherently fail to adapt to diverse task requirements, leading to either excessive token consumption for simple problems or performance bottlenecks for complex ones. To address this challenge, we introduce a novel generative framework called \\textit{Guided Topology Diffusion (GTD)}. Inspired by conditional discrete graph diffusion models, GTD formulates topology synthesis as an iterative construction process. At each step, the generation is steered by a lightweight proxy model that predicts multi-objective rewards (e.g., accuracy, utility, cost), enabling real-time, gradient-free optimization towards task-adaptive topologies. This iterative, guided synthesis process distinguishes GTD from single-step generative frameworks, enabling it to better navigate complex design trade-offs. We validated GTD across multiple benchmarks, and experiments show that this framework can generate highly task-adaptive, sparse, and efficient communication topologies, significantly outperforming existing methods in LLM agent collaboration.", "authors": ["Eric Hanchen Jiang", "Mengting Li", "Guancheng Wan", "Sophia Yin", "Yuchen Wu", "Xiao Liang", "Xinfeng Li", "Yizhou Sun", "Wei Wang", "Kai-Wei Chang", "Ying Nian Wu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-09", "url": "https://arxiv.org/abs/2510.07799", "pdf_url": "https://arxiv.org/pdf/2510.07799v2", "arxiv_id": "2510.07799", "doi": "10.48550/arXiv.2510.07799", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3323}
{"id": "19fdaaf2e525ebbd67be3a3bb80a6f5b7e51d651d79ca29705299eed966db54c", "sources": ["arxiv", "semantic_scholar"], "title": "Traceability and Accountability in Role-Specialized Multi-Agent LLM Pipelines", "abstract": "Sequential multi-agent systems built with large language models (LLMs) can automate complex software tasks, but they are hard to trust because errors quietly pass from one stage to the next. We study a traceable and accountable pipeline, meaning a system with clear roles, structured handoffs, and saved records that let us trace who did what at each step and assign blame when things go wrong. Our setting is a Planner -> Executor -> Critic pipeline. We evaluate eight configurations of three state-of-the-art LLMs on three benchmarks and analyze where errors start, how they spread, and how they can be fixed. Our results show: (1) adding a structured, accountable handoff between agents markedly improves accuracy and prevents the failures common in simple pipelines; (2) models have clear role-specific strengths and risks (e.g., steady planning vs. high-variance critiquing), which we quantify with repair and harm rates; and (3) accuracy-cost-latency trade-offs are task-dependent, with heterogeneous pipelines often the most efficient. Overall, we provide a practical, data-driven method for designing, tracing, and debugging reliable, predictable, and accountable multi-agent systems.", "authors": ["Amine Barrak"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-08", "url": "https://arxiv.org/abs/2510.07614", "pdf_url": "https://arxiv.org/pdf/2510.07614v1", "arxiv_id": "2510.07614", "doi": "10.1109/ASEW67777.2025.00064", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "2025 40th ACM/IEEE International Conference on Automated Software Engineering Workshops", "quality_score": 0.3311}
{"id": "47475536752614c8a71835067429203066f299884105a8e9e7af64edd9d28f79", "sources": ["arxiv", "semantic_scholar"], "title": "COMPASS: Benchmarking Constrained Optimization in LLM Agents", "abstract": "Human decision-making often involves constrained optimization. As LLM agents are deployed to assist with real-world tasks like travel planning, shopping, and scheduling, they must mirror this capability. We introduce COMPASS, a benchmark that evaluates whether LLM agents can perform constrained optimization in realistic travel planning settings. To success in these tasks, agents must engage in multi-turn conversations with user to gather task information as well as use tools to gather information from the database. Then agents must propose a solution that not only satisfies hard constraints but also optimizes user's utility objective. Evaluating state-of-the-art models, we reveal a significant feasible-optimal gap: while models achieve 70-90% feasibility (constraint satisfaction), they reach only 20-60% optimality (utility optimization). Our analysis shows that tool use is not the bottleneck. Instead, the core limitation is insufficient exploration of the search space, with success strongly correlating with information gathered. Coding agents show a promising approach to mitigate this gap. Together, COMPASS provides a testbed for developing LLM agents that can truly mirror human decision-making by both satisfying constraints and optimizing objectives.", "authors": ["Tian Qin", "Felix Bai", "Ting-Yao Hu", "Raviteja Vemulapalli", "Hema Swetha Koppula", "Zhiyang Xu", "Bowen Jin", "Mert Cemri", "Jiarui Lu", "Zirui Wang", "Meng Cao"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-08", "url": "https://arxiv.org/abs/2510.07043", "pdf_url": "https://arxiv.org/pdf/2510.07043v2", "arxiv_id": "2510.07043", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2107}
{"id": "307ee31b7bd994d873116b1a5118361fad3d4b50c39dfb001b874baf3c7c249b", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Assisted Modeling of Semantic Web-Enabled Multi-Agents Systems with AJAN", "abstract": "There are many established semantic Web standards for implementing multi-agent driven applications. The AJAN framework allows to engineer multi-agent systems based on these standards. In particular, agent knowledge is represented in RDF/RDFS and OWL, while agent behavior models are defined with Behavior Trees and SPARQL to access and manipulate this knowledge. However, the appropriate definition of RDF/RDFS and SPARQL-based agent behaviors still remains a major hurdle not only for agent modelers in practice. For example, dealing with URIs is very error-prone regarding typos and dealing with complex SPARQL queries in large-scale environments requires a high learning curve. In this paper, we present an integrated development environment to overcome such hurdles of modeling AJAN agents and at the same time to extend the user community for AJAN by the possibility to leverage Large Language Models for agent engineering.", "authors": ["Hacane Hechehouche", "Andre Antakli", "Matthias Klusch"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-08", "url": "https://arxiv.org/abs/2510.06911", "pdf_url": "https://arxiv.org/pdf/2510.06911v1", "arxiv_id": "2510.06911", "doi": "10.48550/arXiv.2510.06911", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3311}
{"id": "ad8d59a0748191316a6756a9850246f0229be67fc47a2c2679aa02b84f5eada0", "sources": ["arxiv", "semantic_scholar"], "title": "MAPRO: Recasting Multi-Agent Prompt Optimization as Maximum a Posteriori Inference", "abstract": "Large language models (LLMs) have demonstrated remarkable capabilities across diverse tasks, and LLM-based agents further extend these abilities to various practical workflows. While recent progress shows that multi-agent systems (MAS) can outperform single agents by coordinating specialized roles, designing effective MAS remains difficult due to prompt sensitivity and the compounded instability MAS creates. To cope with the challenge, recent efforts in automated prompt design have reduced manual effort. However, multi-agent prompt optimization remains largely unexplored. Challenges like exponentially expanding search space and ambiguous credit assignment together make systematic design intractable without principled methods. Therefore, we introduce M}ulti-Agent PRompt Optimization (MAPRO), a four-stage framework that first formulates MAS prompt optimization as a Maximum a Posteriori (MAP) inference problem and solves it using a language-guided variant of max-product belief propagation algorithm. To address credit assignment and updates the system iteratively, MAPRO employs a topology-aware refinement mechanism that integrates execution feedback and downstream blames to selectively update agent prompts. Through this process, MAPRO progressively converges to a coordinated set of agent-specific prompt policies. Across benchmarks in various tasks, MAPRO achieves state-of-the-art performance, consistently surpassing manually engineered baselines and recent automated alternatives. Beyond performance, our MAP-based formulation also delivers general guidelines for building more reliable and principled multi-agent systems in the future", "authors": ["Zheyuan Zhang", "Lin Ge", "Hongjiang Li", "Weicheng Zhu", "Chuxu Zhang", "Yanfang Ye"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-08", "url": "https://arxiv.org/abs/2510.07475", "pdf_url": "https://arxiv.org/pdf/2510.07475v1", "arxiv_id": "2510.07475", "doi": "10.48550/arXiv.2510.07475", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.3311}
{"id": "14469f399b9144e768d1768644d4accf4afa1bccddd8150df2e88b7eb4591a03", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Framework for Stateful Inference-Time Search", "abstract": "Recent work explores agentic inference-time techniques to perform structured, multi-step reasoning. However, stateless inference often struggles on multi-step tasks due to the absence of persistent state. Moreover, task-specific fine-tuning or instruction-tuning often achieve surface-level code generation but remain brittle on tasks requiring deeper reasoning and long-horizon dependencies. To address these limitations, we propose stateful multi-agent evolutionary search, a training-free framework that departs from prior stateless approaches by combining (i) persistent inference-time state, (ii) adversarial mutation, and (iii) evolutionary preservation. We demonstrate its effectiveness in automated unit test generation through the generation of edge cases. We generate robust edge cases using an evolutionary search process, where specialized agents sequentially propose, mutate, and score candidates. A controller maintains persistent state across generations, while evolutionary preservation ensures diversity and exploration across all possible cases. This yields a generalist agent capable of discovering robust, high-coverage edge cases across unseen codebases. Experiments show our stateful multi-agent inference framework achieves substantial gains in coverage over stateless single-step baselines, evaluated on prevalent unit-testing benchmarks such as HumanEval and TestGenEvalMini and using three diverse LLM families - Llama, Gemma, and GPT. These results indicate that combining persistent inference-time state with evolutionary search materially improves unit-test generation.", "authors": ["Arshika Lalan", "Rajat Ghosh", "Aditya Kolsur", "Debojyoti Dutta"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-08", "url": "https://arxiv.org/abs/2510.07147", "pdf_url": "https://arxiv.org/pdf/2510.07147v1", "arxiv_id": "2510.07147", "doi": "10.48550/arXiv.2510.07147", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3311}
{"id": "22bb674fdcc0e214a659657b8b7b30dd87d747cc82a4f960313221ac789827e7", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Tool-Integrated Policy Optimization", "abstract": "Large language models (LLMs) increasingly rely on multi-turn tool-integrated planning for knowledge-intensive and complex reasoning tasks. Existing implementations typically rely on a single agent, but they suffer from limited context length and noisy tool responses. A natural solution is to adopt a multi-agent framework with planner- and worker-agents to manage context. However, no existing methods support effective reinforcement learning post-training of tool-integrated multi-agent frameworks. To address this gap, we propose Multi-Agent Tool-Integrated Policy Optimization (MATPO), which enables distinct roles (planner and worker) to be trained within a single LLM instance using role-specific prompts via reinforcement learning. MATPO is derived from a principled credit assignment mechanism across planner and worker rollouts. This design eliminates the need to deploy multiple LLMs, which would be memory-intensive, while preserving the benefits of specialization. Experiments on GAIA-text, WebWalkerQA, and FRAMES show that MATPO consistently outperforms single-agent baselines by an average of 18.38% relative improvement in performance and exhibits greater robustness to noisy tool outputs. Our findings highlight the effectiveness of unifying multiple agent roles within a single LLM and provide practical insights for stable and efficient multi-agent RL training.", "authors": ["Zhanfeng Mo", "Xingxuan Li", "Yuntao Chen", "Lidong Bing"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-06", "url": "https://arxiv.org/abs/2510.04678", "pdf_url": "https://arxiv.org/pdf/2510.04678v1", "arxiv_id": "2510.04678", "doi": "10.48550/arXiv.2510.04678", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3289}
{"id": "037e19701d2205f0a217ccd56b76c7472ff0e1f62aece3ed0e4848723de2af3d", "sources": ["arxiv", "semantic_scholar"], "title": "LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation", "abstract": "We introduce LEGOMem, a modular procedural memory framework for multi-agent large language model (LLM) systems in workflow automation. LEGOMem decomposes past task trajectories into reusable memory units and flexibly allocates them across orchestrators and task agents to support planning and execution. To explore the design space of memory in multi-agent systems, we use LEGOMem as a lens and conduct a systematic study of procedural memory in multi-agent systems, examining where memory should be placed, how it should be retrieved, and which agents benefit most. Experiments on the OfficeBench benchmark show that orchestrator memory is critical for effective task decomposition and delegation, while fine-grained agent memory improves execution accuracy. We find that even teams composed of smaller language models can benefit substantially from procedural memory, narrowing the performance gap with stronger agents by leveraging prior execution traces for more accurate planning and tool use. These results position LEGOMem as both a practical framework for memory-augmented agent systems and a research tool for understanding memory design in multi-agent workflow automation.", "authors": ["Dongge Han", "Camille Couturier", "Daniel Madrigal Diaz", "Xuchao Zhang", "Victor Rühle", "Saravan Rajmohan"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-06", "url": "https://arxiv.org/abs/2510.04851", "pdf_url": "https://arxiv.org/pdf/2510.04851v1", "arxiv_id": "2510.04851", "doi": "10.48550/arXiv.2510.04851", "citation_count": 22, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3404}
{"id": "dac2a4dbabb5b9a584b409dbf40c0d4ebd6fbf8acd061f963701496c4fe0a3c4", "sources": ["arxiv", "semantic_scholar"], "title": "AgentRouter: A Knowledge-Graph-Guided LLM Router for Collaborative Multi-Agent Question Answering", "abstract": "Large language models (LLMs) and agent-based frameworks have advanced rapidly, enabling diverse applications. Yet, with the proliferation of models and agentic strategies, practitioners face substantial uncertainty in selecting the best configuration for a downstream task. Prior studies show that different agents and backbones exhibit complementary strengths, and that larger models are not always superior, underscoring the need for adaptive routing mechanisms. Existing approaches to agent routing, however, often emphasize cost efficiency while overlooking the fine-grained contextual and relational structure inherent in QA tasks. In this paper, we propose tAgentRouter, a framework that formulates multi-agent QA as a knowledge-graph-guided routing problem supervised by empirical performance signals. Specifically, we convert QA instance into a knowledge graph that jointly encodes queries, contextual entities, and agents, and then train a heterogeneous graph neural network (GNN) to propagate information across node types and produce task-aware routing distributions over agents. By leveraging soft supervision and weighted aggregation of agent outputs, AgentRouter learns principled collaboration schemes that capture the complementary strengths of diverse agents. Extensive experiments demonstrate that our framework consistently outperforms single-agent and ensemble baselines, while generalizing across benchmarks and LLM backbones. These results highlight the effectiveness and robustness of graph-supervised multi-agent routing for question answering.", "authors": ["Zheyuan Zhang", "Kaiwen Shi", "Zhengqing Yuan", "Zehong Wang", "Tianyi Ma", "Keerthiram Murugesan", "Vincent Galassi", "Chuxu Zhang", "Yanfang Ye"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-06", "url": "https://arxiv.org/abs/2510.05445", "pdf_url": "https://arxiv.org/pdf/2510.05445v1", "arxiv_id": "2510.05445", "doi": "10.48550/arXiv.2510.05445", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3289}
{"id": "2ae5495df50ef09f982aa0b9520dd6c97ae49d99396328149c2e66006e1d3f87", "sources": ["arxiv", "semantic_scholar"], "title": "Biomedical reasoning in action: Multi-agent System for Auditable Biomedical Evidence Synthesis", "abstract": "We present M-Reason, a demonstration system for transparent, agent-based reasoning and evidence integration in the biomedical domain, with a focus on cancer research. M-Reason leverages recent advances in large language models (LLMs) and modular agent orchestration to automate evidence retrieval, appraisal, and synthesis across diverse biomedical data sources. Each agent specializes in a specific evidence stream, enabling parallel processing and fine-grained analysis. The system emphasizes explainability, structured reporting, and user auditability, providing complete traceability from source evidence to final conclusions. We discuss critical tradeoffs between agent specialization, system complexity, and resource usage, as well as the integration of deterministic code for validation. An open, interactive user interface allows researchers to directly observe, explore and evaluate the multi-agent workflow. Our evaluation demonstrates substantial gains in efficiency and output consistency, highlighting M-Reason's potential as both a practical tool for evidence synthesis and a testbed for robust multi-agent LLM systems in scientific research, available at https://m-reason.digitalecmt.com.", "authors": ["Oskar Wysocki", "Magdalena Wysocka", "Mauricio Jacobo", "Harriet Unsworth", "André Freitas"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-06", "url": "https://arxiv.org/abs/2510.05335", "pdf_url": "https://arxiv.org/pdf/2510.05335v1", "arxiv_id": "2510.05335", "doi": "10.48550/arXiv.2510.05335", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3289}
{"id": "6db20f3411c652f9dc6cd665565b604de8aefa5903ae7d3b3b5b55d21cae57a6", "sources": ["arxiv", "semantic_scholar"], "title": "Social Agent: Mastering Dyadic Nonverbal Behavior Generation via Conversational LLM Agents", "abstract": "We present Social Agent, a novel framework for synthesizing realistic and contextually appropriate co-speech nonverbal behaviors in dyadic conversations. In this framework, we develop an agentic system driven by a Large Language Model (LLM) to direct the conversation flow and determine appropriate interactive behaviors for both participants. Additionally, we propose a novel dual-person gesture generation model based on an auto-regressive diffusion model, which synthesizes coordinated motions from speech signals. The output of the agentic system is translated into high-level guidance for the gesture generator, resulting in realistic movement at both the behavioral and motion levels. Furthermore, the agentic system periodically examines the movements of interlocutors and infers their intentions, forming a continuous feedback loop that enables dynamic and responsive interactions between the two participants. User studies and quantitative evaluations show that our model significantly improves the quality of dyadic interactions, producing natural, synchronized nonverbal behaviors.", "authors": ["Zeyi Zhang", "Yanju Zhou", "Heyuan Yao", "Tenglong Ao", "Xiaohang Zhan", "Libin Liu"], "categories": ["cs.GR", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-06", "url": "https://arxiv.org/abs/2510.04637", "pdf_url": "https://arxiv.org/pdf/2510.04637v1", "arxiv_id": "2510.04637", "doi": "10.1145/3757377.3763879", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia", "quality_score": 0.3289}
{"id": "1869366482044a2cb77eebedd406eb55e7d4e04b0ff5eb75fb393a6627cd07ff", "sources": ["arxiv", "semantic_scholar"], "title": "On the Importance of Task Complexity in Evaluating LLM-Based Multi-Agent Systems", "abstract": "Large language model multi-agent systems (LLM-MAS) offer a promising paradigm for harnessing collective intelligence to achieve more advanced forms of AI behaviour. While recent studies suggest that LLM-MAS can outperform LLM single-agent systems (LLM-SAS) on certain tasks, the lack of systematic experimental designs limits the strength and generality of these conclusions. We argue that a principled understanding of task complexity, such as the degree of sequential reasoning required and the breadth of capabilities involved, is essential for assessing the effectiveness of LLM-MAS in task solving. To this end, we propose a theoretical framework characterising tasks along two dimensions: depth, representing reasoning length, and width, representing capability diversity. We theoretically examine a representative class of LLM-MAS, namely the multi-agent debate system, and empirically evaluate its performance in both discriminative and generative tasks with varying depth and width. Theoretical and empirical results show that the benefit of LLM-MAS over LLM-SAS increases with both task depth and width, and the effect is more pronounced with respect to depth. This clarifies when LLM-MAS are beneficial and provides a principled foundation for designing future LLM-MAS methods and benchmarks.", "authors": ["Bohan Tang", "Huidong Liang", "Keyue Jiang", "Xiaowen Dong"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-05", "url": "https://arxiv.org/abs/2510.04311", "pdf_url": "https://arxiv.org/pdf/2510.04311v1", "arxiv_id": "2510.04311", "doi": "10.48550/arXiv.2510.04311", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3277}
{"id": "f6c74762456406f67cf56d5bce06c8e65c758e0afaeba006c6fd9f23b2d05826", "sources": ["arxiv", "semantic_scholar"], "title": "Emergent Coordination in Multi-Agent Language Models", "abstract": "When are multi-agent LLM systems merely a collection of individual agents versus an integrated collective with higher-order structure? We introduce an information-theoretic framework to test -- in a purely data-driven way -- whether multi-agent systems show signs of higher-order structure. This information decomposition lets us measure whether dynamical emergence is present in multi-agent LLM systems, localize it, and distinguish spurious temporal coupling from performance-relevant cross-agent synergy. We implement a practical criterion and an emergence capacity criterion operationalized as partial information decomposition of time-delayed mutual information (TDMI). We apply our framework to experiments using a simple guessing game without direct agent communication and minimal group-level feedback with three randomized interventions. Groups in the control condition exhibit strong temporal synergy but little coordinated alignment across agents. Assigning a persona to each agent introduces stable identity-linked differentiation. Combining personas with an instruction to ``think about what other agents might do'' shows identity-linked differentiation and goal-directed complementarity across agents. Taken together, our framework establishes that multi-agent LLM systems can be steered with prompt design from mere aggregates to higher-order collectives. Our results are robust across emergence measures and entropy estimators, and not explained by coordination-free baselines or temporal dynamics alone. Without attributing human-like cognition to the agents, the patterns of interaction we observe mirror well-established principles of collective intelligence in human groups: effective performance requires both alignment on shared objectives and complementary contributions across members.", "authors": ["Christoph Riedl"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-05", "url": "https://arxiv.org/abs/2510.05174", "pdf_url": "https://arxiv.org/pdf/2510.05174v4", "arxiv_id": "2510.05174", "doi": "10.48550/arXiv.2510.05174", "citation_count": 14, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3277}
{"id": "50d09975fa87fdd450ecf1aab84977a579b085a75d0ffaec5b5678baeeb977cc", "sources": ["arxiv", "semantic_scholar"], "title": "NegotiationGym: Self-Optimizing Agents in a Multi-Agent Social Simulation Environment", "abstract": "We design and implement NegotiationGym, an API and user interface for configuring and running multi-agent social simulations focused upon negotiation and cooperation. The NegotiationGym codebase offers a user-friendly, configuration-driven API that enables easy design and customization of simulation scenarios. Agent-level utility functions encode optimization criteria for each agent, and agents can self-optimize by conducting multiple interaction rounds with other agents, observing outcomes, and modifying their strategies for future rounds.", "authors": ["Shashank Mangla", "Chris Hokamp", "Jack Boylan", "Demian Gholipour Ghalandari", "Yuuv Jauhari", "Lauren Cassidy", "Oisin Duffy"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-05", "url": "https://arxiv.org/abs/2510.04368", "pdf_url": "https://arxiv.org/pdf/2510.04368v1", "arxiv_id": "2510.04368", "doi": "10.48550/arXiv.2510.04368", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3277}
{"id": "6225a988ffa1c102363fe82f99c0ce5ef1137ce50c5a890b00332364fa3487e1", "sources": ["arxiv", "semantic_scholar"], "title": "Lang-PINN: From Language to Physics-Informed Neural Networks via a Multi-Agent Framework", "abstract": "Physics-informed neural networks (PINNs) provide a powerful approach for solving partial differential equations (PDEs), but constructing a usable PINN remains labor-intensive and error-prone. Scientists must interpret problems as PDE formulations, design architectures and loss functions, and implement stable training pipelines. Existing large language model (LLM) based approaches address isolated steps such as code generation or architecture suggestion, but typically assume a formal PDE is already specified and therefore lack an end-to-end perspective. We present Lang-PINN, an LLM-driven multi-agent system that builds trainable PINNs directly from natural language task descriptions. Lang-PINN coordinates four complementary agents: a PDE Agent that parses task descriptions into symbolic PDEs, a PINN Agent that selects architectures, a Code Agent that generates modular implementations, and a Feedback Agent that executes and diagnoses errors for iterative refinement. This design transforms informal task statements into executable and verifiable PINN code. Experiments show that Lang-PINN achieves substantially lower errors and greater robustness than competitive baselines: mean squared error (MSE) is reduced by up to 3--5 orders of magnitude, end-to-end execution success improves by more than 50\\%, and reduces time overhead by up to 74\\%.", "authors": ["Xin He", "Liangliang You", "Hongduan Tian", "Bo Han", "Ivor Tsang", "Yew-Soon Ong"], "categories": ["cs.AI", "cs.CE", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-03", "url": "https://arxiv.org/abs/2510.05158", "pdf_url": "https://arxiv.org/pdf/2510.05158v1", "arxiv_id": "2510.05158", "doi": "10.48550/arXiv.2510.05158", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3254}
{"id": "546fd5bf36b9c7772dc7a2280fa95d1dad9a2ea071328c752a8a6c0e2eb4ac8d", "sources": ["arxiv", "semantic_scholar"], "title": "ALMAS: an Autonomous LLM-based Multi-Agent Software Engineering Framework", "abstract": "Multi-agent Large Language Model (LLM) systems have been leading the way in applied LLM research across a number of fields. One notable area is software development, where researchers have advanced the automation of code implementation, code testing, code maintenance, inter alia, using LLM agents. However, software development is a multifaceted environment that extends beyond just code. As such, a successful LLM system must factor in multiple stages of the software development life-cycle (SDLC). In this paper, we propose a vision for ALMAS, an Autonomous LLM-based Multi-Agent Software Engineering framework, which follows the above SDLC philosophy such that it may work within an agile software development team to perform several tasks end-to-end. ALMAS aligns its agents with agile roles, and can be used in a modular fashion to seamlessly integrate with human developers and their development environment. We showcase the progress towards ALMAS through our published works and a use case demonstrating the framework, where ALMAS is able to seamlessly generate an application and add a new feature.", "authors": ["Vali Tawosi", "Keshav Ramani", "Salwa Alamir", "Xiaomo Liu"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-03", "url": "https://arxiv.org/abs/2510.03463", "pdf_url": "https://arxiv.org/pdf/2510.03463v2", "arxiv_id": "2510.03463", "doi": "10.1109/ASEW67777.2025.00059", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2258}
{"id": "db3d11ac265c914fcd71c56d4fcc09680ec41460dfa05d526810afa48573d2ec", "sources": ["arxiv", "semantic_scholar"], "title": "MALF: A Multi-Agent LLM Framework for Intelligent Fuzzing of Industrial Control Protocols", "abstract": "Industrial control systems (ICS) are vital to modern infrastructure but increasingly vulnerable to cybersecurity threats, particularly through weaknesses in their communication protocols. This paper presents MALF (Multi-Agent LLM Fuzzing Framework), an advanced fuzzing solution that integrates large language models (LLMs) with multi-agent coordination to identify vulnerabilities in industrial control protocols (ICPs). By leveraging Retrieval-Augmented Generation (RAG) for domain-specific knowledge and QLoRA fine-tuning for protocol-aware input generation, MALF enhances fuzz testing precision and adaptability. The multi-agent framework optimizes seed generation, mutation strategies, and feedback-driven refinement, leading to improved vulnerability discovery. Experiments on protocols like Modbus/TCP, S7Comm, and Ethernet/IP demonstrate that MALF surpasses traditional methods, achieving a test case pass rate (TCPR) of 88-92% and generating more exception triggers (ETN). MALF also maintains over 90% seed coverage and Shannon entropy values between 4.2 and 4.6 bits, ensuring diverse, protocol-compliant mutations. Deployed in a real-world Industrial Attack-Defense Range for power plants, MALF identified critical vulnerabilities, including three zero-day flaws, one confirmed and registered by CNVD. These results validate MALF's effectiveness in real-world fuzzing applications. This research highlights the transformative potential of multi-agent LLMs in ICS cybersecurity, offering a scalable, automated framework that sets a new standard for vulnerability discovery and strengthens critical infrastructure security against emerging threats.", "authors": ["Bowei Ning", "Xuejun Zong", "Kan He"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-03", "url": "https://arxiv.org/abs/2510.02694", "pdf_url": "https://arxiv.org/pdf/2510.02694v1", "arxiv_id": "2510.02694", "doi": "10.48550/arXiv.2510.02694", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3254}
{"id": "62e256f20b0dd653e6a34cb379c314eb2c92fa47f2ef74e5fabaa2389978aa01", "sources": ["arxiv", "semantic_scholar"], "title": "AMAS: Adaptively Determining Communication Topology for LLM-based Multi-Agent System", "abstract": "Although large language models (LLMs) have revolutionized natural language processing capabilities, their practical implementation as autonomous multi-agent systems (MAS) for industrial problem-solving encounters persistent barriers. Conventional MAS architectures are fundamentally restricted by inflexible, hand-crafted graph topologies that lack contextual responsiveness, resulting in diminished efficacy across varied academic and commercial workloads. To surmount these constraints, we introduce AMAS, a paradigm-shifting framework that redefines LLM-based MAS through a novel dynamic graph designer. This component autonomously identifies task-specific optimal graph configurations via lightweight LLM adaptation, eliminating the reliance on monolithic, universally applied structural templates. Instead, AMAS exploits the intrinsic properties of individual inputs to intelligently direct query trajectories through task-optimized agent pathways. Rigorous validation across question answering, mathematical deduction, and code generation benchmarks confirms that AMAS systematically exceeds state-of-the-art single-agent and multi-agent approaches across diverse LLM architectures. Our investigation establishes that context-sensitive structural adaptability constitutes a foundational requirement for high-performance LLM MAS deployments.", "authors": ["Hui Yi Leong", "Yuheng Li", "Yuqing Wu", "Wenwen Ouyang", "Wei Zhu", "Jiechao Gao", "Wei Han"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-02", "url": "https://arxiv.org/abs/2510.01617", "pdf_url": "https://arxiv.org/pdf/2510.01617v3", "arxiv_id": "2510.01617", "doi": "10.48550/arXiv.2510.01617", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3243}
{"id": "3e8b5f7b3989f7e86ae1f3c6247193ed1e982b0545c38f3cceddfaf17ed5aa63", "sources": ["arxiv", "semantic_scholar"], "title": "TalkPlay-Tools: Conversational Music Recommendation with LLM Tool Calling", "abstract": "While the recent developments in large language models (LLMs) have successfully enabled generative recommenders with natural language interactions, their recommendation behavior is limited, leaving other simpler yet crucial components such as metadata or attribute filtering underutilized in the system. We propose an LLM-based music recommendation system with tool calling to serve as a unified retrieval-reranking pipeline. Our system positions an LLM as an end-to-end recommendation system that interprets user intent, plans tool invocations, and orchestrates specialized components: boolean filters (SQL), sparse retrieval (BM25), dense retrieval (embedding similarity), and generative retrieval (semantic IDs). Through tool planning, the system predicts which types of tools to use, their execution order, and the arguments needed to find music matching user preferences, supporting diverse modalities while seamlessly integrating multiple database filtering methods. We demonstrate that this unified tool-calling framework achieves competitive performance across diverse recommendation scenarios by selectively employing appropriate retrieval methods based on user queries, envisioning a new paradigm for conversational music recommendation systems.", "authors": ["Seungheon Doh", "Keunwoo Choi", "Juhan Nam"], "categories": ["cs.IR", "cs.MM", "cs.SD", "eess.AS"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-10-02", "url": "https://arxiv.org/abs/2510.01698", "pdf_url": "https://arxiv.org/pdf/2510.01698v4", "arxiv_id": "2510.01698", "doi": "10.48550/arXiv.2510.01698", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3243}
{"id": "6e18244591a71b25024587321a33e8ee3a19c0c38bffc8048bedc8f37eb048d2", "sources": ["arxiv", "semantic_scholar"], "title": "InfoMosaic-Bench: Evaluating Multi-Source Information Seeking in Tool-Augmented Agents", "abstract": "Information seeking is a fundamental requirement for humans. However, existing LLM agents rely heavily on open-web search, which exposes two fundamental weaknesses: online content is noisy and unreliable, and many real-world tasks require precise, domain-specific knowledge unavailable from the web. The emergence of the Model Context Protocol (MCP) now allows agents to interface with thousands of specialized tools, seemingly resolving this limitation. Yet it remains unclear whether agents can effectively leverage such tools -- and more importantly, whether they can integrate them with general-purpose search to solve complex tasks. Therefore, we introduce InfoMosaic-Bench, the first benchmark dedicated to multi-source information seeking in tool-augmented agents. Covering six representative domains (medicine, finance, maps, video, web, and multi-domain integration), InfoMosaic-Bench requires agents to combine general-purpose search with domain-specific tools. Tasks are synthesized with InfoMosaic-Flow, a scalable pipeline that grounds task conditions in verified tool outputs, enforces cross-source dependencies, and filters out shortcut cases solvable by trivial lookup. This design guarantees both reliability and non-triviality. Experiments with 14 state-of-the-art LLM agents reveal three findings: (i) web information alone is insufficient, with GPT-5 achieving only 38.2% accuracy and 67.5% pass rate; (ii) domain tools provide selective but inconsistent benefits, improving some domains while degrading others; and (iii) 22.4% of failures arise from incorrect tool usage or selection, highlighting that current LLMs still struggle with even basic tool handling.", "authors": ["Yaxin Du", "Yuanshuo Zhang", "Xiyuan Yang", "Yifan Zhou", "Cheng Wang", "Gongyi Zou", "Xianghe Pang", "Wenhao Wang", "Menglan Chen", "Shuo Tang", "Zhiyu Li", "Feiyu Xiong", "Siheng Chen"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-02", "url": "https://arxiv.org/abs/2510.02271", "pdf_url": "https://arxiv.org/pdf/2510.02271v2", "arxiv_id": "2510.02271", "doi": "10.48550/arXiv.2510.02271", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3243}
{"id": "f8f08b73a7a179f98c81a8fc8ae9c619e4389c640b520ea22139fb06748888e3", "sources": ["arxiv", "semantic_scholar"], "title": "AgentRec: Next-Generation LLM-Powered Multi-Agent Collaborative Recommendation with Adaptive Intelligence", "abstract": "Interactive conversational recommender systems have gained significant attention for their ability to capture user preferences through natural language interactions. However, existing approaches face substantial challenges in handling dynamic user preferences, maintaining conversation coherence, and balancing multiple ranking objectives simultaneously. This paper introduces AgentRec, a next-generation LLM-powered multi-agent collaborative recommendation framework that addresses these limitations through hierarchical agent networks with adaptive intelligence. Our approach employs specialized LLM-powered agents for conversation understanding, preference modeling, context awareness, and dynamic ranking, coordinated through an adaptive weighting mechanism that learns from interaction patterns. We propose a three-tier learning strategy combining rapid response for simple queries, intelligent reasoning for complex preferences, and deep collaboration for challenging scenarios. Extensive experiments on three real-world datasets demonstrate that AgentRec achieves consistent improvements over state-of-the-art baselines, with 2.8\\% enhancement in conversation success rate, 1.9\\% improvement in recommendation accuracy (NDCG@10), and 3.2\\% better conversation efficiency while maintaining comparable computational costs through intelligent agent coordination.", "authors": ["Bo Ma", "Hang Li", "ZeHua Hu", "XiaoFan Gui", "LuYao Liu", "Simon Lau"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-02", "url": "https://arxiv.org/abs/2510.01609", "pdf_url": "https://arxiv.org/pdf/2510.01609v1", "arxiv_id": "2510.01609", "doi": "10.48550/arXiv.2510.01609", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3243}
{"id": "60a62827fa0a67129addee782d9afc6a359ea89239bc29c2a9532ffe83756f89", "sources": ["arxiv", "semantic_scholar"], "title": "StockBench: Can LLM Agents Trade Stocks Profitably In Real-world Markets?", "abstract": "Large language models (LLMs) demonstrate strong potential as autonomous agents, with promising capabilities in reasoning, tool use, and sequential decision-making. While prior benchmarks have evaluated LLM agents in various domains, the financial domain remains underexplored, despite its significant economic value and complex reasoning requirements. Most existing financial benchmarks focus on static question-answering, failing to capture the dynamics of real-market trading. To address this gap, we introduce STOCKBENCH, a contamination-free benchmark designed to evaluate LLM agents in realistic, multi-month stock trading environments. Agents receive daily market signals -- including prices, fundamentals, and news -- and make sequential buy, sell, or hold decisions. Performance is measured using financial metrics such as cumulative return, maximum drawdown, and the Sortino ratio, capturing both profitability and risk management. We evaluate a wide range of state-of-the-art proprietary and open-source LLMs. Surprisingly, most models struggle to outperform the simple buy-and-hold baseline, while some models demonstrate the potential to achieve higher returns and stronger risk management. These findings highlight both the challenges and opportunities of LLM-based trading agents, showing that strong performance on static financial question-answering do not necessarily translate into effective trading behavior. We release STOCKBENCH as an open-source benchmark to enable future research on LLM-driven financial agents.", "authors": ["Yanxu Chen", "Zijun Yao", "Yantao Liu", "Amy Xin", "Jin Ye", "Jianing Yu", "Lei Hou", "Juanzi Li"], "categories": ["cs.LG", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-02", "url": "https://arxiv.org/abs/2510.02209", "pdf_url": "https://arxiv.org/pdf/2510.02209v2", "arxiv_id": "2510.02209", "doi": "10.48550/arXiv.2510.02209", "citation_count": 18, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5011}
{"id": "7c845bb6c4ebb692e13155278454b3bdc49575284072dccd5d86c1383ac673c5", "sources": ["arxiv", "semantic_scholar"], "title": "The Social Laboratory: A Psychometric Framework for Multi-Agent LLM Evaluation", "abstract": "As Large Language Models (LLMs) transition from static tools to autonomous agents, traditional evaluation benchmarks that measure performance on downstream tasks are becoming insufficient. These methods fail to capture the emergent social and cognitive dynamics that arise when agents communicate, persuade, and collaborate in interactive environments. To address this gap, we introduce a novel evaluation framework that uses multi-agent debate as a controlled \"social laboratory\" to discover and quantify these behaviors. In our framework, LLM-based agents, instantiated with distinct personas and incentives, deliberate on a wide range of challenging topics under the supervision of an LLM moderator. Our analysis, enabled by a new suite of psychometric and semantic metrics, reveals several key findings. Across hundreds of debates, we uncover a powerful and robust emergent tendency for agents to seek consensus, consistently reaching high semantic agreement (μ > 0.88) even without explicit instruction and across sensitive topics. We show that assigned personas induce stable, measurable psychometric profiles, particularly in cognitive effort, and that the moderators persona can significantly alter debate outcomes by structuring the environment, a key finding for external AI alignment. This work provides a blueprint for a new class of dynamic, psychometrically grounded evaluation protocols designed for the agentic setting, offering a crucial methodology for understanding and shaping the social behaviors of the next generation of AI agents. We have released the code and results at https://github.com/znreza/multi-agent-LLM-eval-for-debate.", "authors": ["Zarreen Reza"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-01", "url": "https://arxiv.org/abs/2510.01295", "pdf_url": "https://arxiv.org/pdf/2510.01295v1", "arxiv_id": "2510.01295", "doi": "10.48550/arXiv.2510.01295", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/znreza/multi-agent-LLM-eval-for-debate", "venue": "arXiv.org", "quality_score": 0.4994}
{"id": "52abada152635ab19617b2135f4564539009d59eafebabf7e0209c206c7a65d4", "sources": ["arxiv", "semantic_scholar"], "title": "Stochastic Self-Organization in Multi-Agent Systems", "abstract": "Multi-agent systems (MAS) based on Large Language Models (LLMs) have the potential to solve tasks that are beyond the reach of any single LLM. However, this potential can only be realized when the collaboration mechanism between agents is optimized. Specifically, optimizing the communication structure between agents is critical for fruitful collaboration. Most existing approaches rely on fixed topologies, pretrained graph generators, optimization over edges, or employ external LLM judges, thereby adding to the complexity. In this work, we introduce a response-conditioned framework that adapts communication on-the-fly. Agents independently generate responses to the user query and assess peer contributions using an approximation of the Shapley value. A directed acyclic graph (DAG) is then constructed to regulate the propagation of the responses among agents, which ensures stable and efficient message transmission from high-contributing agents to others. This graph is dynamically updated based on the agent responses from the previous collaboration round. Since the proposed framework enables the self-organization of agents without additional supervision or training, we refer to it as SelfOrg. The SelfOrg framework goes beyond task- and query-level optimization and takes into account the stochastic nature of agent responses. Experiments with both strong and weak LLM backends demonstrate robust performance, with significant gains in the weak regime where prior methods collapse. We also theoretically show that multiple agents increase the chance of correctness and that the correct responses naturally dominate the information flow.", "authors": ["Nurbek Tastan", "Samuel Horvath", "Karthik Nandakumar"], "categories": ["cs.MA", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-10-01", "url": "https://arxiv.org/abs/2510.00685", "pdf_url": "https://arxiv.org/pdf/2510.00685v2", "arxiv_id": "2510.00685", "doi": "10.48550/arXiv.2510.00685", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3231}
{"id": "dbda99590d13447006eab8457f4e949dac6712e0ac19b3d8d3d22135a5fb281a", "sources": ["arxiv", "semantic_scholar"], "title": "STAC: When Innocent Tools Form Dangerous Chains to Jailbreak LLM Agents", "abstract": "As LLMs advance into autonomous agents with tool-use capabilities, they introduce security challenges that extend beyond traditional content-based LLM safety concerns. This paper introduces Sequential Tool Attack Chaining (STAC), a novel multi-turn attack framework that exploits agent tool use. STAC chains together tool calls that each appear harmless in isolation but, when combined, collectively enable harmful operations that only become apparent at the final execution step. We apply our framework to automatically generate and systematically evaluate 483 STAC cases, featuring 1,352 sets of user-agent-environment interactions and spanning diverse domains, tasks, agent types, and 10 failure modes. Our evaluations show that state-of-the-art LLM agents, including GPT-4.1, are highly vulnerable to STAC, with attack success rates (ASR) exceeding 90% in most cases. The core design of STAC's automated framework is a closed-loop pipeline that synthesizes executable multi-step tool chains, validates them through in-environment execution, and reverse-engineers stealthy multi-turn prompts that reliably induce agents to execute the verified malicious sequence. We further perform defense analysis against STAC and find that existing prompt-based defenses provide limited protection. To address this gap, we propose a new reasoning-driven defense prompt that achieves far stronger protection, cutting ASR by up to 28.8%. These results highlight a crucial gap: defending tool-enabled agents requires reasoning over entire action sequences and their cumulative effects, rather than evaluating isolated prompts or responses.", "authors": ["Jing-Jing Li", "Jianfeng He", "Chao Shang", "Devang Kulshreshtha", "Xun Xian", "Yi Zhang", "Hang Su", "Sandesh Swamy", "Yanjun Qi"], "categories": ["cs.CR", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2509.25624", "pdf_url": "https://arxiv.org/pdf/2509.25624v2", "arxiv_id": "2509.25624", "doi": "10.48550/arXiv.2509.25624", "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "4c21af79fe07f787af68bc828d42f860eabdb149ad2e89cb338d86a3fcd2f6ba", "sources": ["arxiv", "semantic_scholar"], "title": "TUMIX: Multi-Agent Test-Time Scaling with Tool-Use Mixture", "abstract": "While integrating tools like Code Interpreter and Search has significantly enhanced Large Language Model (LLM) reasoning in models like ChatGPT Agent and Gemini-Pro, practical guidance on optimal tool use is lacking. The core challenge is effectively combining textual reasoning, coding, and search for diverse questions. In this paper, we propose Tool-Use Mixture (TUMIX), an ensemble framework that runs multiple agents in parallel, each employing distinct tool-use strategies and answer paths. Agents in TUMIX iteratively share and refine responses based on the question and previous answers. In experiments, TUMIX achieves significant gains over state-of-the-art tool-augmented and test-time scaling methods, delivering an average accuracy improvement of up to 3.55% over the best baseline on Gemini-2.5-Pro and Gemini-2.5-Flash across key reasoning benchmarks, with near-equal inference costs. We find that agent diversity and quality are crucial and can be enhanced by using LLMs to auto-optimize agent designs. Furthermore, TUMIX can halt refinement upon reaching sufficient confidence, preserving performance at only 49% of the inference cost. Further scaling can achieve higher performance, albeit at a greater cost.", "authors": ["Yongchao Chen", "Jiefeng Chen", "Rui Meng", "Ji Yin", "Na Li", "Chuchu Fan", "Chi Wang", "Tomas Pfister", "Jinsung Yoon"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2510.01279", "pdf_url": "https://arxiv.org/pdf/2510.01279v1", "arxiv_id": "2510.01279", "doi": "10.48550/arXiv.2510.01279", "citation_count": 10, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "b146e2ce562a56565d5aa04c8ea6537b0bad1a76503ba1cd1ac817268176e3df", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Based Multi-Agent Blackboard System for Information Discovery in Data Science", "abstract": "Advances in large language models (LLMs) have created new opportunities in data science, but their deployment is often limited by the challenge of finding relevant data in large data lakes. Existing methods struggle with this: both single- and multi-agent systems are quickly overwhelmed by large, heterogeneous files, and master-slave multi-agent systems rely on a rigid central controller that requires precise knowledge of each sub-agent's capabilities, which is not possible in large-scale settings where the main agent lacks full observability over sub-agents' knowledge and competencies. We propose a novel multi-agent paradigm inspired by the blackboard architecture for traditional AI models. In our framework, a central agent posts requests to a shared blackboard, and autonomous subordinate agents - either responsible for a partition of the data lake or retrieval from the web - volunteer to respond based on their capabilities. This design improves scalability and flexibility by removing the need for a central coordinator to know each agent's expertise or internal knowledge. We evaluate the approach on three benchmarks that require data discovery: KramaBench and modified versions of DSBench and DA-Code. Results show that the blackboard architecture substantially outperforms strong baselines, achieving 13%-57% relative improvements in end-to-end success and up to a 9% relative gain in data discovery F1 over the best baseline.", "authors": ["Alireza Salemi", "Mihir Parmar", "Palash Goyal", "Yiwen Song", "Jinsung Yoon", "Hamed Zamani", "Tomas Pfister", "Hamid Palangi"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.IR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2510.01285", "pdf_url": "https://arxiv.org/pdf/2510.01285v2", "arxiv_id": "2510.01285", "doi": "10.48550/arXiv.2510.01285", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "5fa9ee0da0eec6b58c7ab43103d6cf12bdd99bbc491d15ec910fa0f929801051", "sources": ["arxiv", "semantic_scholar"], "title": "CORTEX: Collaborative LLM Agents for High-Stakes Alert Triage", "abstract": "Security Operations Centers (SOCs) are overwhelmed by tens of thousands of daily alerts, with only a small fraction corresponding to genuine attacks. This overload creates alert fatigue, leading to overlooked threats and analyst burnout. Classical detection pipelines are brittle and context-poor, while recent LLM-based approaches typically rely on a single model to interpret logs, retrieve context, and adjudicate alerts end-to-end -- an approach that struggles with noisy enterprise data and offers limited transparency. We propose CORTEX, a multi-agent LLM architecture for high-stakes alert triage in which specialized agents collaborate over real evidence: a behavior-analysis agent inspects activity sequences, evidence-gathering agents query external systems, and a reasoning agent synthesizes findings into an auditable decision. To support training and evaluation, we release a dataset of fine-grained SOC investigations from production environments, capturing step-by-step analyst actions and linked tool outputs. Across diverse enterprise scenarios, CORTEX substantially reduces false positives and improves investigation quality over state-of-the-art single-agent LLMs.", "authors": ["Bowen Wei", "Yuan Shen Tay", "Howard Liu", "Jinhao Pan", "Kun Luo", "Ziwei Zhu", "Chris Jordan"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2510.00311", "pdf_url": "https://arxiv.org/pdf/2510.00311v1", "arxiv_id": "2510.00311", "doi": "10.48550/arXiv.2510.00311", "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "ae6ed4d62bbfdaa19df45690430de92ce91f5a5135eb3673066e3b2041d18b44", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Reliable Benchmarking: A Contamination Free, Controllable Evaluation Framework for Multi-step LLM Function Calling", "abstract": "Existing benchmarks for tool-augmented language models (TaLMs) lack fine-grained control over task difficulty and remain vulnerable to data contamination. We present FuncBenchGen, a unified, contamination-free framework that evaluates TaLMs by generating synthetic multi-step tool-use tasks to stress-test TaLMs. The key idea is to cast tool use as traversal over a hidden function-dependency DAG where models must infer the correct sequence of calls to compute a target value. FuncBenchGen allows precise control over task difficulty (e.g., graph size, dependency depth, and distractor functions) while avoiding pretraining/test-time leakage. Our evaluation demonstrates reasoning-optimized models consistently outperform general-purpose models with GPT-5 significantly outperforming other available models. Performance declines sharply as dependency depth increases. Furthermore, connected distractors -- irrelevant functions sharing type-compatible variables with relevant functions -- prove especially difficult to handle. Also, strong models often make syntactically valid function calls but propagate incorrect or stale argument values across steps, revealing brittle state tracking by LLMs in multi-turn tool use. Motivated by this observation, we introduce a simple mitigation strategy that explicitly restates prior variable values to the agent at each step. Surprisingly, this lightweight change yields substantial gains across models. e.g., yielding an improvement in success rate from 62.5% to 81.3% for GPT-5.", "authors": ["Seiji Maekawa", "Jackson Hassell", "Pouya Pezeshkpour", "Tom Mitchell", "Estevam Hruschka"], "categories": ["cs.CL", "cs.PL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2509.26553", "pdf_url": "https://arxiv.org/pdf/2509.26553v2", "arxiv_id": "2509.26553", "doi": "10.48550/arXiv.2509.26553", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "105c5c71f0dcc4ce2c5899a38dc92b8b946115aa4e11ae6b877be47d6e813f77", "sources": ["arxiv", "semantic_scholar"], "title": "ScheduleMe: Multi-Agent Calendar Assistant", "abstract": "Recent advancements in LLMs have contributed to the rise of advanced conversational assistants that can assist with user needs through natural language conversation. This paper presents a ScheduleMe, a multi-agent calendar assistant for users to manage google calendar events in natural language. The system uses a graph-structured coordination mechanism where a central supervisory agent supervises specialized task agents, allowing modularity, conflicts resolution, and context-aware interactions to resolve ambiguities and evaluate user commands. This approach sets an example of how structured reasoning and agent cooperation might convince operators to increase the usability and flexibility of personal calendar assistant tools.", "authors": ["Oshadha Wijerathne", "Amandi Nimasha", "Dushan Fernando", "Nisansa de Silva", "Srinath Perera"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2509.25693", "pdf_url": "https://arxiv.org/pdf/2509.25693v3", "arxiv_id": "2509.25693", "doi": "10.48550/arXiv.2509.25693", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "4306bb44664fdd5346101ee20d860b7a256956c4aa1ce8a437ad4b4cb8311d29", "sources": ["arxiv", "semantic_scholar"], "title": "Interactive Learning for LLM Reasoning", "abstract": "Existing multi-agent learning approaches have developed interactive training environments to explicitly promote collaboration among multiple Large Language Models (LLMs), thereby constructing stronger multi-agent systems (MAS). However, during inference, they require re-executing the MAS to obtain final solutions, which diverges from human cognition that individuals can enhance their reasoning capabilities through interactions with others and resolve questions independently in the future. To investigate whether multi-agent interaction can enhance LLMs' independent problem-solving ability, we introduce ILR, a novel co-learning framework for MAS that integrates two key components: Dynamic Interaction and Perception Calibration. Specifically, Dynamic Interaction first adaptively selects either cooperative or competitive strategies depending on question difficulty and model ability. LLMs then exchange information through Idea3, an innovative interaction paradigm designed to mimic human discussion, before deriving their respective final answers. In Perception Calibration, ILR employs Group Relative Policy Optimization (GRPO) to train LLMs while integrating one LLM's reward distribution characteristics into another's reward function, thereby enhancing the cohesion of multi-agent interactions. We evaluate the effectiveness of ILR across three LLMs from two model families of varying scales on five mathematical, one coding, one general question answering, and one scientific reasoning benchmarks. Experimental results show that ILR consistently outperforms single-agent learning, yielding an improvement of up to 5% over the strongest baseline. We further discover that Idea3 can enhance the robustness of stronger LLMs during multi-agent inference, and dynamic interaction types can boost multi-agent learning compared to pure cooperative or competitive strategies.", "authors": ["Hehai Lin", "Shilei Cao", "Sudong Wang", "Haotian Wu", "Minzhi Li", "Linyi Yang", "Juepeng Zheng", "Chengwei Qin"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2509.26306", "pdf_url": "https://arxiv.org/pdf/2509.26306v4", "arxiv_id": "2509.26306", "doi": "10.48550/arXiv.2509.26306", "citation_count": 5, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/linhh29/Interactive-Learning-for-LLM-Reasoning", "venue": "arXiv.org", "quality_score": 0.4976}
{"id": "21a0675a287c2426c3a0556baf06618dc4ba426402ab4410f95000f54e22b61c", "sources": ["arxiv", "semantic_scholar"], "title": "MAGIC-MASK: Multi-Agent Guided Inter-Agent Collaboration with Mask-Based Explainability for Reinforcement Learning", "abstract": "Understanding the decision-making process of Deep Reinforcement Learning agents remains a key challenge for deploying these systems in safety-critical and multi-agent environments. While prior explainability methods like StateMask, have advanced the identification of critical states, they remain limited by computational cost, exploration coverage, and lack of adaptation to multi-agent settings. To overcome these limitations, we propose a mathematically grounded framework, MAGIC-MASK (Multi-Agent Guided Inter-agent Collaboration with Mask-Based Explainability for Reinforcement Learning), that extends perturbation-based explanation to Multi-Agent Reinforcement Learning. Our method integrates Proximal Policy Optimization, adaptive epsilon-greedy exploration, and lightweight inter-agent collaboration to share masked state information and peer experience. This collaboration enables each agent to perform saliency-guided masking and share reward-based insights with peers, reducing the time required for critical state discovery, improving explanation fidelity, and leading to faster and more robust learning. The core novelty of our approach lies in generalizing explainability from single-agent to multi-agent systems through a unified mathematical formalism built on trajectory perturbation, reward fidelity analysis, and Kullback-Leibler divergence regularization. This framework yields localized, interpretable explanations grounded in probabilistic modeling and multi-agent Markov decision processes. We validate our framework on both single-agent and multi-agent benchmarks, including a multi-agent highway driving environment and Google Research Football, demonstrating that MAGIC-MASK consistently outperforms state-of-the-art baselines in fidelity, learning efficiency, and policy robustness while offering interpretable and transferable explanations.", "authors": ["Maisha Maliha", "Dean Hougen"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-30", "url": "https://arxiv.org/abs/2510.00274", "pdf_url": "https://arxiv.org/pdf/2510.00274v1", "arxiv_id": "2510.00274", "doi": "10.48550/arXiv.2510.00274", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.322}
{"id": "91fb720c4dd6c785f5478c4a989f5a056dbfb487402aa8eba7918e8dccdc24a8", "sources": ["arxiv", "semantic_scholar"], "title": "\"Stop replacing salt with sugar!'': Towards Intuitive Human-Agent Teaching", "abstract": "Humans quickly learn new concepts from a small number of examples. Replicating this capacity with Artificial Intelligence (AI) systems has proven to be challenging. When it comes to learning subjective tasks-where there is an evident scarcity of data-this capacity needs to be recreated. In this work, we propose an intuitive human-agent teaching architecture in which the human can teach an agent how to perform a task by providing demonstrations, i.e., examples. To have an intuitive interaction, we argue that the agent should be able to learn incrementally from a few single examples. To allow for this, our objective is to broaden the agent's task understanding using domain knowledge. Then, using a learning method to enable the agent to learn efficiently from a limited number of examples. Finally, to optimize how human can select the most representative and less redundant examples to provide the agent with. We apply our proposed method to the subjective task of ingredient substitution, where the agent needs to learn how to substitute ingredients in recipes based on human examples. We replicate human input using the Recipe1MSubs dataset. In our experiments, the agent achieves half its task performance after only 100 examples are provided, compared to the complete training set of 50k examples. We show that by providing examples in strategic order along with a learning method that leverages external symbolic knowledge, the agent can generalize more efficiently.", "authors": ["Nikolaos Kondylidis", "Andrea Rafanelli", "Ilaria Tiddi", "Annette ten Teije", "Frank van Harmelen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.24651", "pdf_url": "https://arxiv.org/pdf/2509.24651v1", "arxiv_id": "2509.24651", "doi": "10.48550/arXiv.2509.24651", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3208}
{"id": "3fbf76a85674160948d113c17bce9da447c470092c55cb9624ccca1877f559ca", "sources": ["arxiv", "semantic_scholar"], "title": "FuncPoison: Poisoning Function Library to Hijack Multi-agent Autonomous Driving Systems", "abstract": "Autonomous driving systems increasingly rely on multi-agent architectures powered by large language models (LLMs), where specialized agents collaborate to perceive, reason, and plan. A key component of these systems is the shared function library, a collection of software tools that agents use to process sensor data and navigate complex driving environments. Despite its critical role in agent decision-making, the function library remains an under-explored vulnerability. In this paper, we introduce FuncPoison, a novel poisoning-based attack targeting the function library to manipulate the behavior of LLM-driven multi-agent autonomous systems. FuncPoison exploits two key weaknesses in how agents access the function library: (1) agents rely on text-based instructions to select tools; and (2) these tools are activated using standardized command formats that attackers can replicate. By injecting malicious tools with deceptive instructions, FuncPoison manipulates one agent s decisions--such as misinterpreting road conditions--triggering cascading errors that mislead other agents in the system. We experimentally evaluate FuncPoison on two representative multi-agent autonomous driving systems, demonstrating its ability to significantly degrade trajectory accuracy, flexibly target specific agents to induce coordinated misbehavior, and evade diverse defense mechanisms. Our results reveal that the function library, often considered a simple toolset, can serve as a critical attack surface in LLM-based autonomous driving systems, raising elevated concerns on their reliability.", "authors": ["Yuzhen Long", "Songze Li"], "categories": ["cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.24408", "pdf_url": "https://arxiv.org/pdf/2509.24408v2", "arxiv_id": "2509.24408", "doi": "10.48550/arXiv.2509.24408", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3208}
{"id": "110dbe456fb5126a5e5f5d735e27083f3619acc45c86bec161e218602f8b7f9a", "sources": ["arxiv", "semantic_scholar"], "title": "AIPOM: Agent-aware Interactive Planning for Multi-Agent Systems", "abstract": "Large language models (LLMs) are being increasingly used for planning in orchestrated multi-agent systems. However, existing LLM-based approaches often fall short of human expectations and, critically, lack effective mechanisms for users to inspect, understand, and control their behaviors. These limitations call for enhanced transparency, controllability, and human oversight. To address this, we introduce AIPOM, a system supporting human-in-the-loop planning through conversational and graph-based interfaces. AIPOM enables users to transparently inspect, refine, and collaboratively guide LLM-generated plans, significantly enhancing user control and trust in multi-agent workflows. Our code and demo video are available at https://github.com/megagonlabs/aipom.", "authors": ["Hannah Kim", "Kushan Mitra", "Chen Shen", "Dan Zhang", "Estevam Hruschka"], "categories": ["cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.24826", "pdf_url": "https://arxiv.org/pdf/2509.24826v1", "arxiv_id": "2509.24826", "doi": "10.48550/arXiv.2509.24826", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/megagonlabs/aipom", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.4958}
{"id": "fb778ab7ac832da286f5fcaaa2c832aac9240f95426d4070d5f93c9809858d5b", "sources": ["arxiv", "semantic_scholar"], "title": "ATLAS: Constraints-Aware Multi-Agent Collaboration for Real-World Travel Planning", "abstract": "While Large Language Models (LLMs) have shown remarkable advancements in reasoning and tool use, they often fail to generate optimal, grounded solutions under complex constraints. Real-world travel planning exemplifies these challenges, evaluating agents' abilities to handle constraints that are explicit, implicit, and even evolving based on interactions with dynamic environments and user needs. In this paper, we present ATLAS, a general multi-agent framework designed to effectively handle such complex nature of constraints awareness in real-world travel planning tasks. ATLAS introduces a principled approach to address the fundamental challenges of constraint-aware planning through dedicated mechanisms for dynamic constraint management, iterative plan critique, and adaptive interleaved search. ATLAS demonstrates state-of-the-art performance on the TravelPlanner benchmark, improving the final pass rate from 23.3% to 44.4% over its best alternative. More importantly, our work is the first to demonstrate quantitative effectiveness on real-world travel planning tasks with live information search and multi-turn feedback. In this realistic setting, ATLAS showcases its superior overall planning performance, achieving an 84% final pass rate which significantly outperforms baselines including ReAct (59%) and a monolithic agent (27%).", "authors": ["Jihye Choi", "Jinsung Yoon", "Jiefeng Chen", "Somesh Jha", "Tomas Pfister"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.25586", "pdf_url": "https://arxiv.org/pdf/2509.25586v1", "arxiv_id": "2509.25586", "doi": "10.48550/arXiv.2509.25586", "citation_count": 9, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3208}
{"id": "09a2d27980829804d2a68099f171c9d23448a46ca95006a28f7d18d4413e9d81", "sources": ["arxiv", "semantic_scholar"], "title": "Unifying Agent Interaction and World Information for Multi-agent Coordination", "abstract": "This work presents a novel representation learning framework, *interaction-world* latent (IWoL), to facilitate *team coordination* in multi-agent reinforcement learning (MARL). Building effective representation for team coordination is a challenging problem, due to the intricate dynamics emerging from multi-agent interaction and incomplete information induced by local observations. Our key insight is to construct a learnable representation space that jointly captures inter-agent relations and task-specific world information by directly modeling communication protocols. This representation enables fully decentralized execution with implicit coordination while avoiding the drawbacks of explicit message passing, for example, slower decision-making, vulnerability to malicious attackers, and sensitivity to bandwidth limitations. In practice, our representation can be used not only as an implicit latent for each agent, but also as an explicit message for communication. Across four challenging MARL benchmarks, we evaluate both variants and show that IWoL provides a simple yet powerful key for team coordination. Moreover, we demonstrate that our representation can be combined with existing MARL algorithms to further enhance their performance.", "authors": ["Dongsu Lee", "Daehee Lee", "Yaru Niu", "Honguk Woo", "Amy Zhang", "Ding Zhao"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.25550", "pdf_url": "https://arxiv.org/pdf/2509.25550v4", "arxiv_id": "2509.25550", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2042}
{"id": "fd8201668939342d5c6fc9c7a68c5de8aaf3102ec63827e51ae8a7445243f03e", "sources": ["arxiv", "semantic_scholar"], "title": "MAS$^2$: Self-Generative, Self-Configuring, Self-Rectifying Multi-Agent Systems", "abstract": "The past two years have witnessed the meteoric rise of Large Language Model (LLM)-powered multi-agent systems (MAS), which harness collective intelligence and exhibit a remarkable trajectory toward self-evolution. This paradigm has rapidly progressed from manually engineered systems that require bespoke configuration of prompts, tools, roles, and communication protocols toward frameworks capable of automated orchestration. Yet, dominant automatic multi-agent systems, whether generated by external modules or a single LLM agent, largely adhere to a rigid ``\\textit{generate-once-and-deploy}'' paradigm, rendering the resulting systems brittle and ill-prepared for the dynamism and uncertainty of real-world environments. To transcend this limitation, we introduce MAS$^2$, a paradigm predicated on the principle of recursive self-generation: a multi-agent system that autonomously architects bespoke multi-agent systems for diverse problems. Technically, we devise a ``\\textit{generator-implementer-rectifier}'' tri-agent team capable of dynamically composing and adaptively rectifying a target agent system in response to real-time task demands. Collaborative Tree Optimization is proposed to train and specialize these meta-agents. Extensive evaluation across seven benchmarks reveals that MAS$^2$ achieves performance gains of up to $19.6\\%$ over state-of-the-art MAS in complex scenarios such as deep research and code generation. Moreover, MAS$^2$ exhibits superior cross-backbone generalization, effectively leveraging previously unseen LLMs to yield improvements of up to $15.1\\%$. Crucially, these gains are attained without incurring excessive token costs, as MAS$^2$ consistently resides on the Pareto frontier of cost-performance trade-offs. The source codes are available at https://github.com/yeyeyeah2/MAS2.", "authors": ["Kun Wang", "Guibin Zhang", "ManKit Ye", "Xinyu Deng", "Dongxia Wang", "Xiaobin Hu", "Jinyang Guo", "Yang Liu", "Yufei Guo"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-29", "url": "https://arxiv.org/abs/2509.24323", "pdf_url": "https://arxiv.org/pdf/2509.24323v1", "arxiv_id": "2509.24323", "doi": "10.48550/arXiv.2509.24323", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/yeyeyeah2/MAS2", "venue": "arXiv.org", "quality_score": 0.4958}
{"id": "d395a2553c5734124cf61d2e7c0cdda1103a1d2f9e66a8bf472fef6944165daa", "sources": ["arxiv", "semantic_scholar"], "title": "PartnerMAS: An LLM Hierarchical Multi-Agent Framework for Business Partner Selection on High-Dimensional Features", "abstract": "High-dimensional decision-making tasks, such as business partner selection, involve evaluating large candidate pools with heterogeneous numerical, categorical, and textual features. While large language models (LLMs) offer strong in-context reasoning capabilities, single-agent or debate-style systems often struggle with scalability and consistency in such settings. We propose PartnerMAS, a hierarchical multi-agent framework that decomposes evaluation into three layers: a Planner Agent that designs strategies, Specialized Agents that perform role-specific assessments, and a Supervisor Agent that integrates their outputs. To support systematic evaluation, we also introduce a curated benchmark dataset of venture capital co-investments, featuring diverse firm attributes and ground-truth syndicates. Across 140 cases, PartnerMAS consistently outperforms single-agent and debate-based multi-agent baselines, achieving up to 10--15\\% higher match rates. Analysis of agent reasoning shows that planners are most responsive to domain-informed prompts, specialists produce complementary feature coverage, and supervisors play an important role in aggregation. Our findings demonstrate that structured collaboration among LLM agents can generate more robust outcomes than scaling individual models, highlighting PartnerMAS as a promising framework for high-dimensional decision-making in data-rich domains.", "authors": ["Lingyao Li", "Haolun Wu", "Zhenkun Li", "Jiabei Hu", "Yu Wang", "Xiaoshan Huang", "Wenyue Hua", "Wenqian Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-28", "url": "https://arxiv.org/abs/2509.24046", "pdf_url": "https://arxiv.org/pdf/2509.24046v2", "arxiv_id": "2509.24046", "doi": "10.48550/arXiv.2509.24046", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "ef0ceb9d7c17228ba3d83da8c002a1aad9b2e757fb71031848eb692920b0b28c", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond the Strongest LLM: Multi-Turn Multi-Agent Orchestration vs. Single LLMs on Benchmarks", "abstract": "We study multi-turn multi-agent orchestration, where multiple large language model (LLM) agents interact over multiple turns by iteratively proposing answers or casting votes until reaching consensus. Using four LLMs (Gemini 2.5 Pro, GPT-5, Grok 4, and Claude Sonnet 4) on GPQA-Diamond, IFEval, and MuSR, we conduct two experiments: (i) benchmarking orchestration against single-LLM baselines; and (ii) ablations on GPQA-Diamond that vary whether agents see who authored answers and whether they can observe ongoing votes. Orchestration matches or exceeds the strongest single model and consistently outperforms the others. Analysis of best-achievable orchestration performance shows potential for further gains. The ablations show that revealing authorship increases self-voting and ties, and that showing ongoing votes amplifies herding, which speeds convergence but can sometimes yield premature consensus.", "authors": ["Aaron Xuxiang Tian", "Ruofan Zhang", "Jiayao Tang", "Young Min Cho", "Xueqian Li", "Qiang Yi", "Ji Wang", "Zhunping Zhang", "Danrui Qi", "Zekun Li", "Xingyu Xiang", "Sharath Chandra Guntuku", "Lyle Ungar", "Tianyu Shi", "Chi Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-28", "url": "https://arxiv.org/abs/2509.23537", "pdf_url": "https://arxiv.org/pdf/2509.23537v2", "arxiv_id": "2509.23537", "doi": "10.48550/arXiv.2509.23537", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "3be1c51f61445a0a8f069e2b346b6e82b47a8c974ece8c0523442eb61e7f3ea0", "sources": ["arxiv", "semantic_scholar"], "title": "Reducing Cost of LLM Agents with Trajectory Reduction", "abstract": "Multi-turn agent systems based on Large Language Models (LLMs) have become increasingly popular for software engineering tasks. While LLM agents demonstrate promising effectiveness, the high computational cost of input tokens due to ever-growing trajectories remains a significant efficiency concern. Efficiency has been largely overlooked in existing studies and agent products, and this paper addresses this gap by introducing an inference-time trajectory reduction approach that reduces computational costs. By analyzing existing agent trajectories, we demonstrate that useless, redundant, and expired information is widespread across trajectories. Such waste can be identified and reduced without compromising the agent's performance. We propose a simple yet effective trajectory reduction approach, AgentDiet, which automatically removes such waste during agent execution. We implement AgentDiet on a top-performing coding agent, and our evaluation on two LLMs and two benchmarks shows that AgentDiet can reduce input tokens by 39.9%-59.7% and the total computational cost by 21.1%-35.9%, while maintaining the same agent performance. These results indicate that inference-time trajectory reduction is a promising direction for agent systems.", "authors": ["Yuan-An Xiao", "Pengfei Gao", "Chao Peng", "Yingfei Xiong"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-28", "url": "https://arxiv.org/abs/2509.23586", "pdf_url": "https://arxiv.org/pdf/2509.23586v2", "arxiv_id": "2509.23586", "doi": "10.1145/3797084", "citation_count": 18, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3197}
{"id": "ea157e7957f97a236ee811cf87d4cf60caf5db557bfbbc1d2ce5e8ade4120d78", "sources": ["arxiv", "semantic_scholar"], "title": "Non-Collaborative User Simulators for Tool Agents", "abstract": "Tool agents interact with users through multi-turn dialogues to accomplish various tasks. Recent studies have adopted user simulation methods to develop these agents in multi-turn settings. However, existing user simulators tend to be agent-friendly, exhibiting only cooperative behaviors, failing to train and test agents against non-collaborative users in the real world. We propose a novel user simulator architecture that simulates four categories of non-collaborative behaviors: requesting unavailable services, digressing into tangential conversations, expressing impatience, and providing incomplete utterances. Our user simulator can simulate challenging and natural non-collaborative behaviors while reliably delivering all intents and information necessary to accomplish the task. Our experiments on MultiWOZ and τ-bench reveal significant performance degradation in state-of-the-art tool agents when encountering non-collaborative users, as well as agent weaknesses under each non-collaborative condition such as escalated hallucinations and dialogue breakdowns. Our findings point to the need for methods that can improve agent robustness to the wide range of user behaviors encountered in deployment. We release the extensible simulation framework to help the community develop and stress-test tool agents under realistic conditions within their own service domains. Our code is available at https://github.com/holi-lab/NCUser.", "authors": ["Jeonghoon Shim", "Woojung Song", "Cheyon Jin", "Seungwon KooK", "Yohan Jo"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-27", "url": "https://arxiv.org/abs/2509.23124", "pdf_url": "https://arxiv.org/pdf/2509.23124v5", "arxiv_id": "2509.23124", "doi": "10.48550/arXiv.2509.23124", "citation_count": 8, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/holi-lab/NCUser", "venue": "arXiv.org", "quality_score": 0.4923}
{"id": "da47377c5fa654e294b5e49e41eb5207546d5bf40c6a402497c6c517b9dda096", "sources": ["arxiv", "semantic_scholar"], "title": "Peacemaker or Troublemaker: How Sycophancy Shapes Multi-Agent Debate", "abstract": "Large language models (LLMs) often display sycophancy, a tendency toward excessive agreeability. This behavior poses significant challenges for multi-agent debating systems (MADS) that rely on productive disagreement to refine arguments and foster innovative thinking. LLMs' inherent sycophancy can collapse debates into premature consensus, potentially undermining the benefits of multi-agent debate. While prior studies focus on user--LLM sycophancy, the impact of inter-agent sycophancy in debate remains poorly understood. To address this gap, we introduce the first operational framework that (1) proposes a formal definition of sycophancy specific to MADS settings, (2) develops new metrics to evaluate the agent sycophancy level and its impact on information exchange in MADS, and (3) systematically investigates how varying levels of sycophancy across agent roles (debaters and judges) affects outcomes in both decentralized and centralized debate frameworks. Our findings reveal that sycophancy is a core failure mode that amplifies disagreement collapse before reaching a correct conclusion in multi-agent debates, yields lower accuracy than single-agent baselines, and arises from distinct debater-driven and judge-driven failure modes. Building on these findings, we propose actionable design principles for MADS, effectively balancing productive disagreement with cooperation in agent interactions.", "authors": ["Binwei Yao", "Chao Shang", "Wanyu Du", "Jianfeng He", "Ruixue Lian", "Yi Zhang", "Hang Su", "Sandesh Swamy", "Yanjun Qi"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-27", "url": "https://arxiv.org/abs/2509.23055", "pdf_url": "https://arxiv.org/pdf/2509.23055v1", "arxiv_id": "2509.23055", "doi": "10.48550/arXiv.2509.23055", "citation_count": 16, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3185}
{"id": "02898cfc911c77c08e55cac7cb023be56a85ba8182ff4a17f1678327d0649a36", "sources": ["arxiv", "semantic_scholar"], "title": "Think-on-Graph 3.0: Efficient and Adaptive LLM Reasoning on Heterogeneous Graphs via Multi-Agent Dual-Evolving Context Retrieval", "abstract": "Graph-based Retrieval-Augmented Generation (GraphRAG) has become the important paradigm for enhancing Large Language Models (LLMs) with external knowledge. However, existing approaches are constrained by their reliance on high-quality knowledge graphs: manually built ones are not scalable, while automatically extracted ones are limited by the performance of LLM extractors, especially when using smaller, local-deployed models. To address this, we introduce Think-on-Graph 3.0 (ToG-3), a novel framework featuring a Multi-Agent Context Evolution and Retrieval (MACER) mechanism. Its core contribution is the dynamic construction and iterative refinement of a Chunk-Triplets-Community heterogeneous graph index, powered by a Dual-Evolution process that adaptively evolves both the query and the retrieved sub-graph during reasoning. ToG-3 dynamically builds a targeted graph index tailored to the query, enabling precise evidence retrieval and reasoning even with lightweight LLMs. Extensive experiments demonstrate that ToG-3 outperforms compared baselines on both deep and broad reasoning benchmarks, and ablation studies confirm the efficacy of the components of MACER framework. The source code are available in https://github.com/DataArcTech/ToG-3.", "authors": ["Xiaojun Wu", "Cehao Yang", "Xueyuan Lin", "Chengjin Xu", "Xuhui Jiang", "Yuanliang Sun", "Hui Xiong", "Jia Li", "Jian Guo"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-26", "url": "https://arxiv.org/abs/2509.21710", "pdf_url": "https://arxiv.org/pdf/2509.21710v2", "arxiv_id": "2509.21710", "doi": "10.48550/arXiv.2509.21710", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/DataArcTech/ToG-3", "venue": "arXiv.org", "quality_score": 0.4905}
{"id": "8985256eebc8fa8ffed03b7f19df02dbf541929f69705ece389424bded7d610d", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Belief Reasoning with LLMs for Efficient Multi-Agent Collaboration", "abstract": "Effective real-world multi-agent collaboration requires not only accurate planning but also the ability to reason about collaborators' intents--a crucial capability for avoiding miscoordination and redundant communication under partial observable environments. Due to their strong planning and reasoning capabilities, large language models (LLMs) have emerged as promising autonomous agents for collaborative task solving. However, existing collaboration frameworks for LLMs overlook their reasoning potential for dynamic intent inference, and thus produce inconsistent plans and redundant communication, reducing collaboration efficiency. To bridge this gap, we propose CoBel-World, a novel framework that equips LLM agents with a Collaborative Belief World--an internal representation jointly modeling the physical environment and collaborators' mental states. CoBel-World enables agents to parse external open-world knowledge into structured beliefs via a symbolic belief representation module, and perform zero-shot Bayesian-style belief updates through LLM reasoning. This allows agents to proactively detect potential miscoordination (e.g., conflicting plans) and communicate adaptively. Evaluated on challenging embodied benchmarks (i.e., TDW-MAT and C-WAH), CoBel-World significantly reduces communication costs by 64-79% and improves task completion efficiency by 4-28% compared to the strongest baseline. Our results show that explicit, intent-aware belief modeling is essential for efficient and human-like collaboration in LLM-based multi-agent systems.", "authors": ["Zhimin Wang", "Duo Wu", "Shaokang He", "Jinghe Wang", "Linjia Kang", "Jing Yu", "Kai Zhu", "Jiawei Li", "Zhi Wang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-26", "url": "https://arxiv.org/abs/2509.21981", "pdf_url": "https://arxiv.org/pdf/2509.21981v3", "arxiv_id": "2509.21981", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.202}
{"id": "6b69f4c74fb42ec7b3f4820a1b1db3f2af0d0624a126343c1d0644e48221eaa8", "sources": ["arxiv", "semantic_scholar"], "title": "VibeCodeHPC: An Agent-Based Iterative Prompting Auto-Tuner for HPC Code Generation Using LLMs", "abstract": "In this study, we propose VibeCodeHPC, a multi-agent system based on large language models (LLMs) for the automatic tuning of high-performance computing (HPC) programs on supercomputers. VibeCodeHPC adopts Claude Code as its backend and provides an integrated environment that facilitates program development in supercomputer settings. The system not only brings the Vibe Coding paradigm -- program development through natural language interaction with users -- to HPC programming, but also enables autonomous performance optimization with minimal user intervention through a sophisticated multi-agent design. To achieve these objectives, VibeCodeHPC implements three core functionalities: (1) configuration capabilities tailored to the unique development environments of supercomputers, (2) collaborative operation among multiple LLM agents with distinct roles -- Project Manager (PM), System Engineer (SE), Programmer (PG), and Continuous Deliverer (CD), and (3) long-term autonomous operation through agent activity monitoring and dynamic deployment mechanisms. This paper highlights one of the most powerful features of VibeCodeHPC: fully automated code optimization through autonomous operation without user intervention. Specifically, it demonstrates the performance optimization of CPU-based codes on GPU-equipped systems for matrix multiplication and a Poisson equation solver using Jacobi's iterative method. The results show that the multi-agent configuration employed in VibeCodeHPC enables faster and more reliable development of higher-performance code compared to a single-agent setup.", "authors": ["Shun-ichiro Hayashi", "Koki Morita", "Daichi Mukunoki", "Tetsuya Hoshino", "Takahiro Katagiri"], "categories": ["cs.SE", "cs.AI", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-26", "url": "https://arxiv.org/abs/2510.00031", "pdf_url": "https://arxiv.org/pdf/2510.00031v3", "arxiv_id": "2510.00031", "doi": "10.48550/arXiv.2510.00031", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3174}
{"id": "cef8770f6f2c3c1359e3a76f6f5804aee0a644730ef0b56188edb1b8fec86e9c", "sources": ["arxiv", "semantic_scholar"], "title": "InfiAgent: Self-Evolving Pyramid Agent Framework for Infinite Scenarios", "abstract": "Large Language Model (LLM) agents have demonstrated remarkable capabilities in organizing and executing complex tasks, and many such agents are now widely used in various application scenarios. However, developing these agents requires carefully designed workflows, carefully crafted prompts, and iterative tuning, which requires LLM techniques and domain-specific expertise. These hand-crafted limitations hinder the scalability and cost-effectiveness of LLM agents across a wide range of industries. To address these challenges, we propose \\textbf{InfiAgent}, a Pyramid-like DAG-based Multi-Agent Framework that can be applied to \\textbf{infi}nite scenarios, which introduces several key innovations: a generalized \"agent-as-a-tool\" mechanism that automatically decomposes complex agents into hierarchical multi-agent systems; a dual-audit mechanism that ensures the quality and stability of task completion; an agent routing function that enables efficient task-agent matching; and an agent self-evolution mechanism that autonomously restructures the agent DAG based on new tasks, poor performance, or optimization opportunities. Furthermore, InfiAgent's atomic task design supports agent parallelism, significantly improving execution efficiency. This framework evolves into a versatile pyramid-like multi-agent system capable of solving a wide range of problems. Evaluations on multiple benchmarks demonstrate that InfiAgent achieves 9.9\\% higher performance compared to ADAS (similar auto-generated agent framework), while a case study of the AI research assistant InfiHelper shows that it generates scientific papers that have received recognition from human reviewers at top-tier IEEE conferences.", "authors": ["Chenglin Yu", "Yang Yu", "Songmiao Wang", "Yucheng Wang", "Yifan Yang", "Jinjia Li", "Ming Li", "Hongxia Yang"], "categories": ["cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-26", "url": "https://arxiv.org/abs/2509.22502", "pdf_url": "https://arxiv.org/pdf/2509.22502v2", "arxiv_id": "2509.22502", "doi": "10.48550/arXiv.2509.22502", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3174}
{"id": "f65d60a0af25322cbae2480b9e837e9a12c4fcc1fcbb179b35422ca64b6d8f2a", "sources": ["arxiv", "semantic_scholar"], "title": "Tool Calling for Arabic LLMs: Data Strategies and Instruction Tuning", "abstract": "Tool calling is a critical capability that allows Large Language Models (LLMs) to interact with external systems, significantly expanding their utility. However, research and resources for tool calling are predominantly English-centric, leaving a gap in our understanding of how to enable this functionality for other languages, such as Arabic. This paper investigates three key research questions: (1) the necessity of in-language (Arabic) tool-calling data versus relying on cross-lingual transfer, (2) the effect of general-purpose instruction tuning on tool-calling performance, and (3) the value of fine-tuning on specific, high-priority tools. To address these questions, we conduct extensive experiments using base and post-trained variants of an open-weight Arabic LLM. To enable this study, we bridge the resource gap by translating and adapting two open-source tool-calling datasets into Arabic. Our findings provide crucial insights into the optimal strategies for developing robust tool-augmented agents for Arabic.", "authors": ["Asim Ersoy", "Enes Altinisik", "Husrev Taha Sencar", "Kareem Darwish"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-25", "url": "https://arxiv.org/abs/2509.20957", "pdf_url": "https://arxiv.org/pdf/2509.20957v1", "arxiv_id": "2509.20957", "doi": "10.48550/arXiv.2509.20957", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "ArabicNLP 2025", "quality_score": 0.4887}
{"id": "b75a482168aef2d9ecc3e1219634368a3b4923fcb6b8365ab405db844e3b6047", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agent Meets Agentic AI: Can LLM Agents Simulate Customers to Evaluate Agentic-AI-based Shopping Assistants?", "abstract": "Agentic AI is emerging, capable of executing tasks through natural language, such as Copilot for coding or Amazon Rufus for shopping. Evaluating these systems is challenging, as their rapid evolution outpaces traditional human evaluation. Researchers have proposed LLM Agents to simulate participants as digital twins, but it remains unclear to what extent a digital twin can represent a specific customer in multi-turn interaction with an agentic AI system. In this paper, we recruited 40 human participants to shop with Amazon Rufus, collected their personas, interaction traces, and UX feedback, and then created digital twins to repeat the task. Pairwise comparison of human and digital-twin traces shows that while agents often explored more diverse choices, their action patterns aligned with humans and yielded similar design feedback. This study is the first to quantify how closely LLM agents can mirror human multi-turn interaction with an agentic AI system, highlighting their potential for scalable evaluation.", "authors": ["Lu Sun", "Shihan Fu", "Bingsheng Yao", "Yuxuan Lu", "Wenbo Li", "Hansu Gu", "Jiri Gesi", "Jing Huang", "Chen Luo", "Dakuo Wang"], "categories": ["cs.HC", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-25", "url": "https://arxiv.org/abs/2509.21501", "pdf_url": "https://arxiv.org/pdf/2509.21501v1", "arxiv_id": "2509.21501", "doi": "10.48550/arXiv.2509.21501", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3162}
{"id": "5f6dd194d47e3f760b5dbc0a090d9cb07a6849d8a2e6ae166a78ae21b05bac28", "sources": ["arxiv", "semantic_scholar"], "title": "Recon-Act: A Self-Evolving Multi-Agent Browser-Use System via Web Reconnaissance, Tool Generation, and Task Execution", "abstract": "Recent years, multimodal models have made remarkable strides and pave the way for intelligent browser use agents. However, when solving tasks on real world webpages in multi-turn, long-horizon trajectories, current agents still suffer from disordered action sequencing and excessive trial and error during execution. This paper introduces Recon-Act, a self-evolving multi-agent framework grounded in Reconnaissance-Action behavioral paradigm. The system comprises a Reconnaissance Team and an Action Team: the former conducts comparative analysis and tool generation, while the latter handles intent decomposition, tool orchestration, and execution. By contrasting the erroneous trajectories with successful ones, the Reconnaissance Team infers remedies, and abstracts them into a unified notion of generalized tools, either expressed as hints or as rule-based codes, and register to the tool archive in real time. The Action Team reinference the process empowered with these targeting tools, thus establishing a closed-loop training pipeline of data-tools-action-feedback. Following the 6 level implementation roadmap proposed in this work, we have currently reached Level 3 (with limited human-in-the-loop intervention). Leveraging generalized tools obtained through reconnaissance, Recon-Act substantially improves adaptability to unseen websites and solvability on long-horizon tasks, and achieves state-of-the-art performance on the challenging VisualWebArena dataset.", "authors": ["Kaiwen He", "Zhiwei Wang", "Chenyi Zhuang", "Jinjie Gu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-25", "url": "https://arxiv.org/abs/2509.21072", "pdf_url": "https://arxiv.org/pdf/2509.21072v1", "arxiv_id": "2509.21072", "doi": "10.48550/arXiv.2509.21072", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3162}
{"id": "e7a4d1c1669d902c0d6304292358c8f1fb15c25921fcc4d44d7516cbd113e1a3", "sources": ["arxiv", "semantic_scholar"], "title": "Tree Search for LLM Agent Reinforcement Learning", "abstract": "Recent advances in reinforcement learning (RL) have significantly enhanced the agentic capabilities of large language models (LLMs). In long-term and multi-turn agent tasks, existing approaches driven solely by outcome rewards often suffer from the problem of sparse supervision. To address the challenge, we propose Tree-based Group Relative Policy Optimization (Tree-GRPO), a grouped agent RL method based on tree search, where each tree node represents the complete agent interaction step. By sharing common prefixes, the tree search sampling increases the number of rollouts achievable within a fixed budget of tokens or tool calls. Moreover, we find that the tree-structured trajectory naturally allows the construction of step-wise process supervised signals even using only the outcome reward. Based on this, Tree-GRPO estimates the grouped relative advantages both on intra-tree and inter-tree levels. Through theoretical analysis, we demonstrate that the objective of intra-tree level group relative policy optimization is equivalent to that of step-level direct preference learning. Experiments across 11 datasets and 3 types of QA tasks demonstrate the superiority of the proposed tree-based RL over the chain-based RL method.", "authors": ["Yuxiang Ji", "Ziyu Ma", "Yong Wang", "Guanhua Chen", "Xiangxiang Chu", "Liaoni Wu"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-25", "url": "https://arxiv.org/abs/2509.21240", "pdf_url": "https://arxiv.org/pdf/2509.21240v3", "arxiv_id": "2509.21240", "doi": "10.48550/arXiv.2509.21240", "citation_count": 41, "influential_citation_count": 7, "has_code": true, "code_url": "https://github.com/AMAP-ML/Tree-GRPO", "venue": "arXiv.org", "quality_score": 0.4887}
{"id": "aaac21a8e870c364adac58e64d16ac7189bc8a7c00c3d4b72bdb9b006c30dcf9", "sources": ["arxiv", "semantic_scholar"], "title": "MARS: toward more efficient multi-agent collaboration for LLM reasoning", "abstract": "Large language models (LLMs) have achieved impressive results in natural language understanding, yet their reasoning capabilities remain limited when operating as single agents. Multi-Agent Debate (MAD) has been proposed to address this limitation by enabling collaborative reasoning among multiple models in a round-table debate manner. While effective, MAD introduces substantial computational overhead due to the number of agents involved and the frequent communication required. In this paper, we propose MARS (Multi-Agent Review System), a role-based collaboration framework inspired by the review process. In MARS, an author agent generates an initial solution, reviewer agents provide decisions and comments independently, and a meta-reviewer integrates the feedback to make the final decision and guide further revision. This design enhances reasoning quality while avoiding costly reviewer-to-reviewer interactions, thereby controlling token consumption and inference time. We compared MARS with both MAD and other state-of-the-art reasoning strategies across multiple benchmarks. Extensive experiments with different LLMs show that MARS matches the accuracy of MAD while reducing both token usage and inference time by approximately 50\\%. Code is available at https://github.com/xwang97/MARS.", "authors": ["Xiao Wang", "Jia Wang", "Yijie Wang", "Pengtao Dang", "Sha Cao", "Chi Zhang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2509.20502", "pdf_url": "https://arxiv.org/pdf/2509.20502v2", "arxiv_id": "2509.20502", "doi": "10.48550/arXiv.2509.20502", "citation_count": 9, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/xwang97/MARS", "venue": "arXiv.org", "quality_score": 0.487}
{"id": "618f1673414eccffff1661e4815d14ad5a9568cb9ddb0bbfe4002396e1114b9c", "sources": ["arxiv", "semantic_scholar"], "title": "Online-Optimized RAG for Tool Use and Function Calling", "abstract": "In many applications, retrieval-augmented generation (RAG) drives tool use and function calling by embedding the (user) queries and matching them to pre-specified tool/function descriptions. In this paper, we address an embedding misalignment issue that often arises in practical applications due to imperfect embedding models or noisy descriptions; such misalignment may lead to incorrect retrieval and task failure. We introduce Online-Optimized RAG, a deployment-time framework that continually adapts retrieval embeddings from live interactions using minimal feedback (e.g., task success). Online-Optimized RAG applies lightweight online gradient updates with negligible per-query latency and requires no changes to the underlying LLM. The method is plug-and-play: it supports both single- and multi-hop tool use, dynamic tool inventories, and $K$-retrieval with re-ranking. We provide a problem-dependent theoretical analysis that quantifies how the method's performance depends on the initialization quality of the embeddings and other related quantities. Across diverse tool-use and document-retrieval scenarios, our Online-Optimized RAG consistently improves tool selection accuracy and end-task success, thus providing a simple, practical path to robust, self-improving RAG systems.", "authors": ["Yu Pan", "Xiaocheng Li", "Hanzhao Wang"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2509.20415", "pdf_url": "https://arxiv.org/pdf/2509.20415v2", "arxiv_id": "2509.20415", "doi": "10.48550/arXiv.2509.20415", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3151}
{"id": "7c43328d4d9b229c281aecf5b9a2489a427eca8a780e1e9ae465bf1343ce62e6", "sources": ["arxiv", "semantic_scholar"], "title": "MACD: Multi-Agent Clinical Diagnosis with Self-Learned Knowledge for LLM", "abstract": "Large language models (LLMs) have demonstrated notable potential in medical applications, yet they face substantial challenges in handling complex real-world clinical diagnoses using conventional prompting methods. Current prompt engineering and multi-agent approaches typically optimize isolated inferences, neglecting the accumulation of reusable clinical experience. To address this, this study proposes a novel Multi-Agent Clinical Diagnosis (MACD) framework, which allows LLMs to self-learn clinical knowledge via a multi-agent pipeline that summarizes, refines, and applies diagnostic insights. It mirrors how physicians develop expertise through experience, enabling more focused and accurate diagnosis on key disease-specific cues. We further extend it to a MACD-human collaborative workflow, where multiple LLM-based diagnostician agents engage in iterative consultations, supported by an evaluator agent and human oversight for cases where agreement is not reached. Evaluated on 4,390 real-world patient cases across seven diseases using diverse open-source LLMs (Llama-3.1 8B/70B, DeepSeek-R1-Distill-Llama 70B), MACD significantly improves primary diagnostic accuracy, outperforming established clinical guidelines with gains up to 22.3% (MACD). In direct comparison with physician-only diagnosis under the same evaluation protocol, MACD achieves comparable or superior performance, with improvements up to 16%. Furthermore, the MACD-human workflow yields an 18.6% improvement over physician-only diagnosis, demonstrating the synergistic potential of human-AI collaboration. Notably, the self-learned clinical knowledge exhibits strong cross-model stability, transferability across LLMs, and capacity for model-specific personalization.This work thus presents a scalable self-learning paradigm that bridges the gap between the intrinsic knowledge of LLMs.", "authors": ["Wenliang Li", "Rui Yan", "Xu Zhang", "Li Chen", "Hongji Zhu", "Jing Zhao", "Junjun Li", "Mengru Li", "Wei Cao", "Zihang Jiang", "Wei Wei", "Kun Zhang", "Shaohua Kevin Zhou"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2509.20067", "pdf_url": "https://arxiv.org/pdf/2509.20067v4", "arxiv_id": "2509.20067", "doi": "10.48550/arXiv.2509.20067", "citation_count": 5, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.487}
{"id": "821d83ea757ca1aa7b7c5a2eb073156180858ebe5e158bd1ac1f4ba97d1958dc", "sources": ["arxiv", "semantic_scholar"], "title": "EpidemIQs: Prompt-to-Paper LLM Agents for Epidemic Modeling and Analysis", "abstract": "Large Language Models (LLMs) offer new opportunities to accelerate complex interdisciplinary research domains. Epidemic modeling, characterized by its complexity and reliance on network science, dynamical systems, epidemiology, and stochastic simulations, represents a prime candidate for leveraging LLM-driven automation. We introduce EpidemIQs, a novel multi-agent LLM framework that integrates user inputs and autonomously conducts literature review, analytical derivation, network modeling, mechanistic modeling, stochastic simulations, data visualization and analysis, and finally documentation of findings in a structured manuscript, through five predefined research phases. We introduce two types of agents: a scientist agent for planning, coordination, reflection, and generation of final results, and a task-expert agent to focus exclusively on one specific duty serving as a tool to the scientist agent. The framework consistently generated complete reports in scientific article format. Specifically, using GPT 4.1 and GPT 4.1 Mini as backbone LLMs for scientist and task-expert agents, respectively, the autonomous process completes with average total token usage 870K at a cost of about $1.57 per study, successfully executing all phases and final report. We evaluate EpidemIQs across several different epidemic scenarios, measuring computational cost, workflow reliability, task success rate, and LLM-as-Judge and human expert reviews to estimate the overall quality and technical correctness of the generated results. Through our experiments, the framework consistently addresses evaluation scenarios with an average task success rate of 79%. We compare EpidemIQs to an iterative single-agent LLM, benefiting from the same system prompts and tools, iteratively planning, invoking tools, and revising outputs until task completion. The comparisons suggest a consistently higher performance of EpidemIQs.", "authors": ["Mohammad Hossein Samaei", "Faryad Darabi Sahneh", "Lee W. Cohnstaedt", "Caterina Scoglio"], "categories": ["cs.SI", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2510.00024", "pdf_url": "https://arxiv.org/pdf/2510.00024v2", "arxiv_id": "2510.00024", "doi": "10.48550/arXiv.2510.00024", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3151}
{"id": "dcd44a15ee4d303076f6fcb7f5dc7c7e7f70078ff9d4850ffcebab216f1c7a36", "sources": ["arxiv", "semantic_scholar"], "title": "ToolBrain: A Flexible Reinforcement Learning Framework for Agentic Tools", "abstract": "Effective tool use is essential for agentic AI, yet training agents to utilize tools remains challenging due to manually designed rewards, limited training data, and poor multi-tool selection, resulting in slow adaptation, wasted computational resources, and suboptimal performance. We introduce ToolBrain, a lightweight and user-friendly framework for training tool use in agentic models with flexible reinforcement learning, thereby easing the barriers for researchers and practitioners to adapt LLM-based agents to specific domains. It supports a wide range of training strategies, including reinforcement learning algorithms such as GRPO and DPO, as well as supervised learning. ToolBrain enables custom reward callables directly on an agent's execution traces or simply utilizes an automated LLM-as-a-judge system for reward generation. It is packed with useful capabilities, including knowledge distillation from large to small models, automatic task generation from tool descriptions, seamless tool retrieval, efficient fine-tuning pipelines with QLoRA through Unsloth, and quantized inference via bitsandbytes. We demonstrate ToolBrain through an Email Search Agent case study, showing measurable improvements in tool-use skills under a realistic workflow, while keeping the codebase simple and extensible. Our framework is publicly available at https://toolbrain.org/.", "authors": ["Quy Minh Le", "Minh Sao Khue Luu", "Khanh-Tung Tran", "Duc-Hai Nguyen", "Hoang-Quoc-Viet Pham", "Quan Le", "Hoang Thanh Lam", "Hoang D. Nguyen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2510.00023", "pdf_url": "https://arxiv.org/pdf/2510.00023v2", "arxiv_id": "2510.00023", "doi": "10.48550/arXiv.2510.00023", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2005}
{"id": "4c30ffb458f374ea98d1380c5005343f62ea16330409180ed1b32677610b8430", "sources": ["arxiv", "semantic_scholar"], "title": "Training Task Reasoning LLM Agents for Multi-turn Task Planning via Single-turn Reinforcement Learning", "abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities in knowledge acquisition, reasoning, and tool use, making them promising candidates for autonomous agent applications. However, training LLM agents for complex multi-turn task planning faces significant challenges, including sparse episode-wise rewards, credit assignment across long horizons, and the computational overhead of reinforcement learning in multi-turn interaction settings. To this end, this paper introduces a novel approach that transforms multi-turn task planning into single-turn task reasoning problems, enabling efficient policy optimization through Group Relative Policy Optimization (GRPO) with dense and verifiable reward from expert trajectories. Our theoretical analysis shows that GRPO improvement on single-turn task reasoning results in a lower bound of the multi-turn success probability under the minimal turns, as well as the generalization to subtasks with shorter horizons. Experimental evaluation on the complex task planning benchmark demonstrates that our 1.5B parameter model trained with single-turn GRPO achieves superior performance compared to larger baseline models up to 14B parameters, with success rates of 70% for long-horizon planning tasks.", "authors": ["Hanjiang Hu", "Changliu Liu", "Na Li", "Yebin Wang"], "categories": ["cs.LG", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-09-24", "url": "https://arxiv.org/abs/2509.20616", "pdf_url": "https://arxiv.org/pdf/2509.20616v2", "arxiv_id": "2509.20616", "doi": "10.1109/LCSYS.2025.3642767", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Control Systems Letters", "quality_score": 0.3151}
{"id": "588672532fafbb63c9389b2e5c3b285b777534b631785bfae549dd195ac9a6aa", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic AutoSurvey: Let LLMs Survey LLMs", "abstract": "The exponential growth of scientific literature poses unprecedented challenges for researchers attempting to synthesize knowledge across rapidly evolving fields. We present \\textbf{Agentic AutoSurvey}, a multi-agent framework for automated survey generation that addresses fundamental limitations in existing approaches. Our system employs four specialized agents (Paper Search Specialist, Topic Mining \\& Clustering, Academic Survey Writer, and Quality Evaluator) working in concert to generate comprehensive literature surveys with superior synthesis quality. Through experiments on six representative LLM research topics from COLM 2024 categories, we demonstrate that our multi-agent approach achieves significant improvements over existing baselines, scoring 8.18/10 compared to AutoSurvey's 4.77/10. The multi-agent architecture processes 75--443 papers per topic (847 total across six topics) while targeting high citation coverage (often $\\geq$80\\% on 75--100-paper sets; lower on very large sets such as RLHF) through specialized agent orchestration. Our 12-dimension evaluation captures organization, synthesis integration, and critical analysis beyond basic metrics. These findings demonstrate that multi-agent architectures represent a meaningful advancement for automated literature survey generation in rapidly evolving scientific domains.", "authors": ["Yixin Liu", "Yonghui Wu", "Denghui Zhang", "Lichao Sun"], "categories": ["cs.IR", "cs.CL", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-23", "url": "https://arxiv.org/abs/2509.18661", "pdf_url": "https://arxiv.org/pdf/2509.18661v1", "arxiv_id": "2509.18661", "doi": "10.48550/arXiv.2509.18661", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.314}
{"id": "4519594db537bec347dc4af9947ccf91b83d8e75a1dbd23a18366a30d9c301b2", "sources": ["arxiv", "semantic_scholar"], "title": "The Heterogeneous Multi-Agent Challenge", "abstract": "Multi-Agent Reinforcement Learning (MARL) is a growing research area which gained significant traction in recent years, extending Deep RL applications to a much wider range of problems. A particularly challenging class of problems in this domain is Heterogeneous Multi-Agent Reinforcement Learning (HeMARL), where agents with different sensors, resources, or capabilities must cooperate based on local information. The large number of real-world situations involving heterogeneous agents makes it an attractive research area, yet underexplored, as most MARL research focuses on homogeneous agents (e.g., a swarm of identical robots). In MARL and single-agent RL, standardized environments such as ALE and SMAC have allowed to establish recognized benchmarks to measure progress. However, there is a clear lack of such standardized testbed for cooperative HeMARL. As a result, new research in this field often uses simple environments, where most algorithms perform near optimally, or uses weakly heterogeneous MARL environments.", "authors": ["Charles Dansereau", "Junior-Samuel Lopez-Yepez", "Karthik Soma", "Antoine Fagette"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-23", "url": "https://arxiv.org/abs/2509.19512", "pdf_url": "https://arxiv.org/pdf/2509.19512v1", "arxiv_id": "2509.19512", "doi": "10.48550/arXiv.2509.19512", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.314}
{"id": "7fac3287de8935ed4b62b827b80fe97465e8d7475136d91e0e2fd70d0780be05", "sources": ["arxiv", "semantic_scholar"], "title": "Medical AI Consensus: A Multi-Agent Framework for Radiology Report Generation and Evaluation", "abstract": "Automating radiology report generation poses a dual challenge: building clinically reliable systems and designing rigorous evaluation protocols. We introduce a multi-agent reinforcement learning framework that serves as both a benchmark and evaluation environment for multimodal clinical reasoning in the radiology ecosystem. The proposed framework integrates large language models (LLMs) and large vision models (LVMs) within a modular architecture composed of ten specialized agents responsible for image analysis, feature extraction, report generation, review, and evaluation. This design enables fine-grained assessment at both the agent level (e.g., detection and segmentation accuracy) and the consensus level (e.g., report quality and clinical relevance). We demonstrate an implementation using chatGPT-4o on public radiology datasets, where LLMs act as evaluators alongside medical radiologist feedback. By aligning evaluation protocols with the LLM development lifecycle, including pretraining, finetuning, alignment, and deployment, the proposed benchmark establishes a path toward trustworthy deviance-based radiology report generation.", "authors": ["Ahmed T. Elboardy", "Ghada Khoriba", "Essam A. Rashed"], "categories": ["cs.AI", "eess.IV", "physics.med-ph"], "fields_of_study": ["Computer Science", "Engineering", "Physics"], "published_date": "2025-09-22", "url": "https://arxiv.org/abs/2509.17353", "pdf_url": "https://arxiv.org/pdf/2509.17353v1", "arxiv_id": "2509.17353", "doi": "10.48550/arXiv.2509.17353", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3128}
{"id": "7c5c68210b2cabfcb44dced9b54af42e96f31b41a6b4dd94368fcb6366132223", "sources": ["arxiv", "semantic_scholar"], "title": "MapCoder-Lite: Distilling Multi-Agent Coding into a Single Small LLM", "abstract": "Large language models (LLMs) have advanced code generation from single-function tasks to competitive-programming problems, but existing multi-agent solutions either rely on costly large-scale (>30B) models or collapse when downsized to small open-source models. We present MapCoder-Lite, a framework for distilling the complex reasoning of large, multi-agent coding systems into a single 7B model. Our contribution is a novel, three-pillar methodology that synergistically generates, refines, and encodes multi-agent knowledge: (i) pass-based trajectory distillation from strong LLMs fixes format fragility in retrieval and reduces failures in debugging, (ii) supervisor-guided correction with global feedback strengthens planning and coding agents, and (iii) agent-wise LoRA fine-tuning delivers memory-efficient specialisation. Comprehensive evaluation on xCodeEval, APPS, and CodeContests shows that MapCoder-Lite more than doubles xCodeEval accuracy (from 13.2% to 28.3%), eliminates all format failures, while reducing GPU memory and token-generation time by 4x compared to a 32B model. It also achieves over 10% gains on simpler coding benchmarks, demonstrating broad improvements beyond competitive programming. These results demonstrate that careful agent-wise fine-tuning unleashes high-quality multi-agent coding on a small language model. Our code is publicly available at https://github.com/aiha-lab/MapCoder-Lite.", "authors": ["Woongkyu Lee", "Junhee Cho", "Jungwook Choi"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-22", "url": "https://arxiv.org/abs/2509.17489", "pdf_url": "https://arxiv.org/pdf/2509.17489v2", "arxiv_id": "2509.17489", "doi": "10.18653/v1/2026.findings-eacl.346", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/aiha-lab/MapCoder-Lite", "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.4834}
{"id": "81b45e3ca63246addf555d1c5d64b872affdf1b83972b2eaf76f1702e733c257", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic ReAct: Scalable Tool Selection for Large-Scale MCP Environments", "abstract": "We present Dynamic ReAct, a novel approach for enabling ReAct agents to efficiently operate with extensive Model Control Protocol (MCP) tool sets that exceed the contextual memory limitations of large language models. Our approach addresses the fundamental challenge of tool selection in environments containing hundreds or thousands of available tools, where loading all tools simultaneously is computationally infeasible. We propose and evaluate five distinct architectures that progressively refine the tool selection process, culminating in a search-and-load mechanism that achieves intelligent tool selection with minimal computational overhead. Our experimental results demonstrate that the proposed approach reduces tool loading by up to 50% while maintaining task completion accuracy, advancing the path towards truly general-purpose AI agents capable of dynamically adapting to diverse task environments.", "authors": ["Nishant Gaurav", "Adit Akarsh", "Ankit Ranjan", "Manoj Bajaj"], "categories": ["cs.SE", "cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-22", "url": "https://arxiv.org/abs/2509.20386", "pdf_url": "https://arxiv.org/pdf/2509.20386v1", "arxiv_id": "2509.20386", "doi": "10.48550/arXiv.2509.20386", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3128}
{"id": "ebe619202fa8ad3d257102f974574fb95cc32a118dce71e9ced454e76bdec932", "sources": ["arxiv", "semantic_scholar"], "title": "MSCoRe: A Benchmark for Multi-Stage Collaborative Reasoning in LLM Agents", "abstract": "Large Language Models (LLMs) have excelled in question-answering (QA) tasks within single domains. However, their reasoning and coordination capabilities in complex, multi-stage scenarios remain underexplored. Existing benchmarks typically focus on isolated tasks or narrow domains, overlooking models' abilities for multi-stage collaboration and optimization without explicit external guidance. To bridge this gap, we propose \\textbf{MSCoRe}, a novel benchmark comprising 126696 domain-specific QA instances spanning scenarios in automotive, pharmaceutical, electronics, and energy sectors. The dataset is created using a structured three-phase pipeline: dynamic sampling, iterative question-answer generation, and a multi-level quality assessment to ensure data quality. Tasks are further categorized into three difficulty levels according to stage coverage and complexity. With MSCoRe, we have conducted a comprehensive evaluation of various state-of-the-art LLM agents. The commercial models performed best across all tasks and scenarios, but a notable gap in ROUGE scores remains between simple and complex tasks. We also tested the models' robustness and found that their performance is negatively affected by noisy data. MSCoRe provides a valuable new resource for the community to evaluate and improve multi-stage reasoning in LLM agents. The code and data are available at https://github.com/D3E0-source/MSCoRE.", "authors": ["Yuzhen Lei", "Hongbin Xie", "Jiaxing Zhao", "Shuangxue Liu", "Xuan Song"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-22", "url": "https://arxiv.org/abs/2509.17628", "pdf_url": "https://arxiv.org/pdf/2509.17628v1", "arxiv_id": "2509.17628", "doi": "10.48550/arXiv.2509.17628", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/D3E0-source/MSCoRE", "venue": "arXiv.org", "quality_score": 0.4834}
{"id": "5ee92d3c0b4a6a7bc618c0ae7bc48e904fb90f64a1ef44289fe113ef9424069f", "sources": ["arxiv", "semantic_scholar"], "title": "SFT-TA: Supervised Fine-Tuned Agents in Multi-Agent LLMs for Automated Inductive Thematic Analysis", "abstract": "Thematic Analysis (TA) is a widely used qualitative method that provides a structured yet flexible framework for identifying and reporting patterns in clinical interview transcripts. However, manual thematic analysis is time-consuming and limits scalability. Recent advances in LLMs offer a pathway to automate thematic analysis, but alignment with human results remains limited. To address these limitations, we propose SFT-TA, an automated thematic analysis framework that embeds supervised fine-tuned (SFT) agents within a multi-agent system. Our framework outperforms existing frameworks and the gpt-4o baseline in alignment with human reference themes. We observed that SFT agents alone may underperform, but achieve better results than the baseline when embedded within a multi-agent system. Our results highlight that embedding SFT agents in specific roles within a multi-agent system is a promising pathway to improve alignment with desired outputs for thematic analysis.", "authors": ["Seungjun Yi", "Joakim Nguyen", "Huimin Xu", "Terence Lim", "Joseph Skrovan", "Mehak Beri", "Hitakshi Modi", "Andrew Well", "Liu Leqi", "Mia Markey", "Ying Ding"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-21", "url": "https://arxiv.org/abs/2509.17167", "pdf_url": "https://arxiv.org/pdf/2509.17167v1", "arxiv_id": "2509.17167", "doi": "10.48550/arXiv.2509.17167", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3117}
{"id": "82e4b43490eeb9a3c645227beaca114dbd29347a6c46eaf3a51379096bb53acc", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Transparent and Incentive-Compatible Collaboration in Decentralized LLM Multi-Agent Systems: A Blockchain-Driven Approach", "abstract": "Large Language Models (LLMs) have enabled the emergence of autonomous agents capable of complex reasoning, planning, and interaction. However, coordinating such agents at scale remains a fundamental challenge, particularly in decentralized environments where communication lacks transparency and agent behavior cannot be shaped through centralized incentives. We propose a blockchain-based framework that enables transparent agent registration, verifiable task allocation, and dynamic reputation tracking through smart contracts. The core of our design lies in two mechanisms: a matching score-based task allocation protocol that evaluates agents by reputation, capability match, and workload; and a behavior-shaping incentive mechanism that adjusts agent behavior via feedback on performance and reward. Our implementation integrates GPT-4 agents with Solidity contracts and demonstrates, through 50-round simulations, strong task success rates, stable utility distribution, and emergent agent specialization. The results underscore the potential for trustworthy, incentive-compatible multi-agent coordination in open environments.", "authors": ["Minfeng Qi", "Tianqing Zhu", "Lefeng Zhang", "Ningran Li", "Wanlei Zhou"], "categories": ["cs.MA", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-20", "url": "https://arxiv.org/abs/2509.16736", "pdf_url": "https://arxiv.org/pdf/2509.16736v1", "arxiv_id": "2509.16736", "doi": "10.48550/arXiv.2509.16736", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3105}
{"id": "340d3bb85db22399fd0989b61dabad5e5108324106f5d7da78b15c48f5c4be08", "sources": ["arxiv", "semantic_scholar"], "title": "Can an Individual Manipulate the Collective Decisions of Multi-Agents?", "abstract": "Individual Large Language Models (LLMs) have demonstrated significant capabilities across various domains, such as healthcare and law. Recent studies also show that coordinated multi-agent systems exhibit enhanced decision-making and reasoning abilities through collaboration. However, due to the vulnerabilities of individual LLMs and the difficulty of accessing all agents in a multi-agent system, a key question arises: If attackers only know one agent, could they still generate adversarial samples capable of misleading the collective decision? To explore this question, we formulate it as a game with incomplete information, where attackers know only one target agent and lack knowledge of the other agents in the system. With this formulation, we propose M-Spoiler, a framework that simulates agent interactions within a multi-agent system to generate adversarial samples. These samples are then used to manipulate the target agent in the target system, misleading the system's collaborative decision-making process. More specifically, M-Spoiler introduces a stubborn agent that actively aids in optimizing adversarial samples by simulating potential stubborn responses from agents in the target system. This enhances the effectiveness of the generated adversarial samples in misleading the system. Through extensive experiments across various tasks, our findings confirm the risks posed by the knowledge of an individual agent in multi-agent systems and demonstrate the effectiveness of our framework. We also explore several defense mechanisms, showing that our proposed attack framework remains more potent than baselines, underscoring the need for further research into defensive strategies.", "authors": ["Fengyuan Liu", "Rui Zhao", "Shuo Chen", "Guohao Li", "Philip Torr", "Lei Han", "Jindong Gu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-20", "url": "https://arxiv.org/abs/2509.16494", "pdf_url": "https://arxiv.org/pdf/2509.16494v2", "arxiv_id": "2509.16494", "doi": "10.48550/arXiv.2509.16494", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3105}
{"id": "46d01283c5319a0bcc802084a18ef096167f895cfd21876b810cfe208ea46410", "sources": ["arxiv", "semantic_scholar"], "title": "SLM-Based Agentic AI with P-C-G: Optimized for Korean Tool Use", "abstract": "We propose a small-scale language model (SLM) based agent architecture, Planner-Caller-Generator (P-C-G), optimized for Korean tool use. P-C-G separates planning, calling, and generation by role: the Planner produces an initial batch plan with limited on-demand replanning; the Caller returns a normalized call object after joint schema-value validation; and the Generator integrates tool outputs to produce the final answer. We apply a Korean-first value policy to reduce execution failures caused by frequent Korean-to-English code switching in Korean settings. Evaluation assumes Korean queries and Korean tool/parameter specifications; it covers single-chain, multi-chain, missing-parameters, and missing-functions scenarios, and is conducted via an LLM-as-a-Judge protocol averaged over five runs under a unified I/O interface. Results show that P-C-G delivers competitive tool-use accuracy and end-to-end quality while reducing tokens and maintaining acceptable latency, indicating that role-specialized SLMs are a cost-effective alternative for Korean tool-use agents.", "authors": ["Changhyun Jeon", "Jinhee Park", "Jungwoo Choi", "Keonwoo Kim", "Jisu Kim", "Minji Hong"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-19", "url": "https://arxiv.org/abs/2509.19369", "pdf_url": "https://arxiv.org/pdf/2509.19369v1", "arxiv_id": "2509.19369", "doi": "10.48550/arXiv.2509.19369", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3094}
{"id": "a01dd218f0b4bc8963c628dd99ffdab0d2fc1faa56ef09ece84369a17623e707", "sources": ["arxiv", "semantic_scholar"], "title": "Sentinel Agents for Secure and Trustworthy Agentic AI in Multi-Agent Systems", "abstract": "This paper proposes a novel architectural framework aimed at enhancing security and reliability in multi-agent systems (MAS). A central component of this framework is a network of Sentinel Agents, functioning as a distributed security layer that integrates techniques such as semantic analysis via large language models (LLMs), behavioral analytics, retrieval-augmented verification, and cross-agent anomaly detection. Such agents can potentially oversee inter-agent communications, identify potential threats, enforce privacy and access controls, and maintain comprehensive audit records. Complementary to the idea of Sentinel Agents is the use of a Coordinator Agent. The Coordinator Agent supervises policy implementation, and manages agent participation. In addition, the Coordinator also ingests alerts from Sentinel Agents. Based on these alerts, it can adapt policies, isolate or quarantine misbehaving agents, and contain threats to maintain the integrity of the MAS ecosystem. This dual-layered security approach, combining the continuous monitoring of Sentinel Agents with the governance functions of Coordinator Agents, supports dynamic and adaptive defense mechanisms against a range of threats, including prompt injection, collusive agent behavior, hallucinations generated by LLMs, privacy breaches, and coordinated multi-agent attacks. In addition to the architectural design, we present a simulation study where 162 synthetic attacks of different families (prompt injection, hallucination, and data exfiltration) were injected into a multi-agent conversational environment. The Sentinel Agents successfully detected the attack attempts, confirming the practical feasibility of the proposed monitoring approach. The framework also offers enhanced system observability, supports regulatory compliance, and enables policy evolution over time.", "authors": ["Diego Gosmar", "Deborah A. Dahl"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-18", "url": "https://arxiv.org/abs/2509.14956", "pdf_url": "https://arxiv.org/pdf/2509.14956v1", "arxiv_id": "2509.14956", "doi": "10.48550/arXiv.2509.14956", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3082}
{"id": "1eb191980cd3e0a999c5e2b9b124afc47718d200693a0d49a227d6aab86dd74a", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agents at the Roundtable: A Multi-Perspective and Dialectical Reasoning Framework for Essay Scoring", "abstract": "The emergence of large language models (LLMs) has brought a new paradigm to automated essay scoring (AES), a long-standing and practical application of natural language processing in education. However, achieving human-level multi-perspective understanding and judgment remains a challenge. In this work, we propose Roundtable Essay Scoring (RES), a multi-agent evaluation framework designed to perform precise and human-aligned scoring under a zero-shot setting. RES constructs evaluator agents based on LLMs, each tailored to a specific prompt and topic context. Each agent independently generates a trait-based rubric and conducts a multi-perspective evaluation. Then, by simulating a roundtable-style discussion, RES consolidates individual evaluations through a dialectical reasoning process to produce a final holistic score that more closely aligns with human evaluation. By enabling collaboration and consensus among agents with diverse evaluation perspectives, RES outperforms prior zero-shot AES approaches. Experiments on the ASAP dataset using ChatGPT and Claude show that RES achieves up to a 34.86% improvement in average QWK over straightforward prompting (Vanilla) methods.", "authors": ["Jinhee Jang", "Ayoung Moon", "Minkyoung Jung", "YoungBin Kim", "Seung Jin Lee"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-18", "url": "https://arxiv.org/abs/2509.14834", "pdf_url": "https://arxiv.org/pdf/2509.14834v2", "arxiv_id": "2509.14834", "doi": "10.48550/arXiv.2509.14834", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3082}
{"id": "209b286e8fb0fcca06ccb1ee63326928d68c498747f115fd16f71971b7881c9f", "sources": ["arxiv", "semantic_scholar"], "title": "Diagnostics of cognitive failures in multi-agent expert systems using dynamic evaluation protocols and subsequent mutation of the processing context", "abstract": "The rapid evolution of neural architectures - from multilayer perceptrons to large-scale Transformer-based models - has enabled language models (LLMs) to exhibit emergent agentic behaviours when equipped with memory, planning, and external tool use. However, their inherent stochasticity and multi-step decision processes render classical evaluation methods inadequate for diagnosing agentic performance. This work introduces a diagnostic framework for expert systems that not only evaluates but also facilitates the transfer of expert behaviour into LLM-powered agents. The framework integrates (i) curated golden datasets of expert annotations, (ii) silver datasets generated through controlled behavioural mutation, and (iii) an LLM-based Agent Judge that scores and prescribes targeted improvements. These prescriptions are embedded into a vectorized recommendation map, allowing expert interventions to propagate as reusable improvement trajectories across multiple system instances. We demonstrate the framework on a multi-agent recruiter-assistant system, showing that it uncovers latent cognitive failures - such as biased phrasing, extraction drift, and tool misrouting - while simultaneously steering agents toward expert-level reasoning and style. The results establish a foundation for standardized, reproducible expert behaviour transfer in stochastic, tool-augmented LLM agents, moving beyond static evaluation to active expert system refinement.", "authors": ["Andrejs Sorstkins", "Josh Bailey", "Dr Alistair Baron"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-18", "url": "https://arxiv.org/abs/2509.15366", "pdf_url": "https://arxiv.org/pdf/2509.15366v1", "arxiv_id": "2509.15366", "doi": "10.48550/arXiv.2509.15366", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3082}
{"id": "f0344042b44cecb0ada03d87f2b7af03e3714d080f7ea7f17e1e2355ea031d87", "sources": ["arxiv", "semantic_scholar"], "title": "LEED: A Highly Efficient and Scalable LLM-Empowered Expert Demonstrations Framework for Multi-Agent Reinforcement Learning", "abstract": "Multi-agent reinforcement learning (MARL) holds substantial promise for intelligent decision-making in complex environments. However, it suffers from a coordination and scalability bottleneck as the number of agents increases. To address these issues, we propose the LLM-empowered expert demonstrations framework for multi-agent reinforcement learning (LEED). LEED consists of two components: a demonstration generation (DG) module and a policy optimization (PO) module. Specifically, the DG module leverages large language models to generate instructions for interacting with the environment, thereby producing high-quality demonstrations. The PO module adopts a decentralized training paradigm, where each agent utilizes the generated demonstrations to construct an expert policy loss, which is then integrated with its own policy loss. This enables each agent to effectively personalize and optimize its local policy based on both expert knowledge and individual experience. Experimental results show that LEED achieves superior sample efficiency, time efficiency, and robust scalability compared to state-of-the-art baselines.", "authors": ["Tianyang Duan", "Zongyuan Zhang", "Songxiao Guo", "Dong Huang", "Yuanye Zhao", "Zheng Lin", "Zihan Fang", "Dianxin Luan", "Heming Cui", "Yong Cui"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-18", "url": "https://arxiv.org/abs/2509.14680", "pdf_url": "https://arxiv.org/pdf/2509.14680v1", "arxiv_id": "2509.14680", "doi": "10.48550/arXiv.2509.14680", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3082}
{"id": "b4d2f8fecd6b382b5c1f9fd1318333155c9a98db78fd070d51c4351467c2658f", "sources": ["arxiv", "semantic_scholar"], "title": "Vulnerable Agent Identification in Large-Scale Multi-Agent Reinforcement Learning", "abstract": "Partial agent failure becomes inevitable when systems scale up, making it crucial to identify the subset of agents whose failure causes worst-case system performance degradations. We study this Vulnerable Agent Identification (VAI) problem in large-scale multi-agent reinforcement learning (MARL). We frame VAI as a Hierarchical Adversarial Decentralized Mean Field Control (HAD-MFC), where the upper level selects vulnerable agents as an NP-hard task and the lower level learns their worst-case adversarial policies via mean-field MARL. The two problems are coupled together, making HAD-MFC difficult to solve. To handle this, we first decouple the hierarchical process by Fenchel-Rockafellar transform, resulting a regularized mean-field Bellman operator for upper level that enables independent learning at each level, thus reducing computational complexity. We next reformulate the upper-level NP-hard problem as an MDP with dense rewards, allowing sequential identification of vulnerable agents via greedy and RL algorithms. This decomposition provably preserves the optimal solution. Experiments show our method effectively identifies more vulnerable agents in large-scale MARL and the rule-based system, fooling system into worse failures, and reveals the vulnerability of each agent in large systems. Code available at https://github.com/Waken-dream/VAI", "authors": ["Simin Li", "Zihao Mao", "Zheng Yuwei", "Linhao Wang", "Ruixiao Xu", "Chengdong Ma", "Zhiqian Liu", "Xin Yu", "Yuqing Ma", "Xin Wang", "Jie Luo", "Bo An", "Yaodong Yang", "Weifeng Lv", "Xianglong Liu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-18", "url": "https://arxiv.org/abs/2509.15103", "pdf_url": "https://arxiv.org/pdf/2509.15103v3", "arxiv_id": "2509.15103", "doi": "10.48550/arXiv.2509.15103", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Waken-dream/VAI", "venue": "arXiv.org", "quality_score": 0.4764}
{"id": "cb992ca20cb647b7b9a61767ea0802b1927294cc3e12d7df9efe4a674337cdfd", "sources": ["arxiv", "semantic_scholar"], "title": "Process-Supervised Reinforcement Learning for Interactive Multimodal Tool-Use Agents", "abstract": "Effective interactive tool use requires agents to master Tool Integrated Reasoning (TIR): a complex process involving multi-turn planning and long-context dialogue management. To train agents for this dynamic process, particularly in multi-modal contexts, we introduce a sandbox environment for reinforcement learning (RL) that supports interleaved speech-text rollouts. Our core strategy, Turn-level Adjudicated Reinforcement Learning (TARL), addresses the challenge of credit assignment in long-horizon tasks by employing a Large Language Model (LLM) as a judge to provide turn-level evaluation. To enhance exploration, we integrate a mixed-task training curriculum with mathematical reasoning problems. This unified approach boosts the task pass rate on the text-based $τ$-bench by over 6% compared to strong RL baselines. Crucially, we demonstrate our framework's suitability for fine-tuning a multi-modal foundation model for agentic tasks. By training a base multi-modal LLM on interleaved speech-text rollouts, we equip it with tool-use abilities, paving the way for more natural, voice-driven interactive agents.", "authors": ["Weiting Tan", "Xinghua Qu", "Ming Tu", "Meng Ge", "Andy T. Liu", "Philipp Koehn", "Lu Lu"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-17", "url": "https://arxiv.org/abs/2509.14480", "pdf_url": "https://arxiv.org/pdf/2509.14480v1", "arxiv_id": "2509.14480", "doi": "10.48550/arXiv.2509.14480", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3071}
{"id": "2848d9f8805d6558f73a76bd1bb187313d75cd5bf8e76039bf7d041bd59a1a35", "sources": ["arxiv", "semantic_scholar"], "title": "Foam-Agent 2.0: An End-to-End Composable Multi-Agent Framework for Automating CFD Simulation in OpenFOAM", "abstract": "Computational Fluid Dynamics (CFD) is an essential simulation tool in engineering, yet its steep learning curve and complex manual setup create significant barriers. To address these challenges, we introduce Foam-Agent, a multi-agent framework that automates the entire end-to-end OpenFOAM workflow from a single natural language prompt. Our key innovations address critical gaps in existing systems: 1. An Comprehensive End-to-End Simulation Automation: Foam-Agent is the first system to manage the full simulation pipeline, including advanced pre-processing with a versatile Meshing Agent capable of handling external mesh files and generating new geometries via Gmsh, automatic generation of HPC submission scripts, and post-simulation visualization via ParaView. 2. Composable Service Architecture: Going beyond a monolithic agent, the framework uses Model Context Protocol (MCP) to expose its core functions as discrete, callable tools. This allows for flexible integration and use by other agentic systems, such as Claude-code, for more exploratory workflows. 3. High-Fidelity Configuration Generation: We achieve superior accuracy through a Hierarchical Multi-Index RAG for precise context retrieval and a dependency-aware generation process that ensures configuration consistency. Evaluated on a benchmark of 110 simulation tasks, Foam-Agent achieves an 88.2% success rate with Claude 3.5 Sonnet, significantly outperforming existing frameworks (55.5% for MetaOpenFOAM). Foam-Agent dramatically lowers the expertise barrier for CFD, demonstrating how specialized multi-agent systems can democratize complex scientific computing. The code is public at https://github.com/csml-rpi/Foam-Agent.", "authors": ["Ling Yue", "Nithin Somasekharan", "Tingwen Zhang", "Yadi Cao", "Shaowu Pan"], "categories": ["cs.AI", "cs.CE", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-17", "url": "https://arxiv.org/abs/2509.18178", "pdf_url": "https://arxiv.org/pdf/2509.18178v2", "arxiv_id": "2509.18178", "doi": "10.48550/arXiv.2509.18178", "citation_count": 12, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/csml-rpi/Foam-Agent", "venue": "arXiv.org", "quality_score": 0.4746}
{"id": "b465de9ba5993cef095a40cdad075875dc2091711a059392dbe96ea191ce0603", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM-based multi-agent framework for agile effort estimation", "abstract": "Effort estimation is a crucial activity in agile software development, where teams collaboratively review, discuss, and estimate the effort required to complete user stories in a product backlog. Current practices in agile effort estimation heavily rely on subjective assessments, leading to inaccuracies and inconsistencies in the estimates. While recent machine learning-based methods show promising accuracy, they cannot explain or justify their estimates and lack the capability to interact with human team members. Our paper fills this significant gap by leveraging the powerful capabilities of Large Language Models (LLMs). We propose a novel LLM-based multi-agent framework for agile estimation that not only can produce estimates, but also can coordinate, communicate and discuss with human developers and other agents to reach a consensus. Evaluation results on a real-life dataset show that our approach outperforms state-of-the-art techniques across all evaluation metrics in the majority of the cases. Our human study with software development practitioners also demonstrates an overwhelmingly positive experience in collaborating with our agents in agile effort estimation.", "authors": ["Thanh-Long Bui", "Hoa Khanh Dam", "Rashina Hoda"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-17", "url": "https://arxiv.org/abs/2509.14483", "pdf_url": "https://arxiv.org/pdf/2509.14483v1", "arxiv_id": "2509.14483", "doi": "10.1109/ASE63991.2025.00090", "citation_count": 9, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Automated Software Engineering", "quality_score": 0.3071}
{"id": "97177afa72d8c24c5b62898fd2f67df76f3667450f68f578beefb0aacabe1bc8", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent LLM Defense Pipeline Against Prompt Injection Attacks", "abstract": "Prompt injection attacks represent a major vulnerability in Large Language Model (LLM) deployments, where malicious instructions embedded in user inputs can override system prompts and induce unintended behaviors. This paper presents a novel multi-agent defense framework that employs specialized LLM agents in coordinated pipelines to detect and neutralize prompt injection attacks in real-time. We evaluate our approach using two distinct architectures: a sequential chain-of-agents pipeline and a hierarchical coordinator-based system. Our comprehensive evaluation on 55 unique prompt injection attacks, grouped into 8 categories and totaling 400 attack instances across two LLM platforms (ChatGLM and Llama2), demonstrates significant security improvements. Without defense mechanisms, baseline Attack Success Rates (ASR) reached 30% for ChatGLM and 20% for Llama2. Our multi-agent pipeline achieved 100% mitigation, reducing ASR to 0% across all tested scenarios. The framework demonstrates robustness across multiple attack categories including direct overrides, code execution attempts, data exfiltration, and obfuscation techniques, while maintaining system functionality for legitimate queries.", "authors": ["S M Asif Hossain", "Ruksat Khan Shayoni", "Mohd Ruhul Ameen", "Akif Islam", "M. F. Mridha", "Jungpil Shin"], "categories": ["cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-16", "url": "https://arxiv.org/abs/2509.14285", "pdf_url": "https://arxiv.org/pdf/2509.14285v4", "arxiv_id": "2509.14285", "doi": "10.1109/WIECON-ECE69386.2025.11526251", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE International WIE Conference on Electrical and Computer Engineering", "quality_score": 0.3059}
{"id": "13083579e4ca19cac4631db5d9dae3b15ae8c1655528c6b7bca2827a11b37a02", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM Agentic Approach for Legal-Critical Software: A Case Study for Tax Prep Software", "abstract": "Large language models (LLMs) show promise for translating natural-language statutes into executable logic, but reliability in legally critical settings remains challenging due to ambiguity and hallucinations. We present an agentic approach for developing legal-critical software, using U.S. federal tax preparation as a case study. The key challenge is test-case generation under the oracle problem, where correct outputs require interpreting law. Building on metamorphic testing, we introduce higher-order metamorphic relations that compare system outputs across structured shifts among similar individuals. Because authoring such relations is tedious and error-prone, we use an LLM-driven, role-based framework to automate test generation and code synthesis. We implement a multi-agent system that translates tax code into executable software and incorporates a metamorphic-testing agent that searches for counterexamples. In experiments, our framework using a smaller model (GPT-4o-mini) achieves a worst-case pass rate of 45%, outperforming frontier models (GPT-4o and Claude 3.5, 9-15%) on complex tax-code tasks. These results support agentic LLM methodologies as a path to robust, trustworthy legal-critical software from natural-language specifications.", "authors": ["Sina Gogani-Khiabani", "Ashutosh Trivedi", "Diptikalyan Saha", "Saeid Tizpaz-Niari"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-16", "url": "https://arxiv.org/abs/2509.13471", "pdf_url": "https://arxiv.org/pdf/2509.13471v2", "arxiv_id": "2509.13471", "doi": "10.1145/3744916.3764575", "citation_count": 4, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3059}
{"id": "17bcb65c62eb51bd27662b27dd8ba3c0f767e4fc9741bf8f25b9b47e2cb952e1", "sources": ["arxiv", "semantic_scholar"], "title": "MALLM: Multi-Agent Large Language Models Framework", "abstract": "Multi-agent debate (MAD) has demonstrated the ability to augment collective intelligence by scaling test-time compute and leveraging expertise. Current frameworks for multi-agent debate are often designed towards tool use, lack integrated evaluation, or provide limited configurability of agent personas, response generators, discussion paradigms, and decision protocols. We introduce MALLM (Multi-Agent Large Language Models), an open-source framework that enables systematic analysis of MAD components. MALLM offers more than 144 unique configurations of MAD, including (1) agent personas (e.g., Expert, Personality), (2) response generators (e.g., Critical, Reasoning), (3) discussion paradigms (e.g., Memory, Relay), and (4) decision protocols (e.g., Voting, Consensus). MALLM uses simple configuration files to define a debate. Furthermore, MALLM can load any textual Hugging Face dataset (e.g., MMLU-Pro, WinoGrande) and provides an evaluation pipeline for easy comparison of MAD configurations. MALLM enables researchers to systematically configure, run, and evaluate debates for their problems, facilitating the understanding of the components and their interplay.", "authors": ["Jonas Becker", "Lars Benedikt Kaesberg", "Niklas Bauer", "Jan Philip Wahle", "Terry Ruas", "Bela Gipp"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-15", "url": "https://arxiv.org/abs/2509.11656", "pdf_url": "https://arxiv.org/pdf/2509.11656v3", "arxiv_id": "2509.11656", "doi": "10.48550/arXiv.2509.11656", "citation_count": 10, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.471}
{"id": "5fc880ac30adf12be44db18aaab8961fd4404d304057928921354386bd6cfe0d", "sources": ["arxiv", "semantic_scholar"], "title": "Difficulty-Aware Agentic Orchestration for Query-Specific Multi-Agent Workflows", "abstract": "Large Language Model (LLM)-based agentic systems have shown strong capabilities across various tasks. However, existing multi-agent frameworks often rely on static or task-level workflows, which either over-process simple queries or underperform on complex ones, while also neglecting the efficiency-performance trade-offs across heterogeneous LLMs. To address these limitations, we propose Difficulty-Aware Agentic Orchestration (DAAO), which can dynamically generate query-specific multi-agent workflows guided by predicted query difficulty. DAAO comprises three interdependent modules: a variational autoencoder (VAE) for difficulty estimation, a modular operator allocator, and a cost- and performance-aware LLM router. A self-adjusting policy updates difficulty estimates based on workflow success, enabling simpler workflows for easy queries and more complex strategies for harder ones. Experiments on six benchmarks demonstrate that DAAO surpasses prior multi-agent systems in both accuracy and inference efficiency, validating its effectiveness for adaptive, difficulty-aware reasoning.", "authors": ["Jinwei Su", "Qizhen Lan", "Yinghui Xia", "Lifan Sun", "Weiyou Tian", "Tianyu Shi", "Xinyuan Song", "Lewei He", "Yang Jingsong"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-14", "url": "https://arxiv.org/abs/2509.11079", "pdf_url": "https://arxiv.org/pdf/2509.11079v5", "arxiv_id": "2509.11079", "doi": "10.1145/3774904.3792240", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.294}
{"id": "6fa00c158ca84e23af42033fcedd339d7cad77e43ab152b9311bf454913fd9a5", "sources": ["arxiv", "semantic_scholar"], "title": "QuantAgent: Price-Driven Multi-Agent LLMs for High-Frequency Trading", "abstract": "Recent advances in Large Language Models (LLMs) have shown remarkable capabilities in financial reasoning and market understanding. Multi-agent LLM frameworks such as TradingAgent and FINMEM augment these models to long-horizon investment tasks by leveraging fundamental and sentiment-based inputs for strategic decision-making. However, these approaches are ill-suited for the high-speed, precision-critical demands of High-Frequency Trading (HFT). HFT typically requires rapid, risk-aware decisions driven by structured, short-horizon signals, such as technical indicators, chart patterns, and trend features. These signals stand in sharp contrast to the long-horizon, text-driven reasoning that characterizes most existing LLM-based systems in finance. To bridge this gap, we introduce QuantAgent, the first multi-agent LLM framework explicitly designed for high-frequency algorithmic trading. The system decomposes trading into four specialized agents--Indicator, Pattern, Trend, and Risk--each equipped with domain-specific tools and structured reasoning capabilities to capture distinct aspects of market dynamics over short temporal windows. Extensive experiments across nine financial instruments, including Bitcoin and Nasdaq futures, demonstrate that QuantAgent consistently outperforms baseline methods, achieving higher predictive accuracy at both 1-hour and 4-hour trading intervals across multiple evaluation metrics. Our findings suggest that coupling structured trading signals with LLM-based reasoning provides a viable path for traceable, real-time decision systems in high-frequency financial markets.", "authors": ["Fei Xiong", "Xiang Zhang", "Aosong Feng", "Siqi Sun", "Chenyu You"], "categories": ["cs.CE"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-12", "url": "https://arxiv.org/abs/2509.09995", "pdf_url": "https://arxiv.org/pdf/2509.09995v3", "arxiv_id": "2509.09995", "doi": "10.48550/arXiv.2509.09995", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3014}
{"id": "e7cc006a814a5a6406ab68a3b0198f1be03dd6ffa4af563896b87584b8d269d4", "sources": ["arxiv", "semantic_scholar"], "title": "Bridging the Capability Gap: Joint Alignment Tuning for Harmonizing LLM-based Multi-Agent Systems", "abstract": "The advancement of large language models (LLMs) has enabled the construction of multi-agent systems to solve complex tasks by dividing responsibilities among specialized agents, such as a planning agent for subgoal generation and a grounding agent for executing tool-use actions. Most existing methods typically fine-tune these agents independently, leading to capability gaps among them with poor coordination. To address this, we propose MOAT, a Multi-Agent Joint Alignment Tuning framework that improves agents collaboration through iterative alignment. MOAT alternates between two key stages: (1) Planning Agent Alignment, which optimizes the planning agent to generate subgoal sequences that better guide the grounding agent; and (2) Grounding Agent Improving, which fine-tunes the grounding agent using diverse subgoal-action pairs generated by the agent itself to enhance its generalization capablity. Theoretical analysis proves that MOAT ensures a non-decreasing and progressively convergent training process. Experiments across six benchmarks demonstrate that MOAT outperforms state-of-the-art baselines, achieving average improvements of 3.1% on held-in tasks and 4.4% on held-out tasks.", "authors": ["Minghang Zhu", "Zhengliang Shi", "Zhiwei Xu", "Shiguang Wu", "Lingjie Wang", "Pengjie Ren", "Zhaochun Ren", "Zhumin Chen"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-11", "url": "https://arxiv.org/abs/2509.09629", "pdf_url": "https://arxiv.org/pdf/2509.09629v1", "arxiv_id": "2509.09629", "doi": "10.48550/arXiv.2509.09629", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3002}
{"id": "ac433e5de847259e7efcc09ff08b3c998f23663644f356bd842127b0d8482ec6", "sources": ["arxiv", "semantic_scholar"], "title": "Enabling Regulatory Multi-Agent Collaboration: Architecture, Challenges, and Solutions", "abstract": "Large language models (LLMs)-empowered autonomous agents are transforming both digital and physical environments by enabling adaptive, multi-agent collaboration. While these agents offer significant opportunities across domains such as finance, healthcare, and smart manufacturing, their unpredictable behaviors and heterogeneous capabilities pose substantial governance and accountability challenges. In this paper, we propose a blockchain-enabled layered architecture for regulatory agent collaboration, comprising an agent layer, a blockchain data layer, and a regulatory application layer. Within this framework, we design three key modules: (i) an agent behavior tracing and arbitration module for automated accountability, (ii) a dynamic reputation evaluation module for trust assessment in collaborative scenarios, and (iii) a malicious behavior forecasting module for early detection of adversarial activities. Our approach establishes a systematic foundation for trustworthy, resilient, and scalable regulatory mechanisms in large-scale agent ecosystems. Finally, we discuss the future research directions for blockchain-enabled regulatory frameworks in multi-agent systems.", "authors": ["Qinnan Hu", "Yuntao Wang", "Yuan Gao", "Zhou Su", "Linkang Du", "Qichao Xu"], "categories": ["cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-11", "url": "https://arxiv.org/abs/2509.09215", "pdf_url": "https://arxiv.org/pdf/2509.09215v2", "arxiv_id": "2509.09215", "doi": "10.48550/arXiv.2509.09215", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3002}
{"id": "10db4a8075dd499944bc1dcd352fcef8936e78ae53ebc565cd8aec8d964a72dd", "sources": ["arxiv", "semantic_scholar"], "title": "Strategic Tradeoffs Between Humans and AI in Multi-Agent Bargaining", "abstract": "Markets increasingly accommodate large language models (LLMs) as autonomous decision-making agents. As this transition occurs, it becomes critical to evaluate how these agents behave relative to their human and task-specific statistical predecessors. In this work, we present results from an empirical study comparing humans (N=216), multiple frontier LLMs, and customized Bayesian agents in dynamic multi-player bargaining games under identical conditions. Bayesian agents extract the highest surplus with aggressive trade proposals that are frequently rejected. Humans and LLMs achieve comparable aggregate surplus within their groups, but exhibit different trading strategies. LLMs favor conservative, concessionary proposals that are usually accepted by other LLMs, while humans propose trades that are consistent with fairness norms but are more likely to be rejected. These findings highlight that performance parity -- a common benchmark in agent evaluation -- can mask substantive procedural differences in how LLMs behave in complex multi-agent interactions.", "authors": ["Crystal Qian", "Kehang Zhu", "John Horton", "Benjamin S. Manning", "Vivian Tsai", "James Wexler", "Nithum Thain"], "categories": ["cs.AI", "cs.GT", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-11", "url": "https://arxiv.org/abs/2509.09071", "pdf_url": "https://arxiv.org/pdf/2509.09071v4", "arxiv_id": "2509.09071", "doi": "10.48550/arXiv.2509.09071", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Intelligent User Interfaces", "quality_score": 0.3002}
{"id": "17923ececb60fb466339cf19f7dc2713f086cf39bb43d87d642634e21533c2bc", "sources": ["arxiv", "semantic_scholar"], "title": "GeoJSON Agents:A Multi-Agent LLM Architecture for Geospatial Analysis-Function Calling vs Code Generation", "abstract": "Large Language Models (LLMs) have demonstrated substantial progress in task automation and natural language understanding. However, without domain expertise in geographic information science (GIS), they continue to encounter limitations including reduced accuracy and unstable performance when processing complex tasks. To address these challenges, we propose GeoJSON Agents-a novel multi-agent LLM architecture specifically designed for geospatial analysis. This framework transforms natural language instructions into structured GeoJSON operations through two LLM enhancement techniques: Function Calling and Code Generation. The architecture integrates three core components: task parsing, agent collaboration, and result integration. The Planner agent systematically decomposes user-defined tasks into executable subtasks, while Worker agents perform spatial data processing and analysis either by invoking predefined function APIs or by generating and executing Python-based analytical code. The system produces reusable, standards-compliant GeoJSON outputs through iterative refinement. To evaluate both approaches, we constructed a benchmark comprising 70 tasks spanning basic, intermediate, and advanced complexity levels, conducting experiments with OpenAI's GPT-4o as the core model. Results indicate that the Code Generation-based agent achieved 97.14% accuracy, while the Function Calling-based agent attained 85.71%-both significantly outperforming the best-performing general-purpose model (48.57%). Comparative analysis reveals Code Generation offers superior flexibility for complex, open-ended tasks, whereas Function Calling provides enhanced execution stability for structured operations. This study represents the first systematic integration of GeoJSON data with a multi-agent LLM framework and provides empirical evidence comparing two mainstream enhancement methodologies in geospatial context.", "authors": ["Qianqian Luo", "Qingming Lin", "Liuchang Xu", "Sensen Wu", "Ruichen Mao", "Chao Wang", "Hailin Feng", "Bo Huang", "Zhenhong Du"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-10", "url": "https://arxiv.org/abs/2509.08863", "pdf_url": "https://arxiv.org/pdf/2509.08863v3", "arxiv_id": "2509.08863", "doi": "10.48550/arXiv.2509.08863", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2991}
{"id": "27e4fcb9563dcdc31dbbe2dac9afc45e3d3fd83148619b49592a6729a0dcf543", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Human-AI Collaboration Using Mental Models of Early Adopters of Multi-Agent Generative AI Tools", "abstract": "With recent advancements in multi-agent generative AI (Gen AI), technology organizations like Microsoft are adopting these complex tools, redefining AI agents as active collaborators in complex workflows rather than as passive tools. In this study, we investigated how early adopters and developers conceptualize multi-agent Gen AI tools, focusing on how they understand human-AI collaboration mechanisms, general collaboration dynamics, and transparency in the context of AI tools. We conducted semi-structured interviews with 13 developers, all early adopters of multi-agent Gen AI technology who work at Microsoft. Our findings revealed that these early adopters conceptualize multi-agent systems as \"teams\" of specialized role-based and task-based agents, such as assistants or reviewers, structured similar to human collaboration models and ranging from AI-dominant to AI-assisted, user-controlled interactions. We identified key challenges, including error propagation, unpredictable and unproductive agent loop behavior, and the need for clear communication to mitigate the layered transparency issues. Early adopters' perspectives about the role of transparency underscored its importance as a way to build trust, verify and trace errors, and prevent misuse, errors, and leaks. The insights and design considerations we present contribute to CSCW research about collaborative mechanisms with capabilities ranging from AI-dominant to AI-assisted interactions, transparency and oversight strategies in human-agent and agent-agent interactions, and how humans make sense of these multi-agent systems as dynamic, role-diverse collaborators which are customizable for diverse needs and workflows. We conclude with future research directions that extend CSCW approaches to the design of inter-agent and human mediation interactions.", "authors": ["Suchismita Naik", "Austin L. Toombs", "Amanda Snellinger", "Scott Saponas", "Amanda K. Hall"], "categories": ["cs.HC", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-10", "url": "https://arxiv.org/abs/2510.06224", "pdf_url": "https://arxiv.org/pdf/2510.06224v1", "arxiv_id": "2510.06224", "doi": "10.48550/arXiv.2510.06224", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2991}
{"id": "6cfdcbe68d30e2faa5edf0769e145b177d4e72cbe6d781d13b4c332d3b2dde0c", "sources": ["arxiv", "semantic_scholar"], "title": "A Role-Aware Multi-Agent Framework for Financial Education Question Answering with LLMs", "abstract": "Question answering (QA) plays a central role in financial education, yet existing large language model (LLM) approaches often fail to capture the nuanced and specialized reasoning required for financial problem-solving. The financial domain demands multistep quantitative reasoning, familiarity with domain-specific terminology, and comprehension of real-world scenarios. We present a multi-agent framework that leverages role-based prompting to enhance performance on domain-specific QA. Our framework comprises a Base Generator, an Evidence Retriever, and an Expert Reviewer agent that work in a single-pass iteration to produce a refined answer. We evaluated our framework on a set of 3,532 expert-designed finance education questions from Study.com, an online learning platform. We leverage retrieval-augmented generation (RAG) for contextual evidence from 6 finance textbooks and prompting strategies for a domain-expert reviewer. Our experiments indicate that critique-based refinement improves answer accuracy by 6.6-8.3% over zero-shot Chain-of-Thought baselines, with the highest performance from Gemini-2.0-Flash. Furthermore, our method enables GPT-4o-mini to achieve performance comparable to the finance-tuned FinGPT-mt_Llama3-8B_LoRA. Our results show a cost-effective approach to enhancing financial QA and offer insights for further research in multi-agent financial LLM systems.", "authors": ["Andy Zhu", "Yingjun Du"], "categories": ["cs.CL", "cs.CE"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-10", "url": "https://arxiv.org/abs/2509.09727", "pdf_url": "https://arxiv.org/pdf/2509.09727v1", "arxiv_id": "2509.09727", "doi": "10.48550/arXiv.2509.09727", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2991}
{"id": "f8db82fba2d5eb0d4f0e4b52d2bb9b31491c524479ea902dc679465b4254cc0a", "sources": ["arxiv", "semantic_scholar"], "title": "ChemBOMAS: Accelerated BO in Chemistry with LLM-Enhanced Multi-Agent System", "abstract": "Bayesian optimization (BO) is a powerful tool for scientific discovery in chemistry, yet its efficiency is often hampered by the sparse experimental data and vast search space. Here, we introduce ChemBOMAS: a large language model (LLM)-enhanced multi-agent system that accelerates BO through synergistic data- and knowledge-driven strategies. Firstly, the data-driven strategy involves an 8B-scale LLM regressor fine-tuned on a mere 1% labeled samples for pseudo-data generation, robustly initializing the optimization process. Secondly, the knowledge-driven strategy employs a hybrid Retrieval-Augmented Generation approach to guide LLM in dividing the search space while mitigating LLM hallucinations. An Upper Confidence Bound algorithm then identifies high-potential subspaces within this established partition. Across the LLM-refined subspaces and supported by LLM-generated data, BO achieves the improvement of effectiveness and efficiency. Comprehensive evaluations across multiple scientific benchmarks demonstrate that ChemBOMAS set a new state-of-the-art, accelerating optimization efficiency by up to 5-fold compared to baseline methods.", "authors": ["Dong Han", "Zhehong Ai", "Pengxiang Cai", "Shanya Lu", "Jianpeng Chen", "Zihao Ye", "Shuzhou Sun", "Ben Gao", "Lingli Ge", "Weida Wang", "Xiangxin Zhou", "Xihui Liu", "Mao Su", "Wanli Ouyang", "Lei Bai", "Dongzhan Zhou", "Tao Xu", "Yuqiang Li", "Shufei Zhang"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-10", "url": "https://arxiv.org/abs/2509.08736", "pdf_url": "https://arxiv.org/pdf/2509.08736v2", "arxiv_id": "2509.08736", "doi": "10.48550/arXiv.2509.08736", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2991}
{"id": "9f2082367519df6172a0a3d68e23290d2f328bc472d02cd5d9ba39602cd5faf8", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling up Multi-Turn Off-Policy RL and Multi-Agent Tree Search for LLM Step-Provers", "abstract": "The integration of Large Language Models (LLMs) into automated theorem proving has shown immense promise, yet is fundamentally constrained by challenges in scaling up both training-time reinforcement learning (RL) and inference-time compute. This paper introduces \\texttt{BFS-Prover-V2}, a system designed to address this dual scaling problem. We present two primary innovations. The first is a novel multi-turn off-policy RL framework for continually improving the performance of LLM step-prover at training time. This framework, inspired by the principles of AlphaZero, utilizes a multi-stage expert iteration pipeline featuring adaptive tactic-level data filtering and periodic retraining to surmount the performance plateaus that typically curtail long-term RL in LLM-based agents. The second innovation is a planner-enhanced multi-agent search architecture that scales reasoning capabilities at inference time. This architecture employs a general reasoning model as a high-level planner to iteratively decompose complex theorems into a sequence of simpler subgoals. This hierarchical approach substantially reduces the search space, enabling a team of parallel prover agents to collaborate efficiently by leveraging a shared proof cache. We demonstrate that this dual approach to scaling yields state-of-the-art results on established formal mathematics benchmarks. \\texttt{BFS-Prover-V2} achieves 95.08\\% and 41.4\\% on the MiniF2F and ProofNet test sets respectively. While demonstrated in the domain of formal mathematics, the RL and inference techniques presented in this work are of broader interest and may be applied to other domains requiring long-horizon multi-turn reasoning and complex search.", "authors": ["Ran Xin", "Zeyu Zheng", "Yanchen Nie", "Kun Yuan", "Xia Xiao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-08", "url": "https://arxiv.org/abs/2509.06493", "pdf_url": "https://arxiv.org/pdf/2509.06493v2", "arxiv_id": "2509.06493", "doi": "10.48550/arXiv.2509.06493", "citation_count": 18, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "c9dfc3e64db744f78ad051ae7345b1874d0f3c409af7242203172f7478a20986", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating Multi-Turn Bargain Skills in LLM-Based Seller Agent", "abstract": "In online second-hand marketplaces, multi-turn bargaining is a crucial part of seller-buyer interactions. Large Language Models (LLMs) can act as seller agents, negotiating with buyers on behalf of sellers under given business constraints. A critical ability for such agents is to track and accurately interpret cumulative buyer intents across long negotiations, which directly impacts bargaining effectiveness. We introduce a multi-turn evaluation framework for measuring the bargaining ability of seller agents in e-commerce dialogues. The framework tests whether an agent can extract and track buyer intents. Our contributions are: (1) a large-scale e-commerce bargaining benchmark spanning 622 categories, 9,892 products, and 3,014 tasks; (2) a turn-level evaluation framework grounded in Theory of Mind (ToM) with annotated buyer intents, moving beyond outcome-only metrics; and (3) an automated pipeline that extracts reliable intent from massive dialogue data.", "authors": ["Issue Yishu Wang", "Kakam Chong", "Xiaofeng Wang", "Xu Yan", "DeXin Kong", "Chen Ju", "Ming Chen", "Shuai Xiao", "Shuguang Han", "jufeng chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-08", "url": "https://arxiv.org/abs/2509.06341", "pdf_url": "https://arxiv.org/pdf/2509.06341v1", "arxiv_id": "2509.06341", "doi": "10.48550/arXiv.2509.06341", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2968}
{"id": "72336a00e98704f121636e14e6221db69b36c1f8420ef5e79b5d8d1fc58edcd1", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborate, Deliberate, Evaluate: How LLM Alignment Affects Coordinated Multi-Agent Outcomes", "abstract": "As Large Language Models (LLMs) get integrated into diverse workflows, they are increasingly being regarded as \"collaborators\" with humans, and required to work in coordination with other AI systems. If such AI collaborators are to reliably coordinate their actions and behaviors with humans or other AIs, their properties and behaviors over multi-turn interactions must be known and predictable. This paper examines how different alignment methods affect LLM agents' effectiveness as partners in multi-turn, multi-party collaborations. We study this question through the lens of intervention agents that insert themselves into group dialogues not to provide answers, but to encourage the collaborative group to slow down and reflect upon their reasoning for deliberative decision-making. Common alignment techniques are typically developed under simplified single-user settings and assume the optimality of the underlying token MDP. Using the theoretical lens of the modified-action MDP, we show how they do not account for the dynamics of long-horizon multi-party interactions. We present a novel roleplay simulation methodology, where we align LLMs according to different methods and then deploy them in collaborative task dialogues to quantify how interventions affect the trajectory of group collaboration, belief alignment, and coordination. Our results show that an intervention agent that is robust to action modification significantly outperforms common alignment baselines in supporting correct task outcomes.", "authors": ["Abhijnan Nath", "Carine Graff", "Nikhil Krishnaswamy"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-07", "url": "https://arxiv.org/abs/2509.05882", "pdf_url": "https://arxiv.org/pdf/2509.05882v2", "arxiv_id": "2509.05882", "doi": "10.65109/uqpo8536", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "c1a11dc95aab94244169b68dd2b6b1e011862f27435de79879f4a69e917efb19", "sources": ["arxiv", "semantic_scholar"], "title": "MapAgent: A Hierarchical Agent for Geospatial Reasoning with Dynamic Map Tool Integration", "abstract": "Agentic AI has significantly extended the capabilities of large language models (LLMs) by enabling complex reasoning and tool use. However, most existing frameworks are tailored to domains such as mathematics, coding, or web automation, and fall short on geospatial tasks that require spatial reasoning, multi-hop planning, and real-time map interaction. To address these challenges, we introduce MapAgent, a hierarchical multi-agent plug-and-play framework with customized toolsets and agentic scaffolds for map-integrated geospatial reasoning. Unlike existing flat agent-based approaches that treat tools uniformly-often overwhelming the LLM when handling similar but subtly different geospatial APIs-MapAgent decouples planning from execution. A high-level planner decomposes complex queries into subgoals, which are routed to specialized modules. For tool-heavy modules-such as map-based services-we then design a dedicated map-tool agent that efficiently orchestrates related APIs adaptively in parallel to effectively fetch geospatial data relevant for the query, while simpler modules (e.g., solution generation or answer extraction) operate without additional agent overhead. This hierarchical design reduces cognitive load, improves tool selection accuracy, and enables precise coordination across similar APIs. We evaluate MapAgent on four diverse geospatial benchmarks-MapEval-Textual, MapEval-API, MapEval-Visual, and MapQA-and demonstrate substantial gains over state-of-the-art tool-augmented and agentic baselines. We open-source our framwork at https://github.com/Hasebul/MapAgent.", "authors": ["Md Hasebul Hasan", "Mahir Labib Dihan", "Tanzima Hashem", "Mohammed Eunus Ali", "Md Rizwan Parvez"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-07", "url": "https://arxiv.org/abs/2509.05933", "pdf_url": "https://arxiv.org/pdf/2509.05933v2", "arxiv_id": "2509.05933", "doi": "10.48550/arXiv.2509.05933", "citation_count": 4, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/Hasebul/MapAgent", "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.4569}
{"id": "733a8fea4a77c68eba42d646603646a2866c85e6175c44b71a6c45de80d8c1c4", "sources": ["arxiv", "semantic_scholar"], "title": "PillagerBench: Benchmarking LLM-Based Agents in Competitive Minecraft Team Environments", "abstract": "LLM-based agents have shown promise in various cooperative and strategic reasoning tasks, but their effectiveness in competitive multi-agent environments remains underexplored. To address this gap, we introduce PillagerBench, a novel framework for evaluating multi-agent systems in real-time competitive team-vs-team scenarios in Minecraft. It provides an extensible API, multi-round testing, and rule-based built-in opponents for fair, reproducible comparisons. We also propose TactiCrafter, an LLM-based multi-agent system that facilitates teamwork through human-readable tactics, learns causal dependencies, and adapts to opponent strategies. Our evaluation demonstrates that TactiCrafter outperforms baseline approaches and showcases adaptive learning through self-play. Additionally, we analyze its learning process and strategic evolution over multiple game episodes. To encourage further research, we have open-sourced PillagerBench, fostering advancements in multi-agent AI for competitive environments.", "authors": ["Olivier Schipper", "Yudi Zhang", "Yali Du", "Mykola Pechenizkiy", "Meng Fang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-07", "url": "https://arxiv.org/abs/2509.06235", "pdf_url": "https://arxiv.org/pdf/2509.06235v1", "arxiv_id": "2509.06235", "doi": "10.1109/CoG64752.2025.11114387", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/aialt/PillagerBench", "venue": "2025 IEEE Conference on Games (CoG), Lisbon, Portugal, 2025, pp. 1-15", "quality_score": 0.4569}
{"id": "6a70662947eb6bca90ca58730dfe86d3f8721f9a7952fd9e175aa19e3bad7738", "sources": ["arxiv", "semantic_scholar"], "title": "DRF: LLM-AGENT Dynamic Reputation Filtering Framework", "abstract": "With the evolution of generative AI, multi - agent systems leveraging large - language models(LLMs) have emerged as a powerful tool for complex tasks. However, these systems face challenges in quantifying agent performance and lack mechanisms to assess agent credibility. To address these issues, we introduce DRF, a dynamic reputation filtering framework. DRF constructs an interactive rating network to quantify agent performance, designs a reputation scoring mechanism to measure agent honesty and capability, and integrates an Upper Confidence Bound - based strategy to enhance agent selection efficiency. Experiments show that DRF significantly improves task completion quality and collaboration efficiency in logical reasoning and code - generation tasks, offering a new approach for multi - agent systems to handle large - scale tasks.", "authors": ["Yuwei Lou", "Hao Hu", "Shaocong Ma", "Zongfei Zhang", "Liang Wang", "Jidong Ge", "Xianping Tao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-06", "url": "https://arxiv.org/abs/2509.05764", "pdf_url": "https://arxiv.org/pdf/2509.05764v1", "arxiv_id": "2509.05764", "doi": "10.48550/arXiv.2509.05764", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Neural Information Processing", "quality_score": 0.2945}
{"id": "e82d81b9c13ababa519dfaeb786538014bad61690b2687e25bf4f7b0328b1ee2", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Enabled Multi-Agent System for 6G Networks: Framework and Method of Dual-Loop Edge-Terminal Collaboration", "abstract": "The ubiquitous computing resources in 6G networks provide ideal environments for the fusion of large language models (LLMs) and intelligent services through the agent framework. With auxiliary modules and planning cores, LLM-enabled agents can autonomously plan and take actions to deal with diverse environment semantics and user intentions. However, the limited resources of individual network devices significantly hinder the efficient operation of LLM-enabled agents with complex tool calls, highlighting the urgent need for efficient multi-level device collaborations. To this end, the framework and method of the LLM-enabled multi-agent system with dual-loop terminal-edge collaborations are proposed in 6G networks. Firstly, the outer loop consists of the iterative collaborations between the global agent and multiple sub-agents deployed on edge servers and terminals, where the planning capability is enhanced through task decomposition and parallel sub-task distribution. Secondly, the inner loop utilizes sub-agents with dedicated roles to circularly reason, execute, and replan the sub-task, and the parallel tool calling generation with offloading strategies is incorporated to improve efficiency. The improved task planning capability and task execution efficiency are validated through the conducted case study in 6G-supported urban safety governance. Finally, the open challenges and future directions are thoroughly analyzed in 6G networks, accelerating the advent of the 6G era.", "authors": ["Zheyan Qu", "Wenbo Wang", "Zitong Yu", "Boquan Sun", "Yang Li", "Xing Zhang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-05", "url": "https://arxiv.org/abs/2509.04993", "pdf_url": "https://arxiv.org/pdf/2509.04993v1", "arxiv_id": "2509.04993", "doi": "10.1109/MCOM.001.2500148", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Communications Magazine", "quality_score": 0.2933}
{"id": "ceab9a5a7a55506141e9e800da961c2d69ecee635f8d7aa96efaca397cee653a", "sources": ["arxiv", "semantic_scholar"], "title": "OSC: Cognitive Orchestration through Dynamic Knowledge Alignment in Multi-Agent LLM Collaboration", "abstract": "This paper introduces OSC (Orchestrating Cognitive Synergy), a knowledge-aware adaptive collaboration framework designed to enhance cognitive synergy in multi-agent systems with large language models. While prior work has advanced agent selection and result aggregation, efficient linguistic interactions for deep collaboration among expert agents remain a critical bottleneck. OSC addresses this gap as a pivotal intermediate layer between selection and aggregation, introducing Collaborator Knowledge Models (CKM) to enable each agent to dynamically perceive its collaborators' cognitive states. Through real-time cognitive gap analysis, agents adaptively adjust communication behaviors, including content focus, detail level, and expression style, using learned strategies. Experiments on complex reasoning and problem-solving benchmarks demonstrate that OSC significantly improves task performance and communication efficiency, transforming \"parallel-working individuals'' into a \"deeply collaborative cognitive team.'' This framework not only optimizes multi-agent collaboration but also offers new insights into LLM agent interaction behaviors.", "authors": ["Jusheng Zhang", "Yijia Fan", "Kaitong Cai", "Xiaofei Sun", "Keze Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-05", "url": "https://arxiv.org/abs/2509.04876", "pdf_url": "https://arxiv.org/pdf/2509.04876v1", "arxiv_id": "2509.04876", "doi": "10.48550/arXiv.2509.04876", "citation_count": 34, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.386}
{"id": "5fe6f9f21c68e6e257c01de700744a501092c05432fe745a4f83ecae5e963e06", "sources": ["arxiv", "semantic_scholar"], "title": "Learning to Deliberate: Meta-policy Collaboration for Agentic LLMs with Multi-agent Reinforcement Learning", "abstract": "Multi-agent systems of large language models (LLMs) show promise for complex reasoning, but their effectiveness is often limited by fixed collaboration protocols. These frameworks typically focus on macro-level orchestration while overlooking agents' internal deliberative capabilities. This critical meta-cognitive blindspot treats agents as passive executors unable to adapt their strategy based on internal cognitive states like uncertainty or confidence. We introduce the Meta-Policy Deliberation Framework (MPDF), where agents learn a decentralized policy over a set of high-level meta-cognitive actions: Persist, Refine, and Concede. To overcome the instability of traditional policy gradients in this setting, we develop SoftRankPO, a novel reinforcement learning algorithm. SoftRankPO stabilizes training by shaping advantages based on the rank of rewards mapped through smooth normal quantiles, making the learning process robust to reward variance. Experiments show that MPDF with SoftRankPO achieves a a 4-5% absolute gain in average accuracy across five mathematical and general reasoning benchmarks compared to six state-of-the-art heuristic and learning-based multi-agent reasoning algorithms. Our work presents a paradigm for learning adaptive, meta-cognitive policies for multi-agent LLM systems, shifting the focus from designing fixed protocols to learning dynamic, deliberative strategies.", "authors": ["Wei Yang", "Jesse Thomason"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-04", "url": "https://arxiv.org/abs/2509.03817", "pdf_url": "https://arxiv.org/pdf/2509.03817v2", "arxiv_id": "2509.03817", "doi": "10.48550/arXiv.2509.03817", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.2922}
{"id": "bbef5f8bf52b248b7c1d810707ee89c28c89c9372c936a85a7d2ff0c3db38ce4", "sources": ["arxiv", "semantic_scholar"], "title": "AgenTracer: Who Is Inducing Failure in the LLM Agentic Systems?", "abstract": "Large Language Model (LLM)-based agentic systems, often comprising multiple models, complex tool invocations, and orchestration protocols, substantially outperform monolithic agents. Yet this very sophistication amplifies their fragility, making them more prone to system failure. Pinpointing the specific agent or step responsible for an error within long execution traces defines the task of agentic system failure attribution. Current state-of-the-art reasoning LLMs, however, remain strikingly inadequate for this challenge, with accuracy generally below 10%. To address this gap, we propose AgenTracer, the first automated framework for annotating failed multi-agent trajectories via counterfactual replay and programmed fault injection, producing the curated dataset TracerTraj. Leveraging this resource, we develop AgenTracer-8B, a lightweight failure tracer trained with multi-granular reinforcement learning, capable of efficiently diagnosing errors in verbose multi-agent interactions. On the Who&When benchmark, AgenTracer-8B outperforms giant proprietary LLMs like Gemini-2.5-Pro and Claude-4-Sonnet by up to 18.18%, setting a new standard in LLM agentic failure attribution. More importantly, AgenTracer-8B delivers actionable feedback to off-the-shelf multi-agent systems like MetaGPT and MaAS with 4.8-14.2% performance gains, empowering self-correcting and self-evolving agentic AI.", "authors": ["Guibin Zhang", "Junhao Wang", "Junjie Chen", "Wangchunshu Zhou", "Kun Wang", "Shuicheng Yan"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-03", "url": "https://arxiv.org/abs/2509.03312", "pdf_url": "https://arxiv.org/pdf/2509.03312v2", "arxiv_id": "2509.03312", "doi": "10.48550/arXiv.2509.03312", "citation_count": 51, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.429}
{"id": "16f7cf92f968a4a6613e186de997da756e5debe6f4d6b4e4cb3482449fc5af67", "sources": ["arxiv", "semantic_scholar"], "title": "Contemporary Agent Technology: LLM-Driven Advancements vs Classic Multi-Agent Systems", "abstract": "This contribution provides our comprehensive reflection on the contemporary agent technology, with a particular focus on the advancements driven by Large Language Models (LLM) vs classic Multi-Agent Systems (MAS). It delves into the models, approaches, and characteristics that define these new systems. The paper emphasizes the critical analysis of how the recent developments relate to the foundational MAS, as articulated in the core academic literature. Finally, it identifies key challenges and promising future directions in this rapidly evolving domain.", "authors": ["Costin Bădică", "Amelia Bădică", "Maria Ganzha", "Mirjana Ivanović", "Marcin Paprzycki", "Dan Selişteanu", "Zofia Wrona"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-02", "url": "https://arxiv.org/abs/2509.02515", "pdf_url": "https://arxiv.org/pdf/2509.02515v1", "arxiv_id": "2509.02515", "doi": "10.48550/arXiv.2509.02515", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2899}
{"id": "d344a70e3b1b58f2603eadc84a7c66f82ee050a33e5fe44603bf097b384158e0", "sources": ["arxiv", "semantic_scholar"], "title": "ProST: Progressive Sub-task Training for Pareto-Optimal Multi-agent Systems Using Small Language Models", "abstract": "Multi-agent systems with smaller language models (SLMs) present a viable alternative to single agent systems powered by large language models (LLMs) for addressing complex problems. In this work, we study how these alternatives compare in terms of both effectiveness and efficiency. To study this trade-off, we instantiate single and multi-agent systems for the complex problems in the AppWorld environment using different sized language models. We find that difficulties with long-trajectory learning in smaller language models (SLMs) limit their performance. Even when trained for specialized roles, SLMs fail to learn all subtasks effectively. To address this issue, we introduce a simple progressive sub-task training strategy, which introduces new sub-tasks progressively in each training epoch. We find that this novel strategy, analogous to instance level curriculum learning, consistently improves the effectiveness of multi-agents at all configurations. Our Pareto analysis shows that fine-tuned multi-agent systems yield better effectiveness-efficiency trade-offs. Additional ablations and analyses shows the importance of our progressive training strategy and its ability to reduce subtask error rates.", "authors": ["Biddut Sarker Bijoy", "Mohammad Saqib Hasan", "Pegah Alipoormolabashi", "Avirup Sil", "Aruna Balasubramanian", "Niranjan Balasubramanian"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-09-02", "url": "https://arxiv.org/abs/2509.04508", "pdf_url": "https://arxiv.org/pdf/2509.04508v2", "arxiv_id": "2509.04508", "doi": "10.48550/arXiv.2509.04508", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1845}
{"id": "e6a75efa46ac3c24793dd98c789a6293c4d087df3fea830c5e1de5c21e0e5380", "sources": ["arxiv", "semantic_scholar"], "title": "CoComposer: LLM Multi-agent Collaborative Music Composition", "abstract": "Existing AI Music composition tools are limited in generation duration, musical quality, and controllability. We introduce CoComposer, a multi-agent system that consists of five collaborating agents, each with a task based on the traditional music composition workflow. Using the AudioBox-Aesthetics system, we experimentally evaluate CoComposer on four compositional criteria. We test with three LLMs (GPT-4o, DeepSeek-V3-0324, Gemini-2.5-Flash), and find (1) that CoComposer outperforms existing multi-agent LLM-based systems in music quality, and (2) compared to a single-agent system, in production complexity. Compared to non- LLM MusicLM, CoComposer has better interpretability and editability, although MusicLM still produces better music.", "authors": ["Peiwen Xing", "Aske Plaat", "Niki van Stein"], "categories": ["cs.SD", "cs.AI", "cs.MM", "eess.AS"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-08-29", "url": "https://arxiv.org/abs/2509.00132", "pdf_url": "https://arxiv.org/pdf/2509.00132v1", "arxiv_id": "2509.00132", "doi": "10.48550/arXiv.2509.00132", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2853}
{"id": "53b0fd251052ed2248a3a681506aa9e1f616217c8a4e4c9eaf9d02d65ed789f1", "sources": ["arxiv", "semantic_scholar"], "title": "Automated Clinical Problem Detection from SOAP Notes using a Collaborative Multi-Agent LLM Architecture", "abstract": "Accurate interpretation of clinical narratives is critical for patient care, but the complexity of these notes makes automation challenging. While Large Language Models (LLMs) show promise, single-model approaches can lack the robustness required for high-stakes clinical tasks. We introduce a collaborative multi-agent system (MAS) that models a clinical consultation team to address this gap. The system is tasked with identifying clinical problems by analyzing only the Subjective (S) and Objective (O) sections of SOAP notes, simulating the diagnostic reasoning process of synthesizing raw data into an assessment. A Manager agent orchestrates a dynamically assigned team of specialist agents who engage in a hierarchical, iterative debate to reach a consensus. We evaluated our MAS against a single-agent baseline on a curated dataset of 420 MIMIC-III notes. The dynamic multi-agent configuration demonstrated consistently improved performance in identifying congestive heart failure, acute kidney injury, and sepsis. Qualitative analysis of the agent debates reveals that this structure effectively surfaces and weighs conflicting evidence, though it can occasionally be susceptible to groupthink. By modeling a clinical team's reasoning process, our system offers a promising path toward more accurate, robust, and interpretable clinical decision support tools.", "authors": ["Yeawon Lee", "Xiaoyang Wang", "Christopher C. Yang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-29", "url": "https://arxiv.org/abs/2508.21803", "pdf_url": "https://arxiv.org/pdf/2508.21803v1", "arxiv_id": "2508.21803", "doi": "10.1145/3765612.3767792", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "ACM International Conference on Bioinformatics, Computational Biology and Biomedicine", "quality_score": 0.2853}
{"id": "bd92fab642937d7bc9b023c7ee1eb4ca9518f27dc54d6c646bfbee6531b304a4", "sources": ["arxiv", "semantic_scholar"], "title": "MCP-Bench: Benchmarking Tool-Using LLM Agents with Complex Real-World Tasks via MCP Servers", "abstract": "We introduce MCP-Bench, a benchmark for evaluating large language models (LLMs) on realistic, multi-step tasks that demand tool use, cross-tool coordination, precise parameter control, and planning/reasoning for solving tasks. Built on the Model Context Protocol (MCP), MCP-Bench connects LLMs to 28 representative live MCP servers spanning 250 tools across domains such as finance, traveling, scientific computing, and academic search. Unlike prior API-based benchmarks, each MCP server provides a set of complementary tools designed to work together, enabling the construction of authentic, multi-step tasks with rich input-output coupling. Tasks in MCP-Bench test agents' ability to retrieve relevant tools from fuzzy instructions without explicit tool names, plan multi-hop execution trajectories for complex objectives, ground responses in intermediate tool outputs, and orchestrate cross-domain workflows - capabilities not adequately evaluated by existing benchmarks that rely on explicit tool specifications, shallow few-step workflows, and isolated domain operations. We propose a multi-faceted evaluation framework covering tool-level schema understanding and usage, trajectory-level planning, and task completion. Experiments on 20 advanced LLMs reveal persistent challenges in MCP-Bench. Code and data: https://github.com/Accenture/mcp-bench.", "authors": ["Zhenting Wang", "Qi Chang", "Hemani Patel", "Shashank Biju", "Cheng-En Wu", "Quan Liu", "Aolin Ding", "Alireza Rezazadeh", "Ankit Shah", "Yujia Bao", "Eugene Siow"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-28", "url": "https://arxiv.org/abs/2508.20453", "pdf_url": "https://arxiv.org/pdf/2508.20453v1", "arxiv_id": "2508.20453", "doi": "10.48550/arXiv.2508.20453", "citation_count": 69, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/Accenture/mcp-bench", "venue": "arXiv.org", "quality_score": 0.4613}
{"id": "a64de35eca103198f791166b51c41d555911fdb6d705d1427af9e5dc0e6ccccb", "sources": ["arxiv", "semantic_scholar"], "title": "CyberSleuth: Autonomous Blue-Team LLM Agent for Web Attack Forensics", "abstract": "Post-mortem analysis of compromised systems is a key aspect of cyber forensics, today a mostly manual, slow, and error-prone task. Agentic AI, i.e., LLM-powered agents, is a promising avenue for automation. However, applying such agents to cybersecurity remains largely unexplored and difficult, as this domain demands long-term reasoning, contextual memory, and consistent evidence correlation - capabilities that current LLM agents struggle to master. In this paper, we present the first systematic study of LLM agents to automate post-mortem investigation. As a first scenario, we consider realistic attacks in which remote attackers try to abuse online services using well-known CVEs (30 controlled cases). The agent receives as input the network traces of the attack and extracts forensic evidence. We compare three AI agent architectures, six LLM backends, and assess their ability to i) identify compromised services, ii) map exploits to exact CVEs, and iii) prepare thorough reports. Our best-performing system, CyberSleuth, achieves 80% accuracy on 2025 incidents, producing complete, coherent, and practically useful reports (judged by a panel of 25 experts). We next illustrate how readily CyberSleuth adapts to face the analysis of infected machine traffic, showing that the effective AI agent design can transfer across forensic tasks. Our findings show that (i) multi-agent specialisation is key to sustained reasoning; (ii) simple orchestration outperforms nested hierarchical architectures; and (iii) the CyberSleuth design generalises across different forensic tasks.", "authors": ["Stefano Fumero", "Kai Huang", "Matteo Boffa", "Danilo Giordano", "Marco Mellia", "Dario Rossi"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-28", "url": "https://arxiv.org/abs/2508.20643", "pdf_url": "https://arxiv.org/pdf/2508.20643v2", "arxiv_id": "2508.20643", "doi": "10.48550/arXiv.2508.20643", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2842}
{"id": "d49b51794e055afdaa957a2bd19c02a1d4bbc8572012adcd4fbd737c60bbe989", "sources": ["arxiv", "semantic_scholar"], "title": "Your AI Bosses Are Still Prejudiced: The Emergence of Stereotypes in LLM-Based Multi-Agent Systems", "abstract": "While stereotypes are well-documented in human social interactions, AI systems are often presumed to be less susceptible to such biases. Previous studies have focused on biases inherited from training data, but whether stereotypes can emerge spontaneously in AI agent interactions merits further exploration. Through a novel experimental framework simulating workplace interactions with neutral initial conditions, we investigate the emergence and evolution of stereotypes in LLM-based multi-agent systems. Our findings reveal that (1) LLM-Based AI agents develop stereotype-driven biases in their interactions despite beginning without predefined biases; (2) stereotype effects intensify with increased interaction rounds and decision-making power, particularly after introducing hierarchical structures; (3) these systems exhibit group effects analogous to human social behavior, including halo effects, confirmation bias, and role congruity; and (4) these stereotype patterns manifest consistently across different LLM architectures. Through comprehensive quantitative analysis, these findings suggest that stereotype formation in AI systems may arise as an emergent property of multi-agent interactions, rather than merely from training data biases. Our work underscores the need for future research to explore the underlying mechanisms of this phenomenon and develop strategies to mitigate its ethical impacts.", "authors": ["Jingyu Guo", "Yingying Xu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-27", "url": "https://arxiv.org/abs/2508.19919", "pdf_url": "https://arxiv.org/pdf/2508.19919v2", "arxiv_id": "2508.19919", "doi": "10.48550/arXiv.2508.19919", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.283}
{"id": "1371a02f25f3496fc799ab5bd52cae168c46270a1c281add4f2a7a4aa8e7dca4", "sources": ["arxiv", "semantic_scholar"], "title": "MUA-RL: Multi-turn User-interacting Agent Reinforcement Learning for agentic tool use", "abstract": "With the recent rapid advancement of Agentic Intelligence, agentic tool use in LLMs has become increasingly important. During multi-turn interactions between agents and users, the dynamic, uncertain, and stochastic nature of user demands poses significant challenges to the agent's tool invocation capabilities. Agents are no longer expected to simply call tools to deliver a result; rather, they must iteratively refine their understanding of user needs through communication while simultaneously invoking tools to resolve user queries. Existing reinforcement learning (RL) approaches for tool use lack the integration of genuinely dynamic users during the RL training process. To bridge this gap, we introduce MUA-RL (Multi-turn User-interacting Agent Reinforcement Learning for agentic tool use), a novel reinforcement learning framework that, for the first time in the field of agentic tool use, integrates LLM-simulated users into the reinforcement learning loop. MUA-RL aims to enable autonomous learning of models to communicate with users efficiently and use various tools to solve practical problems in dynamic multi-turn interactions. Evaluations are done on several multi-turn tool-using benchmarks (see Figure 1). Specifically, MUA-RL-32B achieves 67.3 on TAU2 Retail, 45.4 on TAU2 Airline, 28.3 on TAU2 Telecom, 28.4 on BFCL-V3 Multi Turn, and 82.5 on ACEBench Agent -- outperforming or matching the performance of larger open-source models such as DeepSeek-V3-0324 and Qwen3-235B-A22B in non-thinking settings.", "authors": ["Weikang Zhao", "Xili Wang", "Chengdi Ma", "Lingbin Kong", "Zhaohua Yang", "Mingxiang Tuo", "Xiaowei Shi", "Yitao Zhai", "Xunliang Cai"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-26", "url": "https://arxiv.org/abs/2508.18669", "pdf_url": "https://arxiv.org/pdf/2508.18669v1", "arxiv_id": "2508.18669", "doi": "10.48550/arXiv.2508.18669", "citation_count": 21, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4356}
{"id": "2e400529f927e628524a88973d7cdfc43e01dfea338af273891c164117244a40", "sources": ["arxiv", "semantic_scholar"], "title": "QAgent: An LLM-based Multi-Agent System for Autonomous OpenQASM programming", "abstract": "Noisy Intermediate-Scale Quantum (NISQ) devices have begun to exhibit early quantum advantages on classically intractable problems, spanning physics simulations to Gaussian boson sampling. Yet, realizing these benefits remains challenging for non-experts, primarily due to the complexities of programming in Open Quantum Assembly Language (OpenQASM). Although Large Language Model (LLM)-based agents have shown promise in automating classical programming workflows, their quantum counterparts have largely been restricted to specialized tasks such as quantum chemistry or error correction. In this paper, we present QAgent, an LLM-powered multi-agent system that fully automates OpenQASM programming. By integrating task planning, in-context few-shot learning, retrieval-augmented generation (RAG) for long-term context, predefined generation tools, and chain-of-thought (CoT) reasoning, the agents systematically improve both compilation and functional correctness. Our evaluations demonstrate substantial improvements: across multiple LLMs of varying sizes, QAgent enhances the accuracy of QASM code generation by 71.6\\% compared to previous static LLM-based approaches. We envision this multi-agent system as a key enabler for democratizing quantum programming, bridging expertise gaps, and accelerating the practical adoption of quantum computing.", "authors": ["Zhenxiao Fu", "Fan Chen", "Lei Jiang"], "categories": ["cs.AI", "cs.ET", "quant-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-08-26", "url": "https://arxiv.org/abs/2508.20134", "pdf_url": "https://arxiv.org/pdf/2508.20134v1", "arxiv_id": "2508.20134", "doi": "10.48550/arXiv.2508.20134", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2819}
{"id": "f7f05d7d7df9c6e141485ce728c03a2cf2ef393f660c65898f56fc170f2211be", "sources": ["arxiv", "semantic_scholar"], "title": "CausalMACE: Causality Empowered Multi-Agents in Minecraft Cooperative Tasks", "abstract": "Minecraft, as an open-world virtual interactive environment, has become a prominent platform for research on agent decision-making and execution. Existing works primarily adopt a single Large Language Model (LLM) agent to complete various in-game tasks. However, for complex tasks requiring lengthy sequences of actions, single-agent approaches often face challenges related to inefficiency and limited fault tolerance. Despite these issues, research on multi-agent collaboration remains scarce. In this paper, we propose CausalMACE, a holistic causality planning framework designed to enhance multi-agent systems, in which we incorporate causality to manage dependencies among subtasks. Technically, our proposed framework introduces two modules: an overarching task graph for global task planning and a causality-based module for dependency management, where inherent rules are adopted to perform causal intervention. Experimental results demonstrate our approach achieves state-of-the-art performance in multi-agent cooperative tasks of Minecraft.", "authors": ["Qi Chai", "Zhang Zheng", "Junlong Ren", "Deheng Ye", "Zichuan Lin", "Hao Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-26", "url": "https://arxiv.org/abs/2508.18797", "pdf_url": "https://arxiv.org/pdf/2508.18797v1", "arxiv_id": "2508.18797", "doi": "10.48550/arXiv.2508.18797", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.2819}
{"id": "1ceeeb18f811718abcf2d2ac554ab87b4a8a2b67a56bf20b81e095370354082a", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based Agentic Reasoning Frameworks: A Survey from Methods to Scenarios", "abstract": "Recent advances in the intrinsic reasoning capabilities of large language models (LLMs) have given rise to LLM-based agent systems that exhibit near-human performance on a variety of automated tasks. However, although these systems share similarities in terms of their use of LLMs, different reasoning frameworks of the agent system steer and organize the reasoning process in different ways. In this survey, we propose a systematic taxonomy that decomposes agentic reasoning frameworks and analyze how these frameworks dominate framework-level reasoning by comparing their applications across different scenarios. Specifically, we propose an unified formal language to further classify agentic reasoning systems into single-agent methods, tool-based methods, and multi-agent methods. After that, we provide a comprehensive review of their key application scenarios in scientific discovery, healthcare, software engineering, social simulation, and economics. We also analyze the characteristic features of each framework and summarize different evaluation strategies. Our survey aims to provide the research community with a panoramic view to facilitate understanding of the strengths, suitable scenarios, and evaluation practices of different agentic reasoning frameworks.", "authors": ["Bingxi Zhao", "Lin Geng Foo", "Ping Hu", "Christian Theobalt", "Hossein Rahmani", "Jun Liu"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-25", "url": "https://arxiv.org/abs/2508.17692", "pdf_url": "https://arxiv.org/pdf/2508.17692v1", "arxiv_id": "2508.17692", "doi": "10.48550/arXiv.2508.17692", "citation_count": 28, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3656}
{"id": "a33d1f3635bb353899a97b0470a255c21e22e4a7fa06b5d30fbe948f21b6311c", "sources": ["arxiv", "semantic_scholar"], "title": "RepoTransAgent: Multi-Agent LLM Framework for Repository-Aware Code Translation", "abstract": "Repository-aware code translation is critical for modernizing legacy systems, enhancing maintainability, and enabling interoperability across diverse programming languages. While recent advances in large language models (LLMs) have improved code translation quality, existing approaches face significant challenges in practical scenarios: insufficient contextual understanding, inflexible prompt designs, and inadequate error correction mechanisms. These limitations severely hinder accurate and efficient translation of complex, real-world code repositories. To address these challenges, we propose RepoTransAgent, a novel multi-agent LLM framework for repository-aware code translation. RepoTransAgent systematically decomposes the translation process into specialized subtasks-context retrieval, dynamic prompt construction, and iterative code refinement-each handled by dedicated agents. Our approach leverages retrieval-augmented generation (RAG) for contextual information gathering, employs adaptive prompts tailored to varying repository scenarios, and introduces a reflection-based mechanism for systematic error correction. We evaluate RepoTransAgent on hundreds of Java-C# translation pairs from six popular open-source projects. Experimental results demonstrate that RepoTransAgent significantly outperforms state-of-the-art baselines in both compile and pass rates. Specifically, RepoTransAgent achieves up to 55.34% compile rate and 45.84% pass rate. Comprehensive analysis confirms the robustness and generalizability of RepoTransAgent across different LLMs, establishing its effectiveness for real-world repository-aware code translation.", "authors": ["Ziqi Guan", "Xin Yin", "Zhiyuan Peng", "Chao Ni"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-25", "url": "https://arxiv.org/abs/2508.17720", "pdf_url": "https://arxiv.org/pdf/2508.17720v1", "arxiv_id": "2508.17720", "doi": "10.48550/arXiv.2508.17720", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4339}
{"id": "4393f30a7d04d5dfc1c35686648db76f08c3b512ce625f0417b561e1e8b2601f", "sources": ["arxiv", "semantic_scholar"], "title": "PosterGen: Aesthetic-Aware Multi-Modal Paper-to-Poster Generation via Multi-Agent LLMs", "abstract": "Multi-agent systems built upon large language models (LLMs) have demonstrated remarkable capabilities in tackling complex compositional tasks. In this work, we apply this paradigm to the paper-to-poster generation problem, a practical yet time-consuming process faced by researchers preparing for conferences. While recent approaches have attempted to automate this task, most neglect core design and aesthetic principles, resulting in posters that require substantial manual refinement. To address these design limitations, we propose PosterGen, a multi-agent framework that mirrors the workflow of professional poster designers. It consists of four collaborative specialized agents: (1) Parser and Curator agents extract content from the paper and organize storyboard; (2) Layout agent maps the content into a coherent spatial layout; (3) Stylist agents apply visual design elements such as color and typography; and (4) Renderer composes the final poster. Together, these agents produce posters that are both semantically grounded and visually appealing. To evaluate design quality, we introduce a vision-language model (VLM)-based rubric that measures layout balance, readability, and aesthetic coherence. Experimental results show that PosterGen consistently matches in content fidelity, and significantly outperforms existing methods in visual designs, generating posters that are presentation-ready with minimal human refinements.", "authors": ["Zhilin Zhang", "Xiang Zhang", "Jiaqi Wei", "Yiwei Xu", "Chenyu You"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-24", "url": "https://arxiv.org/abs/2508.17188", "pdf_url": "https://arxiv.org/pdf/2508.17188v2", "arxiv_id": "2508.17188", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1779}
{"id": "a38befae80094424b27d887a2cbd76ea68015dbcf85954c259a3287c54e569f7", "sources": ["arxiv", "semantic_scholar"], "title": "From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users", "abstract": "The pursuit of human-level artificial intelligence (AI) has significantly advanced the development of autonomous agents and Large Language Models (LLMs). LLMs are now widely utilized as decision-making agents for their ability to interpret instructions, manage sequential tasks, and adapt through feedback. This review examines recent developments in employing LLMs as autonomous agents and tool users and comprises seven research questions. We only used the papers published between 2023 and 2025 in conferences of the A* and A rank and Q1 journals. A structured analysis of the LLM agents' architectural design principles, dividing their applications into single-agent and multi-agent systems, and strategies for integrating external tools is presented. In addition, the cognitive mechanisms of LLM, including reasoning, planning, and memory, and the impact of prompting methods and fine-tuning procedures on agent performance are also investigated. Furthermore, we evaluated current benchmarks and assessment protocols and have provided an analysis of 68 publicly available datasets to assess the performance of LLM-based agents in various tasks. In conducting this review, we have identified critical findings on verifiable reasoning of LLMs, the capacity for self-improvement, and the personalization of LLM-based agents. Finally, we have discussed ten future research directions to overcome these gaps.", "authors": ["Sadia Sultana Chowa", "Riasad Alvi", "Subhey Sadi Rahman", "Md Abdur Rahman", "Mohaimenul Azam Khan Raiaan", "Md Rafiqul Islam", "Mukhtar Hussain", "Sami Azam"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-24", "url": "https://arxiv.org/abs/2508.17281", "pdf_url": "https://arxiv.org/pdf/2508.17281v2", "arxiv_id": "2508.17281", "doi": "10.1007/s10462-025-11471-9", "citation_count": 44, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Artificial Intelligence Review", "quality_score": 0.4133}
{"id": "25532c953af156098ef19d85f9ae372e8e8caa9fb233c1dd68a14fb742a34c8a", "sources": ["arxiv", "semantic_scholar"], "title": "Anemoi: A Semi-Centralized Multi-agent System Based on Agent-to-Agent Communication MCP server from Coral Protocol", "abstract": "Recent advances in generalist multi-agent systems (MAS) have largely followed a context-engineering plus centralized paradigm, where a planner agent coordinates multiple worker agents through unidirectional prompt passing. While effective under strong planner models, this design suffers from two critical limitations: (1) strong dependency on the planner's capability, which leads to degraded performance when a smaller LLM powers the planner; and (2) limited inter-agent communication, where collaboration relies on prompt concatenation rather than genuine refinement through structured discussions. To address these challenges, we propose Anemoi, a semi-centralized MAS built on the Agent-to-Agent (A2A) communication MCP server from Coral Protocol. Unlike traditional designs, Anemoi enables structured and direct inter-agent collaboration, allowing all agents to monitor progress, assess results, identify bottlenecks, and propose refinements in real time. This paradigm reduces reliance on a single planner, supports adaptive plan updates, and minimizes redundant context passing, resulting in more scalable execution. Evaluated on the GAIA benchmark, Anemoi achieved 52.73% accuracy with a small LLM (GPT-4.1-mini) as the planner, surpassing the strongest open-source baseline OWL (43.63%) by +9.09% under identical LLM settings. Our implementation is publicly available at https://github.com/Coral-Protocol/Anemoi.", "authors": ["Xinxing Ren", "Caelum Forder", "Qianbo Zang", "Ahsen Tahir", "Roman J. Georgio", "Suman Deb", "Peter Carroll", "Önder Gürcan", "Zekun Guo"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-23", "url": "https://arxiv.org/abs/2508.17068", "pdf_url": "https://arxiv.org/pdf/2508.17068v3", "arxiv_id": "2508.17068", "doi": "10.48550/arXiv.2508.17068", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Coral-Protocol/Anemoi", "venue": "arXiv.org", "quality_score": 0.4303}
{"id": "2fbededf51c7974eabef661b828fa9d14f3a9908903edca4ade219c8d242ce41", "sources": ["arxiv", "semantic_scholar"], "title": "ASIC-Agent: An Autonomous Multi-Agent System for ASIC Design with Benchmark Evaluation", "abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities in Register Transfer Level (RTL) design, enabling high-quality code generation from natural language descriptions. However, LLMs alone face significant limitations in real-world hardware design workflows, including the inability to execute code, lack of debugging capabilities, and absence of long-term memory. To address these challenges, we present ASIC-Agent, an autonomous system designed specifically for digital ASIC design tasks. ASIC-Agent enhances base LLMs with a multi-agent architecture incorporating specialized sub-agents for RTL generation, verification, OpenLane hardening, and Caravel chip integration, all operating within a comprehensive sandbox environment with access to essential hardware design tools. The system leverages a vector database containing documentation, API references, error knowledge, and curated insights from the open-source silicon community. To evaluate ASIC-Agent's performance, we introduce ASIC-Agent-Bench, the first benchmark specifically designed to assess agentic systems in hardware design tasks. We evaluate ASIC-Agent with various base LLMs, providing quantitative comparisons and qualitative insights into agent behavior across different design scenarios. Our results demonstrate that ASIC-Agent, when powered by Claude 4 Sonnet, successfully automates a broad range of ASIC design tasks spanning varying levels of complexity, showing the potential of significantly accelerating the ASIC design workflow.", "authors": ["Ahmed Allam", "Youssef Mansour", "Mohamed Shalan"], "categories": ["cs.AR", "cs.AI", "cs.CL", "cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-21", "url": "https://arxiv.org/abs/2508.15940", "pdf_url": "https://arxiv.org/pdf/2508.15940v1", "arxiv_id": "2508.15940", "doi": "10.1109/ICLAD65226.2025.00033", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.3264}
{"id": "8eddeb2ed9d73e11778df55b801917419fbb4d05148afc297d0d47dbd86e8867", "sources": ["arxiv", "semantic_scholar"], "title": "An Efficient Open World Environment for Multi-Agent Social Learning", "abstract": "Many challenges remain before AI agents can be deployed in real-world environments. However, one virtue of such environments is that they are inherently multi-agent and contain human experts. Using advanced social intelligence in such an environment can help an AI agent learn adaptive skills and behaviors that a known expert exhibits. While social intelligence could accelerate training, it is currently difficult to study due to the lack of open-ended multi-agent environments. In this work, we present an environment in which multiple self-interested agents can pursue complex and independent goals, reflective of real world challenges. This environment will enable research into the development of socially intelligent AI agents in open-ended multi-agent settings, where agents may be implicitly incentivized to cooperate to defeat common enemies, build and share tools, and achieve long horizon goals. In this work, we investigate the impact on agent performance due to social learning in the presence of experts and implicit cooperation such as emergent collaborative tool use, and whether agents can benefit from either cooperation or competition in this environment.", "authors": ["Eric Ye", "Ren Tao", "Natasha Jaques"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-21", "url": "https://arxiv.org/abs/2508.15679", "pdf_url": "https://arxiv.org/pdf/2508.15679v1", "arxiv_id": "2508.15679", "doi": "10.48550/arXiv.2508.15679", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2761}
{"id": "36b55d1f9f92d47aec42ae79f19127847df734aaaaa67ab665adb2562dfc3080", "sources": ["arxiv", "semantic_scholar"], "title": "From Bits to Boardrooms: A Cutting-Edge Multi-Agent LLM Framework for Business Excellence", "abstract": "Large Language Models (LLMs) have shown promising potential in business applications, particularly in enterprise decision support and strategic planning, yet current approaches often struggle to reconcile intricate operational analyses with overarching strategic goals across diverse market environments, leading to fragmented workflows and reduced collaboration across organizational levels. This paper introduces BusiAgent, a novel multi-agent framework leveraging LLMs for advanced decision-making in complex corporate environments. BusiAgent integrates three core innovations: an extended Continuous Time Markov Decision Process (CTMDP) for dynamic agent modeling, a generalized entropy measure to optimize collaborative efficiency, and a multi-level Stackelberg game to handle hierarchical decision processes. Additionally, contextual Thompson sampling is employed for prompt optimization, supported by a comprehensive quality assurance system to mitigate errors. Extensive empirical evaluations across diverse business scenarios validate BusiAgent's efficacy, demonstrating its capacity to generate coherent, client-focused solutions that smoothly integrate granular insights with high-level strategy, significantly outperforming established approaches in both solution quality and user satisfaction. By fusing cutting-edge AI technologies with deep business insights, BusiAgent marks a substantial step forward in AI-driven enterprise decision-making, empowering organizations to navigate complex business landscapes more effectively.", "authors": ["Zihao Wang", "Junming Zhang"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-21", "url": "https://arxiv.org/abs/2508.15447", "pdf_url": "https://arxiv.org/pdf/2508.15447v2", "arxiv_id": "2508.15447", "doi": "10.48550/arXiv.2508.15447", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.2761}
{"id": "0d4adf17f47aa5bdfae9e7fd9f448a2490032143b6362558cdc8a162523c6d98", "sources": ["arxiv", "semantic_scholar"], "title": "Can LLM Agents Solve Collaborative Tasks? A Study on Urgency-Aware Planning and Coordination", "abstract": "The ability to coordinate actions across multiple agents is critical for solving complex, real-world problems. Large Language Models (LLMs) have shown strong capabilities in communication, planning, and reasoning, raising the question of whether they can also support effective collaboration in multi-agent settings. In this work, we investigate the use of LLM agents to solve a structured victim rescue task that requires division of labor, prioritization, and cooperative planning. Agents operate in a fully known graph-based environment and must allocate resources to victims with varying needs and urgency levels. We systematically evaluate their performance using a suite of coordination-sensitive metrics, including task success rate, redundant actions, room conflicts, and urgency-weighted efficiency. This study offers new insights into the strengths and failure modes of LLMs in physically grounded multi-agent collaboration tasks, contributing to future benchmarks and architectural improvements.", "authors": ["João Vitor de Carvalho Silva", "Douglas G. Macharet"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-20", "url": "https://arxiv.org/abs/2508.14635", "pdf_url": "https://arxiv.org/pdf/2508.14635v1", "arxiv_id": "2508.14635", "doi": "10.1109/ICAR65334.2025.11338680", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Advanced Robotics", "quality_score": 0.275}
{"id": "8439be70a5b3e736ecbdc926f14b153dd8c07b2948eca8ee4e6143f91ec83b3b", "sources": ["arxiv", "semantic_scholar"], "title": "CausalPlan: Empowering Efficient LLM Multi-Agent Collaboration Through Causality-Driven Planning", "abstract": "Large language model (LLM) agents-especially smaller, open-source models-often produce causally invalid or incoherent actions in collaborative tasks due to their reliance on surface-level correlations rather than grounded causal reasoning. This limitation undermines their performance in terms of coordination and planning in dynamic environments. We address this challenge with CausalPlan, a two-phase framework that integrates explicit structural causal reasoning into the LLM planning process. At the core of CausalPlan is the Structural Causal Action (SCA) model, which learns a causal graph from agent trajectories to capture how prior actions and current environment states influence future decisions. This structure is then used to guide action selection by assigning causal scores to LLM-generated proposals, reweighting them accordingly, or falling back to causally grounded alternatives when needed. By embedding this causal knowledge directly into the decision loop, CausalPlan constrains planning to intervention-consistent behaviours without requiring fine-tuning of the LLM itself. We evaluate CausalPlan on the Overcooked-AI benchmark across five multi-agent coordination tasks and four LLMs of varying sizes: Gemma-7B, Llama-8B, Qwen-14B, and Llama-70B. Experimental results show that CausalPlan consistently reduces invalid actions and improves collaboration in both AI-AI and human-AI settings, outperforming strong reinforcement learning baselines. Our findings highlight the value of causality-driven planning for deploying efficient, interpretable, and generalisable multi-agent LLM systems.", "authors": ["Minh Hoang Nguyen", "Van Dai Do", "Dung Nguyen", "Thin Nguyen", "Hung Le"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-19", "url": "https://arxiv.org/abs/2508.13721", "pdf_url": "https://arxiv.org/pdf/2508.13721v1", "arxiv_id": "2508.13721", "doi": "10.48550/arXiv.2508.13721", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4232}
{"id": "e2e8a45fc495a626f8725be4fba84f92e0faea34ef42136bd354471bd0a7d7fd", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Organizing Agent Network for LLM-based Workflow Automation", "abstract": "Recent multi-agent frameworks built upon large language models (LLMs) have demonstrated remarkable capabilities in complex task planning. However, in real-world enterprise environments, business workflows are typically composed through modularization and reuse of numerous subprocesses, resulting in intricate workflows characterized by lengthy and deeply nested execution paths. Such complexity poses significant challenges for LLM-driven orchestration, as extended reasoning chains and state-space explosions severely impact planning effectiveness and the proper sequencing of tool invocations. Therefore, developing an orchestration method with controllable structures capable of handling multi-layer nesting becomes a critical issue. To address this, we propose a novel structure-driven orchestration framework Self-Organizing Agent Network (SOAN). SOAN incrementally builds a formalized agent network by identifying and encapsulating structural units as independent agents, enhancing modularity and clarity in orchestration. Extensive evaluations were performed using multiple benchmarks as well as a real-world enterprise workflow dataset. Experimental results demonstrate that SOAN significantly outperforms state-of-the-art methods in terms of adaptability, fault tolerance, and execution efficiency.", "authors": ["Yiming Xiong", "Jian Wang", "Bing Li", "Yuhan Zhu", "Yuqi Zhao"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-19", "url": "https://arxiv.org/abs/2508.13732", "pdf_url": "https://arxiv.org/pdf/2508.13732v2", "arxiv_id": "2508.13732", "doi": "10.48550/arXiv.2508.13732", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2739}
{"id": "28a9c0dead85124e2b7b25ffce4082378fea1285fcde3c2b9872e5d1536b3188", "sources": ["arxiv", "semantic_scholar"], "title": "GraphCogent: Mitigating LLMs' Working Memory Constraints via Multi-Agent Collaboration in Complex Graph Understanding", "abstract": "Large language models (LLMs) show promising performance on small-scale graph reasoning tasks but fail when handling real-world graphs with complex queries. This phenomenon arises from LLMs' working memory constraints, which result in their inability to retain long-range graph topology over extended contexts while sustaining coherent multi-step reasoning. However, real-world graphs are often structurally complex, such as Web, Transportation, Social, and Citation networks. To address these limitations, we propose GraphCogent, a collaborative agent framework inspired by human Working Memory Model that decomposes graph reasoning into specialized cognitive processes: sense, buffer, and execute. The framework consists of three modules: Sensory Module standardizes diverse graph text representations via subgraph sampling, Buffer Module integrates and indexes graph data across multiple formats, and Execution Module combines tool calling and tool creation for efficient reasoning. We also introduce Graph4real, a comprehensive benchmark that contains four domains of real-world graphs (Web, Transportation, Social, and Citation) to evaluate LLMs' graph reasoning capabilities. Our Graph4real covers 21 different graph reasoning tasks, categorized into three types (Structural Querying, Algorithmic Reasoning, and Predictive Modeling tasks), with graph scales up to 10 times larger than existing benchmarks. Experiments show that Llama3.1-8B based GraphCogent achieves a 50% improvement over massive-scale LLMs like DeepSeek-R1 (671B). Compared to state-of-the-art agent-based baseline, our framework outperforms by 20% in accuracy while reducing token usage by 80% for in-toolset tasks and 30% for out-toolset tasks. Code will be available after review.", "authors": ["Rongzheng Wang", "Shuang Liang", "Qizhi Chen", "Yihong Huang", "Muquan Li", "Yizhuo Ma", "Dongyang Zhang", "Ke Qin", "Man-Fai Leung"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-17", "url": "https://arxiv.org/abs/2508.12379", "pdf_url": "https://arxiv.org/pdf/2508.12379v2", "arxiv_id": "2508.12379", "doi": "10.1145/3774904.3792314", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2258}
{"id": "b6a19c2341c9e67cbe09574e3dbf6694443ad0c31e8e0feaffdd362ebe128bda", "sources": ["arxiv", "semantic_scholar"], "title": "CORE: Measuring Multi-Agent LLM Interaction Quality under Game-Theoretic Pressures", "abstract": "Game-theoretic interactions between agents with Large Language Models (LLMs) have revealed many emergent capabilities, yet the linguistic diversity of these interactions has not been sufficiently quantified. In this paper, we present the Conversational Robustness Evaluation Score: CORE, a metric to quantify the effectiveness of language use within multi-agent systems across different game-theoretic interactions. CORE integrates measures of cluster entropy, lexical repetition, and semantic similarity, providing a direct lens of dialog quality. We apply CORE to pairwise LLM dialogs across competitive, cooperative, and neutral settings, further grounding our analysis in Zipf's and Heaps' Laws to characterize word frequency distributions and vocabulary growth. Our findings show that cooperative settings exhibit both steeper Zipf distributions and higher Heap exponents, indicating more repetition alongside greater vocabulary expansion. In contrast, competitive interactions display lower Zipf and Heaps exponents, reflecting less repetition and more constrained vocabularies. These results provide new insights into how social incentives influence language adaptation, and highlight CORE as a robust diagnostic for measuring linguistic robustness in multi-agent LLM systems. Our code is available at https://github.com/psyonp/core.", "authors": ["Punya Syon Pandey", "Yongjin Yang", "Jiarui Liu", "Zhijing Jin"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-16", "url": "https://arxiv.org/abs/2508.11915", "pdf_url": "https://arxiv.org/pdf/2508.11915v2", "arxiv_id": "2508.11915", "doi": "10.48550/arXiv.2508.11915", "citation_count": 4, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/psyonp/core", "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.4179}
{"id": "dbb46cb94abb1f0d2a231eae8a4a2fd98f1952826492500735d6f35922f80614", "sources": ["arxiv", "semantic_scholar"], "title": "FACET: Teacher-Centred LLM-Based Multi-Agent Systems-Towards Personalized Educational Worksheets", "abstract": "The increasing heterogeneity of student populations poses significant challenges for teachers, particularly in mathematics education, where cognitive, motivational, and emotional differences strongly influence learning outcomes. While AI-driven personalization tools have emerged, most remain performance-focused, offering limited support for teachers and neglecting broader pedagogical needs. This paper presents the FACET framework, a teacher-facing, large language model (LLM)-based multi-agent system designed to generate individualized classroom materials that integrate both cognitive and motivational dimensions of learner profiles. The framework comprises three specialized agents: (1) learner agents that simulate diverse profiles incorporating topic proficiency and intrinsic motivation, (2) a teacher agent that adapts instructional content according to didactical principles, and (3) an evaluator agent that provides automated quality assurance. We tested the system using authentic grade 8 mathematics curriculum content and evaluated its feasibility through a) automated agent-based assessment of output quality and b) exploratory feedback from K-12 in-service teachers. Results from ten internal evaluations highlighted high stability and alignment between generated materials and learner profiles, and teacher feedback particularly highlighted structure and suitability of tasks. The findings demonstrate the potential of multi-agent LLM architectures to provide scalable, context-aware personalization in heterogeneous classroom settings, and outline directions for extending the framework to richer learner profiles and real-world classroom trials.", "authors": ["Jana Gonnermann-Müller", "Jennifer Haase", "Konstantin Fackeldey", "Sebastian Pokutta"], "categories": ["cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-15", "url": "https://arxiv.org/abs/2508.11401", "pdf_url": "https://arxiv.org/pdf/2508.11401v5", "arxiv_id": "2508.11401", "doi": "10.48550/arXiv.2508.11401", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2693}
{"id": "93bcb8dc1fa71d5d64d85489c8d6964a6c96ea137f9284d744e6a3e86b94cac2", "sources": ["arxiv", "semantic_scholar"], "title": "SafeSieve: From Heuristics to Experience in Progressive Pruning for LLM-based Multi-Agent Communication", "abstract": "LLM-based multi-agent systems exhibit strong collaborative capabilities but often suffer from redundant communication and excessive token overhead. Existing methods typically enhance efficiency through pretrained GNNs or greedy algorithms, but often isolate pre- and post-task optimization, lacking a unified strategy. To this end, we present SafeSieve, a progressive and adaptive multi-agent pruning algorithm that dynamically refines the inter-agent communication through a novel dual-mechanism. SafeSieve integrates initial LLM-based semantic evaluation with accumulated performance feedback, enabling a smooth transition from heuristic initialization to experience-driven refinement. Unlike existing greedy Top-k pruning methods, SafeSieve employs 0-extension clustering to preserve structurally coherent agent groups while eliminating ineffective links. Experiments across benchmarks (SVAMP, HumanEval, etc.) showcase that SafeSieve achieves 94.01% average accuracy while reducing token usage by 12.4%-27.8%. Results further demonstrate robustness under prompt injection attacks (1.23% average accuracy drop). In heterogeneous settings, SafeSieve reduces deployment costs by 13.3% while maintaining performance. These results establish SafeSieve as an efficient, GPU-free, and scalable framework for practical multi-agent systems. Our code can be found here: https://github.com/csgen/SafeSieve", "authors": ["Ruijia Zhang", "Xinyan Zhao", "Ruixiang Wang", "Sigen Chen", "Guibin Zhang", "An Zhang", "Kun Wang", "Qingsong Wen"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-15", "url": "https://arxiv.org/abs/2508.11733", "pdf_url": "https://arxiv.org/pdf/2508.11733v3", "arxiv_id": "2508.11733", "doi": "10.48550/arXiv.2508.11733", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/csgen/SafeSieve", "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.4161}
{"id": "e282a2c2e121e7eed834912a1fdefbfeab7f089895351e24df1a9d97a93509c8", "sources": ["arxiv", "semantic_scholar"], "title": "Trustworthy AI Psychotherapy: Multi-Agent LLM Workflow for Counseling and Explainable Mental Disorder Diagnosis", "abstract": "LLM-based agents have emerged as transformative tools capable of executing complex tasks through iterative planning and action, achieving significant advancements in understanding and addressing user needs. Yet, their effectiveness remains limited in specialized domains such as mental health diagnosis, where they underperform compared to general applications. Current approaches to integrating diagnostic capabilities into LLMs rely on scarce, highly sensitive mental health datasets, which are challenging to acquire. These methods also fail to emulate clinicians' proactive inquiry skills, lack multi-turn conversational comprehension, and struggle to align outputs with expert clinical reasoning. To address these gaps, we propose DSM5AgentFlow, the first LLM-based agent workflow designed to autonomously generate DSM-5 Level-1 diagnostic questionnaires. By simulating therapist-client dialogues with specific client profiles, the framework delivers transparent, step-by-step disorder predictions, producing explainable and trustworthy results. This workflow serves as a complementary tool for mental health diagnosis, ensuring adherence to ethical and legal standards. Through comprehensive experiments, we evaluate leading LLMs across three critical dimensions: conversational realism, diagnostic accuracy, and explainability. Our datasets and implementations are fully open-sourced.", "authors": ["Mithat Can Ozgun", "Jiahuan Pei", "Koen Hindriks", "Lucia Donatelli", "Qingzhi Liu", "Junxiao Wang"], "categories": ["cs.HC", "cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-15", "url": "https://arxiv.org/abs/2508.11398", "pdf_url": "https://arxiv.org/pdf/2508.11398v2", "arxiv_id": "2508.11398", "doi": "10.1145/3746252.3761164", "citation_count": 11, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "International Conference on Information and Knowledge Management", "quality_score": 0.4161}
{"id": "788a6c270d0da8aca1da4d1884fae9e90c01103bb9edbcca200d4c73757d624a", "sources": ["arxiv", "semantic_scholar"], "title": "FROGENT: An End-to-End Full-process Drug Design Multi-Agent System", "abstract": "Drug discovery is a complex, multi-step pipeline that remains heavily dependent on manual, experience-driven operations; meanwhile, existing customized artificial intelligence tools are fragmented across web applications, desktop software, and code libraries, resulting in incompatible interfaces and inefficient, burdensome workflows. To overcome these challenges, we propose FROGENT, a full-process drug design multi-agent system that leverages the planning, reasoning, and tool-use capabilities of large language models (LLMs) to unify drug discovery within a closed-loop and autonomous framework. FROGENT is a collaborative multi-agent system comprising a central Orchestrate Agent for strategic workflow coordination and three distributed agents, Retrieve, Forge, and Gauge, that employ dynamic biochemical databases, extensible tool libraries, and task-specific computational models via the Model Context Protocol. This architecture enables end-to-end execution of complex drug discovery pipelines, covering target identification, small-molecule generation, peptide optimization, and retrosynthetic planning. Across eight benchmarks spanning core drug discovery tasks, FROGENT consistently outperforms six increasingly advanced ReAct-style agents. Case studies further demonstrate its practicality and generalization across real-world small-molecule and peptide design scenarios. Overall, FROGENT not only achieves substantial gains in efficiency and accuracy, but also demonstrates the potential of LLM-based agentic systems to autonomously orchestrate drug development pipelines, reducing, or even replacing, reliance on manual, experience-driven human intervention.", "authors": ["Qihua Pan", "Dong Xu", "Qianwei Yang", "Jenna Xinyi Yao", "Sisi Yuan", "Zexuan Zhu", "Jianqiang Li", "Junkai Ji"], "categories": ["q-bio.BM", "cs.AI"], "fields_of_study": ["Biology", "Computer Science"], "published_date": "2025-08-14", "url": "https://arxiv.org/abs/2508.10760", "pdf_url": "https://arxiv.org/pdf/2508.10760v2", "arxiv_id": "2508.10760", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "b5b59040ab1558789630992e48f9b2dfb769b07acf4f07479972d8130c879962", "sources": ["arxiv", "semantic_scholar"], "title": "Searching for Privacy Risks in LLM Agents via Simulation", "abstract": "The widespread deployment of LLM-based agents is likely to introduce a critical privacy threat: malicious agents that proactively engage others in multi-turn interactions to extract sensitive information. However, the evolving nature of such dynamic dialogues makes it challenging to anticipate emerging vulnerabilities and design effective defenses. To tackle this problem, we present a search-based framework that alternates between improving attack and defense strategies through the simulation of privacy-critical agent interactions. Specifically, we employ LLMs as optimizers to analyze simulation trajectories and iteratively propose new agent instructions. To explore the strategy space more efficiently, we further utilize parallel search with multiple threads and cross-thread propagation. Through this process, we find that attack strategies escalate from direct requests to sophisticated tactics, such as impersonation and consent forgery, while defenses evolve from simple rule-based constraints to robust identity-verification state machines. The discovered attacks and defenses generalize across diverse scenarios and backbone models, providing useful insights for developing privacy-aware agents.", "authors": ["Yanzhe Zhang", "Diyi Yang"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-14", "url": "https://arxiv.org/abs/2508.10880", "pdf_url": "https://arxiv.org/pdf/2508.10880v3", "arxiv_id": "2508.10880", "doi": "10.48550/arXiv.2508.10880", "citation_count": 18, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "cf3bea97e31b1a8b493a0f9ad8b8af11754d3413f8b2f85314718a7b3e781090", "sources": ["arxiv", "semantic_scholar"], "title": "MCP-Orchestrated Multi-Agent System for Automated Disinformation Detection", "abstract": "The large spread of disinformation across digital platforms creates significant challenges to information integrity. This paper presents a multi-agent system that uses relation extraction to detect disinformation in news articles, focusing on titles and short text snippets. The proposed Agentic AI system combines four agents: (i) a machine learning agent (logistic regression), (ii) a Wikipedia knowledge check agent (which relies on named entity recognition), (iii) a coherence detection agent (using LLM prompt engineering), and (iv) a web-scraped data analyzer that extracts relational triplets for fact checking. The system is orchestrated via the Model Context Protocol (MCP), offering shared context and live learning across components. Results demonstrate that the multi-agent ensemble achieves 95.3% accuracy with an F1 score of 0.964, significantly outperforming individual agents and traditional approaches. The weighted aggregation method, mathematically derived from individual agent misclassification rates, proves superior to algorithmic threshold optimization. The modular architecture makes the system easily scalable, while also maintaining details of the decision processes.", "authors": ["Alexandru-Andrei Avram", "Adrian Groza", "Alexandru Lecu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-13", "url": "https://arxiv.org/abs/2508.10143", "pdf_url": "https://arxiv.org/pdf/2508.10143v1", "arxiv_id": "2508.10143", "doi": "10.1109/SYNASC69064.2025.00051", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Symposium on Symbolic and Numeric Algorithms for Scientific Computing", "quality_score": 0.267}
{"id": "77cd352b3d1e7b3cef195aa597e82395cc5511da48292c849477f71954bf32d7", "sources": ["arxiv", "semantic_scholar"], "title": "Profile-Aware Maneuvering: A Dynamic Multi-Agent System for Robust GAIA Problem Solving by AWorld", "abstract": "The rapid advancement of large language models (LLMs) has empowered intelligent agents to leverage diverse external tools for solving complex real-world problems. However, this reliance introduces new challenges, as extended contexts and noisy tool outputs can undermine system reliability. To address this, we propose a dynamic Multi-Agent System (MAS) in our AWorld framework, where an Execution Agent is supervised by a Guard Agent that provides on-demand dynamic maneuvering, verifying and correcting the reasoning process to improve robustness over single-agent systems. To move beyond this generic supervision, we enhance the architecture with a methodology inspired by System Identification from control theory. This method first profiles the Execution Agent offline on a benchmark dataset to create a \"performance fingerprint\" of its unique weaknesses. The Guard Agent then leverages this fingerprint online to deliver profile-aware supervision, making targeted interventions based on known failure patterns rather than merely reacting to immediate logical flaws. Extensive experiments on the GAIA dataset demonstrate that this profile-aware MAS significantly improves both effectiveness and stability, outperforming not only single-agent systems but also its naive counterpart. This superior performance led our system to achieve first place among open-source projects on the prestigious GAIA leaderboard. These findings highlight that building truly trustworthy intelligent systems requires not just collaboration, but a deep, empirically-grounded understanding of each agent's unique capabilities and limitations.", "authors": ["Zhitian Xie", "Qintong Wu", "Chengyue Yu", "Chenyi Zhuang", "Jinjie Gu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-13", "url": "https://arxiv.org/abs/2508.09889", "pdf_url": "https://arxiv.org/pdf/2508.09889v4", "arxiv_id": "2508.09889", "doi": "10.48550/arXiv.2508.09889", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4126}
{"id": "aa4ecbcfe956c295c9350bc0616ed5d84914db09a3056f2dc6b9de37c3524611", "sources": ["arxiv", "semantic_scholar"], "title": "Extending the OWASP Multi-Agentic System Threat Modeling Guide: Insights from Multi-Agent Security Research", "abstract": "We propose an extension to the OWASP Multi-Agentic System (MAS) Threat Modeling Guide, translating recent anticipatory research in multi-agent security (MASEC) into practical guidance for addressing challenges unique to large language model (LLM)-driven multi-agent architectures. Although OWASP's existing taxonomy covers many attack vectors, our analysis identifies gaps in modeling failures, including, but not limited to: reasoning collapse across planner-executor chains, metric overfitting, unsafe delegation escalation, emergent covert coordination, and heterogeneous multi-agent exploits. We introduce additional threat classes and scenarios grounded in practical MAS deployments, highlighting risks from benign goal drift, cross-agent hallucination propagation, affective prompt framing, and multi-agent backdoors. We also outline evaluation strategies, including robustness testing, coordination assessment, safety enforcement, and emergent behavior monitoring, to ensure complete coverage. This work complements the framework of OWASP by expanding its applicability to increasingly complex, autonomous, and adaptive multi-agent systems, with the goal of improving security posture and resilience in real world deployments.", "authors": ["Klaudia Krawiecka", "Christian Schroeder de Witt"], "categories": ["cs.MA", "cs.CR", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-13", "url": "https://arxiv.org/abs/2508.09815", "pdf_url": "https://arxiv.org/pdf/2508.09815v1", "arxiv_id": "2508.09815", "doi": "10.48550/arXiv.2508.09815", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.267}
{"id": "5d06354b25b58173abc9736c927baa95d110da1163f77c4135d52e38df166902", "sources": ["arxiv", "semantic_scholar"], "title": "Reducing Cognitive Overhead in Tool Use via Multi-Small-Agent Reinforcement Learning", "abstract": "Recent advances in multi-agent systems highlight the potential of specialized small agents that collaborate via division of labor. Existing tool-integrated reasoning systems, however, often follow a single-agent paradigm in which one large model interleaves long-horizon reasoning with precise tool operations, leading to cognitive-load interference and unstable coordination. We present MSARL, a Multi-Small-Agent Reinforcement Learning framework that explicitly decouples reasoning from tool use. In MSARL, a Reasoning Agent decomposes problems and plans tool invocations, while multiple Tool Agents specialize in specific external tools, each trained via a combination of imitation learning and reinforcement learning with role-specific rewards. On mathematical problem solving with code execution, MSARL significantly improves reasoning stability and final-answer accuracy over single-agent baselines. Moreover, the architecture generalizes to diverse tool-use tasks, demonstrating that cognitive-role decoupling with small agents is a scalable blueprint for multi-agent AI design.", "authors": ["Dayu Wang", "Jiaye Yang", "Weikang Li", "Jiahui Liang", "Yang Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-12", "url": "https://arxiv.org/abs/2508.08882", "pdf_url": "https://arxiv.org/pdf/2508.08882v4", "arxiv_id": "2508.08882", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1692}
{"id": "e7360f57126fe4633211ddcbede4b12c8f911dcc13cbb63960eb753a84745fc1", "sources": ["arxiv", "semantic_scholar"], "title": "Intrinsic Memory Agents: Heterogeneous Multi-Agent LLM Systems through Structured Contextual Memory", "abstract": "Multi-agent systems built on Large Language Models (LLMs) show exceptional promise for complex collaborative problem-solving, yet they face fundamental challenges stemming from context window limitations that impair memory consistency, role adherence, and procedural integrity. This paper introduces Intrinsic Memory Agents, a novel framework that addresses these limitations through agent-specific memories that evolve intrinsically with agent outputs. Specifically, our method maintains role-aligned memory that preserves specialized perspectives while focusing on task-relevant information. Our approach utilises a generic memory template applicable to new problems without the need to hand-craft specific memory prompts. We benchmark our approach on the PDDL, FEVER, and ALFWorld datasets, comparing its performance to existing state-of-the-art multi-agentic memory approaches and showing state-of-the-art or comparable performance across all three, with the highest consistency. An additional evaluation is performed on a complex data pipeline design task, and we demonstrate that our approach produces higher quality designs across 5 metrics: scalability, reliability, usability, cost-effectiveness, and documentation, plus additional qualitative evidence of the improvements. Our findings suggest that addressing memory limitations through intrinsic approaches can improve the capabilities of multi-agent LLM systems on structured planning tasks.", "authors": ["Sizhe Yuen", "Francisco Gomez Medina", "Ting Su", "Yali Du", "Adam J. Sobey"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-12", "url": "https://arxiv.org/abs/2508.08997", "pdf_url": "https://arxiv.org/pdf/2508.08997v2", "arxiv_id": "2508.08997", "doi": "10.48550/arXiv.2508.08997", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "a853c53d24a7bd377f87c95b8b28d21868110367d0899a18ad51821ab3d83bc4", "sources": ["arxiv", "semantic_scholar"], "title": "CRADLE: Conversational RTL Design Space Exploration with LLM-based Multi-Agent Systems", "abstract": "This paper presents CRADLE, a conversational framework for design space exploration of RTL designs using LLM-based multi-agent systems. Unlike existing rigid approaches, CRADLE enables user-guided flows with internal self-verification, correction, and optimization. We demonstrate the framework with a generator-critic agent system targeting FPGA resource minimization using state-of-the-art LLMs. Experimental results on the RTLLM benchmark show that CRADLE achieves significant reductions in resource usage with averages of 48% and 40% in LUTs and FFs across all benchmark designs.", "authors": ["Lukas Krupp", "Maximilian Schöffel", "Elias Biehl", "Norbert Wehn"], "categories": ["cs.RO", "cs.AR", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-12", "url": "https://arxiv.org/abs/2508.08709", "pdf_url": "https://arxiv.org/pdf/2508.08709v1", "arxiv_id": "2508.08709", "doi": "10.1109/ISOCC66390.2025.11329873", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International SoC Design Conference", "quality_score": 0.2658}
{"id": "f374664f5ae4a49a117d239a4a6ae5ad43a48469fc295b2c4428b0fa4f001a48", "sources": ["arxiv", "semantic_scholar"], "title": "OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows", "abstract": "Autonomous agents powered by large language models (LLMs) are increasingly deployed in real-world applications requiring complex, long-horizon workflows. However, existing benchmarks predominantly focus on atomic tasks that are self-contained and independent, failing to capture the long-term contextual dependencies and multi-interaction coordination required in realistic scenarios. To address this gap, we introduce OdysseyBench, a comprehensive benchmark for evaluating LLM agents on long-horizon workflows across diverse office applications including Word, Excel, PDF, Email, and Calendar. Our benchmark comprises two complementary splits: OdysseyBench+ with 300 tasks derived from real-world use cases, and OdysseyBench-Neo with 302 newly synthesized complex tasks. Each task requires agent to identify essential information from long-horizon interaction histories and perform multi-step reasoning across various applications. To enable scalable benchmark creation, we propose HomerAgents, a multi-agent framework that automates the generation of long-horizon workflow benchmarks through systematic environment exploration, task generation, and dialogue synthesis. Our extensive evaluation demonstrates that OdysseyBench effectively challenges state-of-the-art LLM agents, providing more accurate assessment of their capabilities in complex, real-world contexts compared to existing atomic task benchmarks. We believe that OdysseyBench will serve as a valuable resource for advancing the development and evaluation of LLM agents in real-world productivity scenarios. In addition, we release OdysseyBench and HomerAgents to foster research along this line.", "authors": ["Weixuan Wang", "Dongge Han", "Daniel Madrigal Diaz", "Jin Xu", "Victor Rühle", "Saravan Rajmohan"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-12", "url": "https://arxiv.org/abs/2508.09124", "pdf_url": "https://arxiv.org/pdf/2508.09124v1", "arxiv_id": "2508.09124", "doi": "10.48550/arXiv.2508.09124", "citation_count": 30, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3728}
{"id": "65e62778ecefcc7c0320cd524f66296a8528e9db57269c6789b1b6fa7be8ee66", "sources": ["arxiv", "semantic_scholar"], "title": "MAViS: A Multi-Agent Framework for Long-Sequence Video Storytelling", "abstract": "Despite recent advances, long-sequence video generation frameworks still suffer from significant limitations: poor assistive capability, suboptimal visual quality, and limited expressiveness. To mitigate these limitations, we propose MAViS, a multi-agent collaborative framework designed to assist in long-sequence video storytelling by efficiently translating ideas into visual narratives. MAViS orchestrates specialized agents across multiple stages, including script writing, shot designing, character modeling, keyframe generation, video animation, and audio generation. In each stage, agents operate under the 3E Principle -- Explore, Examine, and Enhance -- to ensure the completeness of intermediate outputs. Considering the capability limitations of current generative models, we propose the Script Writing Guidelines to optimize compatibility between scripts and generative tools. Experimental results demonstrate that MAViS achieves state-of-the-art performance in assistive capability, visual quality, and video expressiveness. Its modular framework further enables scalability with diverse generative models and tools. With just a brief idea description, MAViS enables users to rapidly explore diverse visual storytelling and creative directions for sequential video generation by efficiently producing high-quality, complete long-sequence videos. To the best of our knowledge, MAViS is the only framework that provides multimodal design output -- videos with narratives and background music.", "authors": ["Qian Wang", "Ziqi Huang", "Ruoxi Jia", "Paul Debevec", "Ning Yu"], "categories": ["cs.CV", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-11", "url": "https://arxiv.org/abs/2508.08487", "pdf_url": "https://arxiv.org/pdf/2508.08487v5", "arxiv_id": "2508.08487", "doi": "10.48550/arXiv.2508.08487", "citation_count": 13, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.2865}
{"id": "c2e1b13b0b89df92b9b7ffc3135c93e5d57f19d0087c1facc4eb49f1dd61dbaf", "sources": ["arxiv", "semantic_scholar"], "title": "MCPToolBench++: A Large Scale AI Agent Model Context Protocol MCP Tool Use Benchmark", "abstract": "LLMs' capabilities are enhanced by using function calls to integrate various data sources or API results into the context window. Typical tools include search, web crawlers, maps, financial data, file systems, and browser usage, etc. Integrating these data sources or functions requires a standardized method. The Model Context Protocol (MCP) provides a standardized way to supply context to LLMs. However, the evaluation of LLMs and AI Agents' MCP tool use abilities suffer from several issues. First, there's a lack of comprehensive datasets or benchmarks to evaluate various MCP tools. Second, the diverse formats of response from MCP tool call execution further increase the difficulty of evaluation. Additionally, unlike existing tool-use benchmarks with high success rates in functions like programming and math functions, the success rate of real-world MCP tool is not guaranteed and varies across different MCP servers. Furthermore, the LLMs' context window also limits the number of available tools that can be called in a single run, because the textual descriptions of tool and the parameters have long token length for an LLM to process all at once. To help address the challenges of evaluating LLMs' performance on calling MCP tools, we propose MCPToolBench++, a large-scale, multi-domain AI Agent tool use benchmark. As of July 2025, this benchmark is build upon marketplace of over 4k MCP servers from more than 40 categories, collected from the MCP marketplaces and GitHub communities. The datasets consist of both single-step and multi-step tool calls across different categories. We evaluated SOTA LLMs with agentic abilities on this benchmark and reported the results.", "authors": ["Shiqing Fan", "Xichen Ding", "Liang Zhang", "Linjian Mo"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-11", "url": "https://arxiv.org/abs/2508.07575", "pdf_url": "https://arxiv.org/pdf/2508.07575v1", "arxiv_id": "2508.07575", "doi": "10.48550/arXiv.2508.07575", "citation_count": 27, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3618}
{"id": "90054ba9f8c3e1a7b4ddc5557770063ca69ae12bbc58420450cba0412082b24d", "sources": ["arxiv", "semantic_scholar"], "title": "1-2-3 Check: Enhancing Contextual Privacy in LLM via Multi-Agent Reasoning", "abstract": "Addressing contextual privacy concerns remains challenging in interactive settings where large language models (LLMs) process information from multiple sources (e.g., summarizing meetings with private and public information). We introduce a multi-agent framework that decomposes privacy reasoning into specialized subtasks (extraction, classification), reducing the information load on any single agent while enabling iterative validation and more reliable adherence to contextual privacy norms. To understand how privacy errors emerge and propagate, we conduct a systematic ablation over information-flow topologies, revealing when and why upstream detection mistakes cascade into downstream leakage. Experiments on the ConfAIde and PrivacyLens benchmark with several open-source and closed-sourced LLMs demonstrate that our best multi-agent configuration substantially reduces private information leakage (\\textbf{18\\%} on ConfAIde and \\textbf{19\\%} on PrivacyLens with GPT-4o) while preserving the fidelity of public content, outperforming single-agent baselines. These results highlight the promise of principled information-flow design in multi-agent systems for contextual privacy with LLMs.", "authors": ["Wenkai Li", "Liwen Sun", "Zhenxiang Guan", "Xuhui Zhou", "Maarten Sap"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-11", "url": "https://arxiv.org/abs/2508.07667", "pdf_url": "https://arxiv.org/pdf/2508.07667v3", "arxiv_id": "2508.07667", "doi": "10.48550/arXiv.2508.07667", "citation_count": 11, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "cbfe09715d9e1cb59c4f6a924a8a133ee28f4b3857fb5e6a6e0f80042162924f", "sources": ["arxiv", "semantic_scholar"], "title": "BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks", "abstract": "The security of LLM-based multi-agent systems (MAS) is critically threatened by propagation vulnerability, where malicious agents can distort collective decision-making through inter-agent message interactions. While existing supervised defense methods demonstrate promising performance, they may be impractical in real-world scenarios due to their heavy reliance on labeled malicious agents to train a supervised malicious detection model. To enable practical and generalizable MAS defenses, in this paper, we propose BlindGuard, an unsupervised defense method that learns without requiring any attack-specific labels or prior knowledge of malicious behaviors. To this end, we establish a hierarchical agent encoder to capture individual, neighborhood, and global interaction patterns of each agent, providing a comprehensive understanding for malicious agent detection. Meanwhile, we design a corruption-guided detector that consists of directional noise injection and contrastive learning, allowing effective detection model training solely on normal agent behaviors. Extensive experiments show that BlindGuard effectively detects diverse attack types (i.e., prompt injection, memory poisoning, and tool attack) across MAS with various communication patterns while maintaining superior generalizability compared to supervised baselines. The code is available at: https://github.com/MR9812/BlindGuard.", "authors": ["Rui Miao", "Yixin Liu", "Yili Wang", "Xu Shen", "Yue Tan", "Yiwei Dai", "Shirui Pan", "Xin Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-11", "url": "https://arxiv.org/abs/2508.08127", "pdf_url": "https://arxiv.org/pdf/2508.08127v2", "arxiv_id": "2508.08127", "doi": "10.48550/arXiv.2508.08127", "citation_count": 25, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/MR9812/BlindGuard", "venue": "arXiv.org", "quality_score": 0.4091}
{"id": "67ef748d248f417d4db3959b3582efcdece8ec3f92efe2abfb221c9f0dd53feb", "sources": ["arxiv", "semantic_scholar"], "title": "Chimera: Harnessing Multi-Agent LLMs for Automatic Insider Threat Simulation", "abstract": "Insider threats pose a persistent and critical security risk, yet are notoriously difficult to detect in complex enterprise environments, where malicious actions are often hidden within seemingly benign user behaviors. Although machine-learning-based insider threat detection (ITD) methods have shown promise, their effectiveness is fundamentally limited by the scarcity of high-quality and realistic training data. Enterprise internal data is highly sensitive and rarely accessible, while existing public and synthetic datasets are either small-scale or lack sufficient realism, semantic richness, and behavioral diversity. To address this challenge, we propose Chimera, an LLM-based multi-agent framework that automatically simulates both benign and malicious insider activities and generates comprehensive system logs across diverse enterprise environments. Chimera models each agent as an individual employee with fine-grained roles and supports group meetings, pairwise interactions, and self-organized scheduling to capture realistic organizational dynamics. Based on 15 insider attacks abstracted from real-world incidents, we deploy Chimera in three representative data-sensitive organizational scenarios and construct ChimeraLog, a new dataset for developing and evaluating ITD methods. We evaluate ChimeraLog through human studies and quantitative analyses, demonstrating its diversity and realism. Experiments with existing ITD methods show substantially lower detection performance on ChimeraLog compared to prior datasets, indicating a more challenging and realistic benchmark. Moreover, despite distribution shifts, models trained on ChimeraLog exhibit strong generalization, highlighting the practical value of LLM-based multi-agent simulation for advancing insider threat detection.", "authors": ["Jiongchi Yu", "Xiaofei Xie", "Qiang Hu", "Yuhan Ma", "Ziming Zhao"], "categories": ["cs.CR", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-11", "url": "https://arxiv.org/abs/2508.07745", "pdf_url": "https://arxiv.org/pdf/2508.07745v4", "arxiv_id": "2508.07745", "doi": "10.48550/arXiv.2508.07745", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "0996af1e600f29579fb77b4ec820e19ea3b04cfaf9a158c2e0069ec425535307", "sources": ["arxiv", "semantic_scholar"], "title": "Grounding Natural Language for Multi-agent Decision-Making with Multi-agentic LLMs", "abstract": "Language is a ubiquitous tool that is foundational to reasoning and collaboration, ranging from everyday interactions to sophisticated problem-solving tasks. The establishment of a common language can serve as a powerful asset in ensuring clear communication and understanding amongst agents, facilitating desired coordination and strategies. In this work, we extend the capabilities of large language models (LLMs) by integrating them with advancements in multi-agent decision-making algorithms. We propose a systematic framework for the design of multi-agentic large language models (LLMs), focusing on key integration practices. These include advanced prompt engineering techniques, the development of effective memory architectures, multi-modal information processing, and alignment strategies through fine-tuning algorithms. We evaluate these design choices through extensive ablation studies on classic game settings with significant underlying social dilemmas and game-theoretic considerations.", "authors": ["Dom Huh", "Prasant Mohapatra"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-10", "url": "https://arxiv.org/abs/2508.07466", "pdf_url": "https://arxiv.org/pdf/2508.07466v1", "arxiv_id": "2508.07466", "doi": "10.48550/arXiv.2508.07466", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2635}
{"id": "9c78d0f7a5874c21f2b322a3c4a3ffb94c9201583d7f4ad0791a20d1aa14aa78", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Approach to Neurological Clinical Reasoning", "abstract": "Large language models (LLMs) have shown promise in medical domains, but their ability to handle specialized neurological reasoning requires systematic evaluation. We developed a comprehensive benchmark using 305 questions from Israeli Board Certification Exams in Neurology, classified along three complexity dimensions: factual knowledge depth, clinical concept integration, and reasoning complexity. We evaluated ten LLMs using base models, retrieval-augmented generation (RAG), and a novel multi-agent system. Results showed significant performance variation. OpenAI-o1 achieved the highest base performance (90.9% accuracy), while specialized medical models performed poorly (52.9% for Meditron-70B). RAG provided modest benefits but limited effectiveness on complex reasoning questions. In contrast, our multi-agent framework, decomposing neurological reasoning into specialized cognitive functions including question analysis, knowledge retrieval, answer synthesis, and validation, achieved dramatic improvements, especially for mid-range models. The LLaMA 3.3-70B-based agentic system reached 89.2% accuracy versus 69.5% for its base model, with substantial gains on level 3 complexity questions. The multi-agent approach transformed inconsistent subspecialty performance into uniform excellence, addressing neurological reasoning challenges that persisted with RAG enhancement. We validated our approach using an independent dataset of 155 neurological cases from MedQA. Results confirm that structured multi-agent approaches designed to emulate specialized cognitive processes significantly enhance complex medical reasoning, offering promising directions for AI assistance in challenging clinical contexts.", "authors": ["Moran Sorka", "Alon Gorenshtein", "Dvir Aran", "Shahar Shelly"], "categories": ["cs.IR", "cs.AI"], "fields_of_study": ["Computer Science", "Medicine"], "published_date": "2025-08-10", "url": "https://arxiv.org/abs/2508.14063", "pdf_url": "https://arxiv.org/pdf/2508.14063v1", "arxiv_id": "2508.14063", "doi": "10.1371/journal.pdig.0001106", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "PLOS Digital Health", "quality_score": 0.2635}
{"id": "97ce2f6032aaa14f174432e261c3379cd6311a9b7c5990112fd5d6770634b879", "sources": ["arxiv", "semantic_scholar"], "title": "Context Engineering for Multi-Agent LLM Code Assistants Using Elicit, NotebookLM, ChatGPT, and Claude Code", "abstract": "Large Language Models (LLMs) have shown promise in automating code generation and software engineering tasks, yet they often struggle with complex, multi-file projects due to context limitations and knowledge gaps. We propose a novel context engineering workflow that combines multiple AI components: an Intent Translator (GPT-5) for clarifying user requirements, an Elicit-powered semantic literature retrieval for injecting domain knowledge, NotebookLM-based document synthesis for contextual understanding, and a Claude Code multi-agent system for code generation and validation. Our integrated approach leverages intent clarification, retrieval-augmented generation, and specialized sub-agents orchestrated via Claude's agent framework. We demonstrate that this method significantly improves the accuracy and reliability of code assistants in real-world repositories, yielding higher single-shot success rates and better adherence to project context than baseline single-agent approaches. Qualitative results on a large Next.js codebase show the multi-agent system effectively plans, edits, and tests complex features with minimal human intervention. We compare our system with recent frameworks like CodePlan, MASAI, and HyperAgent, highlighting how targeted context injection and agent role decomposition lead to state-of-the-art performance. Finally, we discuss the implications for deploying LLM-based coding assistants in production, along with lessons learned on context management and future research directions.", "authors": ["Muhammad Haseeb"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-09", "url": "https://arxiv.org/abs/2508.08322", "pdf_url": "https://arxiv.org/pdf/2508.08322v1", "arxiv_id": "2508.08322", "doi": "10.48550/arXiv.2508.08322", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2624}
{"id": "276da4164d06ccb6eb339d0edf6f9d94c6af3840cbde80f41f6da1f1bacf8644", "sources": ["arxiv", "semantic_scholar"], "title": "Kairos: Low-latency Multi-Agent Serving with Shared LLMs and Excessive Loads in the Public Cloud", "abstract": "Multi-agent applications utilize the advanced capabilities of large language models (LLMs) for intricate task completion through agent collaboration in a workflow. Under this situation, requests from different agents usually access the same shared LLM to perform different kinds of tasks, forcing the shared LLM to suffer excessive loads. However, existing works have low serving performance for these multi-agent applications, mainly due to the ignorance of inter-agent latency and resource differences for request scheduling. We therefore propose Kairos, a multi-agent orchestration system that optimizes end-to-end latency for multi-agent applications. Kairos consists of a workflow orchestrator, a workflow-aware priority scheduler, and a memory-aware dispatcher. The orchestrator collects agent-specific information for online workflow analysis. The scheduler decides the serving priority of the requests based on their latency characteristics to reduce the overall queuing. The dispatcher dispatches the requests to different LLM instances based on their memory demands to avoid GPU overloading. Experimental results show that Kairos reduces end-to-end latency by 17.8% to 28.4% compared to state-of-the-art works.", "authors": ["Jinyuan Chen", "Jiuchen Shi", "Quan Chen", "Minyi Guo"], "categories": ["cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-09", "url": "https://arxiv.org/abs/2508.06948", "pdf_url": "https://arxiv.org/pdf/2508.06948v1", "arxiv_id": "2508.06948", "doi": "10.48550/arXiv.2508.06948", "citation_count": 10, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2624}
{"id": "8364e795de53359a6c56afe86e901e16b2d22ae6e8914a66d06ac1b2bf10cd39", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-level Advantage Credit Assignment for Cooperative Multi-Agent Reinforcement Learning", "abstract": "Cooperative multi-agent reinforcement learning (MARL) aims to coordinate multiple agents to achieve a common goal. A key challenge in MARL is credit assignment, which involves assessing each agent's contribution to the shared reward. Given the diversity of tasks, agents may perform different types of coordination, with rewards attributed to diverse and often overlapping agent subsets. In this work, we formalize the credit assignment level as the number of agents cooperating to obtain a reward, and address scenarios with multiple coexisting levels. We introduce a multi-level advantage formulation that performs explicit counterfactual reasoning to infer credits across distinct levels. Our method, Multi-level Advantage Credit Assignment (MACA), captures agent contributions at multiple levels by integrating advantage functions that reason about individual, joint, and correlated actions. Utilizing an attention-based framework, MACA identifies correlated agent relationships and constructs multi-level advantages to guide policy learning. Comprehensive experiments on challenging Starcraft v1\\&v2 tasks demonstrate MACA's superior performance, underscoring its efficacy in complex credit assignment scenarios.", "authors": ["Xutong Zhao", "Yaqi Xie"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-09", "url": "https://arxiv.org/abs/2508.06836", "pdf_url": "https://arxiv.org/pdf/2508.06836v1", "arxiv_id": "2508.06836", "doi": "10.48550/arXiv.2508.06836", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Artificial Intelligence and Statistics", "quality_score": 0.2624}
{"id": "9cda0bd353e6577915c70aa5d769d943dbd68a18b41437e623be01d6f1330435", "sources": ["arxiv", "semantic_scholar"], "title": "Mediator-Guided Multi-Agent Collaboration among Open-Source Models for Medical Decision-Making", "abstract": "Complex medical decision-making involves cooperative workflows operated by different clinicians. Designing AI multi-agent systems can expedite and augment human-level clinical decision-making. Existing multi-agent researches primarily focus on language-only tasks, yet their extension to multimodal scenarios remains challenging. A blind combination of diverse vision-language models (VLMs) can amplify an erroneous outcome interpretation. VLMs in general are less capable in instruction following and importantly self-reflection, compared to large language models (LLMs) of comparable sizes. This disparity largely constrains VLMs' ability in cooperative workflows. In this study, we propose MedOrch, a mediator-guided multi-agent collaboration framework for medical multimodal decision-making. MedOrch employs an LLM-based mediator agent that enables multiple VLM-based expert agents to exchange and reflect on their outputs towards collaboration. We utilize multiple open-source general-purpose and domain-specific VLMs instead of costly GPT-series models, revealing the strength of heterogeneous models. We show that the collaboration within distinct VLM-based agents can surpass the capabilities of any individual agent. We validate our approach on five medical vision question answering benchmarks, demonstrating superior collaboration performance without model training. Our findings underscore the value of mediator-guided multi-agent collaboration in advancing medical multimodal intelligence.", "authors": ["Kaitao Chen", "Mianxin Liu", "Daoming Zong", "Chaoyue Ding", "Shaohao Rui", "Yankai Jiang", "Mu Zhou", "Xiaosong Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-08", "url": "https://arxiv.org/abs/2508.05996", "pdf_url": "https://arxiv.org/pdf/2508.05996v2", "arxiv_id": "2508.05996", "doi": "10.48550/arXiv.2508.05996", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4037}
{"id": "33585328ca170ba3b0be9296e1c8555bca2277c679930f06a38866685cd8b9ac", "sources": ["arxiv", "semantic_scholar"], "title": "Grid-Agent: An LLM-Powered Multi-Agent System for Power Grid Control", "abstract": "Modern power grids face unprecedented complexity from Distributed Energy Resources (DERs), Electric Vehicles (EVs), and extreme weather, while also being increasingly exposed to cyberattacks that can trigger grid violations. This paper introduces Grid-Agent, an autonomous AI-driven framework that leverages Large Language Models (LLMs) within a multi-agent system to detect and remediate violations. Grid-Agent integrates semantic reasoning with numerical precision through modular agents: a planning agent generates coordinated action sequences using power flow solvers, while a validation agent ensures stability and safety through sandboxed execution with rollback mechanisms. To enhance scalability, the framework employs an adaptive multi-scale network representation that dynamically adjusts encoding schemes based on system size and complexity. Violation resolution is achieved through optimizing switch configurations, battery deployment, and load curtailment. Our experiments on IEEE and CIGRE benchmark networks, including the IEEE 69-bus, CIGRE MV, IEEE 30-bus test systems, demonstrate superior mitigation performance, highlighting Grid-Agent's suitability for modern smart grids requiring rapid, adaptive response.", "authors": ["Yan Zhang", "Ahmad Mohammad Saber", "Amr Youssef", "Deepa Kundur"], "categories": ["cs.MA", "cs.AI", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-08-07", "url": "https://arxiv.org/abs/2508.05702", "pdf_url": "https://arxiv.org/pdf/2508.05702v3", "arxiv_id": "2508.05702", "doi": "10.48550/arXiv.2508.05702", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "ef90daba5a5f38af5d344756a9241e9bcaec98a80e6b7669789253d0d1aeba4f", "sources": ["arxiv", "semantic_scholar"], "title": "CLAPP: The CLASS LLM Agent for Pair Programming", "abstract": "We introduce CLAPP (CLASS LLM Agent for Pair Programming), an interactive AI assistant designed to support researchers working with the Einstein-Boltzmann solver CLASS. CLAPP leverages large language models (LLMs) and domain-specific retrieval to provide conversational coding support for CLASS-answering questions, generating code, debugging errors, and producing plots. Its architecture combines multi-agent LLM orchestration, semantic search across CLASS documentation, and a live Python execution environment. Deployed as a user-friendly web application, CLAPP lowers the entry barrier for scientists unfamiliar with AI tools and enables more productive human-AI collaboration in computational and numerical cosmology. The app is available at https://classclapp.streamlit.app", "authors": ["Santiago Casas", "Christian Fidler", "Boris Bolliet", "Francisco Villaescusa-Navarro", "Julien Lesgourgues"], "categories": ["astro-ph.IM", "astro-ph.CO", "cs.AI", "cs.MA"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2025-08-07", "url": "https://arxiv.org/abs/2508.05728", "pdf_url": "https://arxiv.org/pdf/2508.05728v1", "arxiv_id": "2508.05728", "doi": "10.48550/arXiv.2508.05728", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/santiagocasas/clapp", "venue": "arXiv.org", "quality_score": 0.402}
{"id": "52745b20928e44d54ba0573ce676cce09850b159869b750ce47525266a814ad4", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based Multi-Agent Copilot for Quantum Sensor", "abstract": "Large language models (LLM) exhibit broad utility but face limitations in quantum sensor development, stemming from interdisciplinary knowledge barriers and involving complex optimization processes. Here we present QCopilot, an LLM-based multi-agent framework integrating external knowledge access, active learning, and uncertainty quantification for quantum sensor design and diagnosis. Comprising commercial LLMs with few-shot prompt engineering and vector knowledge base, QCopilot employs specialized agents to adaptively select optimization methods, automate modeling analysis, and independently perform problem diagnosis. Applying QCopilot to atom cooling experiments, we generated 10${}^{\\rm{8}}$ sub-$\\rmμ$K atoms without any human intervention within a few hours, representing $\\sim$100$\\times$ speedup over manual experimentation. Notably, by continuously accumulating prior knowledge and enabling dynamic modeling, QCopilot can autonomously identify anomalous parameters in multi-parameter experimental settings. Our work reduces barriers to large-scale quantum sensor deployment and readily extends to other quantum information systems.", "authors": ["Rong Sha", "Binglin Wang", "Jun Yang", "Xiaoxiao Ma", "Chengkun Wu", "Liang Yan", "Chao Zhou", "Jixun Liu", "Guochao Wang", "Shuhua Yan", "Lingxiao Zhu"], "categories": ["quant-ph", "cs.AI", "physics.atom-ph"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2025-08-07", "url": "https://arxiv.org/abs/2508.05421", "pdf_url": "https://arxiv.org/pdf/2508.05421v1", "arxiv_id": "2508.05421", "doi": "10.48550/arXiv.2508.05421", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2601}
{"id": "32b93217e813bbd9a680249af75bdaf8bff8f59beb57de1859e70abcebbfe35b", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Collaboration With Multi-Agent Reinforcement Learning", "abstract": "A large amount of work has been done in Multi-Agent Systems (MAS) for modeling and solving problems with multiple interacting agents. However, most LLMs are pretrained independently and not specifically optimized for coordination. Existing LLM fine-tuning frameworks rely on individual rewards, which require complex reward designs for each agent to encourage collaboration. To address these challenges, we model LLM collaboration as a cooperative Multi-Agent Reinforcement Learning (MARL) problem. We develop a multi-agent, multi-turn algorithm, Multi-Agent Group Relative Policy Optimization (MAGRPO), to solve it, building on current RL approaches for LLMs as well as MARL techniques. Our experiments on LLM writing and coding collaboration demonstrate that fine-tuning MAS with MAGRPO enables agents to generate high-quality responses efficiently through effective cooperation. Our approach opens the door to using other MARL methods for LLMs and highlights the associated challenges. Our code is available at https://github.com/OpenMLRL/CoMLRL.", "authors": ["Shuo Liu", "Tianle Chen", "Zeyu Liang", "Xueguang Lyu", "Christopher Amato"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-06", "url": "https://arxiv.org/abs/2508.04652", "pdf_url": "https://arxiv.org/pdf/2508.04652v7", "arxiv_id": "2508.04652", "doi": "10.48550/arXiv.2508.04652", "citation_count": 29, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/OpenMLRL/CoMLRL", "venue": "arXiv.org", "quality_score": 0.4002}
{"id": "8433543403da393de44b456af80f7022ab10080b197919e221867777ae7a0672", "sources": ["arxiv", "semantic_scholar"], "title": "Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RL", "abstract": "Recent advances in large language models (LLMs) and multi-agent systems have demonstrated remarkable capabilities in complex problem-solving tasks such as deep research, vibe coding, and mathematical reasoning. However, most existing multi-agent systems are built upon manual prompt/workflow engineering with sophisticated agent frameworks, making them computationally inefficient, less capable, and can not benefit from data-centric learning. In this work, we introduce Chain-of-Agents (CoA), a novel paradigm of LLM reasoning that enables native end-to-end complex problem-solving in the same way as a multi-agent system (i.e., multi-turn problem solving with multiple tools and multiple agents) within one model. In chain-of-agents problem-solving, the model dynamically activates different tool agents and role-playing agents to simulate multi-agent collaboration in an end-to-end fashion. To elicit end-to-end chain-of-agents problem-solving abilities in LLMs, we introduce a multi-agent distillation framework to distill state-of-the-art multi-agent systems into chain-of-agents trajectories for agentic supervised fine-tuning. We then use agentic reinforcement learning on verifiable agentic tasks to further improve the models' capabilities on chain-of-agents problem solving. We call the resulting models Agent Foundation Models (AFMs). Our empirical studies demonstrate that AFM establishes new state-of-the-art performance across diverse benchmarks in both web agent and code agent settings. We make the entire research, including the model weights, code for training and evaluation, and the training data, fully open-sourced, which offers a solid starting point for future research on agent models and agentic RL.", "authors": ["Weizhen Li", "Jianbo Lin", "Zhuosong Jiang", "Jingyi Cao", "Xinpeng Liu", "Jiayu Zhang", "Zhenqiang Huang", "Qianben Chen", "Weichen Sun", "Qiexiang Wang", "Hongxuan Lu", "Tianrui Qin", "Chenghao Zhu", "Yi Yao", "Shuying Fan", "Xiaowan Li", "Tiannan Wang", "Pai Liu", "King Zhu", "He Zhu", "Dingfeng Shi", "Piaohong Wang", "Yeyi Guan", "Xiangru Tang", "Minghao Liu", "Yuchen Eleanor Jiang", "Jian Yang", "Jiaheng Liu", "Ge Zhang", "Wangchunshu Zhou"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-06", "url": "https://arxiv.org/abs/2508.13167", "pdf_url": "https://arxiv.org/pdf/2508.13167v1", "arxiv_id": "2508.13167", "doi": "10.48550/arXiv.2508.13167", "citation_count": 61, "influential_citation_count": 8, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4771}
{"id": "74a64ca3dad5c56fa7401a66cabd91e2597cfcc3e1fbb136cb6590d354963765", "sources": ["arxiv", "semantic_scholar"], "title": "Risk Analysis Techniques for Governed LLM-based Multi-Agent Systems", "abstract": "Organisations are starting to adopt LLM-based AI agents, with their deployments naturally evolving from single agents towards interconnected, multi-agent networks. Yet a collection of safe agents does not guarantee a safe collection of agents, as interactions between agents over time create emergent behaviours and induce novel failure modes. This means multi-agent systems require a fundamentally different risk analysis approach than that used for a single agent. This report addresses the early stages of risk identification and analysis for multi-agent AI systems operating within governed environments where organisations control their agent configurations and deployment. In this setting, we examine six critical failure modes: cascading reliability failures, inter-agent communication failures, monoculture collapse, conformity bias, deficient theory of mind, and mixed motive dynamics. For each, we provide a toolkit for practitioners to extend or integrate into their existing frameworks to assess these failure modes within their organisational contexts. Given fundamental limitations in current LLM behavioural understanding, our approach centres on analysis validity, and advocates for progressively increasing validity through staged testing across stages of abstraction and deployment that gradually increases exposure to potential negative impacts, while collecting convergent evidence through simulation, observational analysis, benchmarking, and red teaming. This methodology establishes the groundwork for robust organisational risk management as these LLM-based multi-agent systems are deployed and operated.", "authors": ["Alistair Reid", "Simon O'Callaghan", "Liam Carroll", "Tiberio Caetano"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-06", "url": "https://arxiv.org/abs/2508.05687", "pdf_url": "https://arxiv.org/pdf/2508.05687v1", "arxiv_id": "2508.05687", "doi": "10.48550/arXiv.2508.05687", "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "bf585859e2971daec96563523d91fcddce597c4778e4d3d0b69b1d198eeb581c", "sources": ["arxiv", "semantic_scholar"], "title": "RCR-Router: Efficient Role-Aware Context Routing for Multi-Agent LLM Systems with Structured Memory", "abstract": "Multi-agent large language model (LLM) systems have shown strong potential in complex reasoning and collaborative decision-making tasks. However, most existing coordination schemes rely on static or full-context routing strategies, which lead to excessive token consumption, redundant memory exposure, and limited adaptability across interaction rounds. We introduce RCR-Router, a modular and role-aware context routing framework designed to enable efficient, adaptive collaboration in multi-agent LLMs. To our knowledge, this is the first routing approach that dynamically selects semantically relevant memory subsets for each agent based on its role and task stage, while adhering to a strict token budget. A lightweight scoring policy guides memory selection, and agent outputs are iteratively integrated into a shared memory store to facilitate progressive context refinement. To better evaluate model behavior, we further propose an Answer Quality Score metric that captures LLM-generated explanations beyond standard QA accuracy. Experiments on three multi-hop QA benchmarks -- HotPotQA, MuSiQue, and 2WikiMultihop -- demonstrate that RCR-Router reduces token usage (up to 30%) while improving or maintaining answer quality. These results highlight the importance of structured memory routing and output-aware evaluation in advancing scalable multi-agent LLM systems.", "authors": ["Jun Liu", "Zhenglun Kong", "Changdi Yang", "Fan Yang", "Tianqi Li", "Peiyan Dong", "Joannah Nanjekye", "Hao Tang", "Geng Yuan", "Wei Niu", "Wenbin Zhang", "Pu Zhao", "Xue Lin", "Dong Huang", "Yanzhi Wang"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-06", "url": "https://arxiv.org/abs/2508.04903", "pdf_url": "https://arxiv.org/pdf/2508.04903v3", "arxiv_id": "2508.04903", "doi": "10.48550/arXiv.2508.04903", "citation_count": 14, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "b2357361f8907a8100b67a714aa73b93116cc542c5ecc7fa0973214f6b50a1a8", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Verifiable Misinformation Detection: A Multi-Tool LLM Agent Framework", "abstract": "With the proliferation of Large Language Models (LLMs), the detection of misinformation has become increasingly important and complex. This research proposes an innovative verifiable misinformation detection LLM agent that goes beyond traditional true/false binary judgments. The agent actively verifies claims through dynamic interaction with diverse web sources, assesses information source credibility, synthesizes evidence, and provides a complete verifiable reasoning process. Our designed agent architecture includes three core tools: precise web search tool, source credibility assessment tool and numerical claim verification tool. These tools enable the agent to execute multi-step verification strategies, maintain evidence logs, and form comprehensive assessment conclusions. We evaluate using standard misinformation datasets such as FakeNewsNet, comparing with traditional machine learning models and LLMs. Evaluation metrics include standard classification metrics, quality assessment of reasoning processes, and robustness testing against rewritten content. Experimental results show that our agent outperforms baseline methods in misinformation detection accuracy, reasoning transparency, and resistance to information rewriting, providing a new paradigm for trustworthy AI-assisted fact-checking.", "authors": ["Zikun Cui", "Tianyi Huang", "Chia-En Chiang", "Cuiqianhe Du"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-05", "url": "https://arxiv.org/abs/2508.03092", "pdf_url": "https://arxiv.org/pdf/2508.03092v1", "arxiv_id": "2508.03092", "doi": "10.1145/3766918.3766948", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1641}
{"id": "a5020b3e7816be2674be55ad937d9721334708c1f170f1a566886a3e10926d23", "sources": ["arxiv", "semantic_scholar"], "title": "Unified Tool Integration for LLMs: A Protocol-Agnostic Approach to Function Calling", "abstract": "The proliferation of tool-augmented Large Language Models (LLMs) has created a fragmented ecosystem where developers must navigate multiple protocols, manual schema definitions, and complex execution workflows. We address this challenge by proposing a unified approach to tool integration that abstracts protocol differences while optimizing execution performance. Our solution demonstrates how protocol-agnostic design principles can significantly reduce development overhead through automated schema generation, dual-mode concurrent execution, and seamless multi-source tool management. Experimental results show 60-80% code reduction across integration scenarios, performance improvements up to 3.1x through optimized concurrency, and full compatibility with existing function calling standards. This work contributes both theoretical insights into tool integration architecture and practical solutions for real-world LLM application development.", "authors": ["Peng Ding", "Rick Stevens"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-05", "url": "https://arxiv.org/abs/2508.02979", "pdf_url": "https://arxiv.org/pdf/2508.02979v1", "arxiv_id": "2508.02979", "doi": "10.48550/arXiv.2508.02979", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2578}
{"id": "ea3597a16befb59edddee5599a6e371601467ae638b0cd7a2ad55cc6d8b64efe", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Effective Offensive Security LLM Agents: Hyperparameter Tuning, LLM as a Judge, and a Lightweight CTF Benchmark", "abstract": "Recent advances in LLM agentic systems have improved the automation of offensive security tasks, particularly for Capture the Flag (CTF) challenges. We systematically investigate the key factors that drive agent success and provide a detailed recipe for building effective LLM-based offensive security agents. First, we present CTFJudge, a framework leveraging LLM as a judge to analyze agent trajectories and provide granular evaluation across CTF solving steps. Second, we propose a novel metric, CTF Competency Index (CCI) for partial correctness, revealing how closely agent solutions align with human-crafted gold standards. Third, we examine how LLM hyperparameters, namely temperature, top-p, and maximum token length, influence agent performance and automated cybersecurity task planning. For rapid evaluation, we present CTFTiny, a curated benchmark of 50 representative CTF challenges across binary exploitation, web, reverse engineering, forensics, and cryptography. Our findings identify optimal multi-agent coordination settings and lay the groundwork for future LLM agent research in cybersecurity. We make CTFTiny open source to public https://github.com/NYU-LLM-CTF/CTFTiny along with CTFJudge on https://github.com/NYU-LLM-CTF/CTFJudge.", "authors": ["Minghao Shao", "Nanda Rani", "Kimberly Milner", "Haoran Xi", "Meet Udeshi", "Saksham Aggarwal", "Venkata Sai Charan Putrevu", "Sandeep Kumar Shukla", "Prashanth Krishnamurthy", "Farshad Khorrami", "Ramesh Karri", "Muhammad Shafique"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-05", "url": "https://arxiv.org/abs/2508.05674", "pdf_url": "https://arxiv.org/pdf/2508.05674v2", "arxiv_id": "2508.05674", "doi": "10.48550/arXiv.2508.05674", "citation_count": 9, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/NYU-LLM-CTF/CTFTiny", "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3984}
{"id": "bda61aec62ae808f93a81143557297b79cfb04eb424e2a32f417d68d81c113bc", "sources": ["arxiv", "semantic_scholar"], "title": "Attack the Messages, Not the Agents: A Multi-round Adaptive Stealthy Tampering Framework for LLM-MAS", "abstract": "Large language model-based multi-agent systems (LLM-MAS) effectively accomplish complex and dynamic tasks through inter-agent communication, but this reliance introduces substantial safety vulnerabilities. Existing attack methods targeting LLM-MAS either compromise agent internals or rely on direct and overt persuasion, which limit their effectiveness, adaptability, and stealthiness. In this paper, we propose MAST, a Multi-round Adaptive Stealthy Tampering framework designed to exploit communication vulnerabilities within the system. MAST integrates Monte Carlo Tree Search with Direct Preference Optimization to train an attack policy model that adaptively generates effective multi-round tampering strategies. Furthermore, to preserve stealthiness, we impose dual semantic and embedding similarity constraints during the tampering process. Comprehensive experiments across diverse tasks, communication architectures, and LLMs demonstrate that MAST consistently achieves high attack success rates while significantly enhancing stealthiness compared to baselines. These findings highlight the effectiveness, stealthiness, and adaptability of MAST, underscoring the need for robust communication safeguards in LLM-MAS.", "authors": ["Bingyu Yan", "Ziyi Zhou", "Xiaoming Zhang", "Chaozhuo Li", "Ruilin Zeng", "Yirui Qi", "Tianbo Wang", "Litian Zhang"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-05", "url": "https://arxiv.org/abs/2508.03125", "pdf_url": "https://arxiv.org/pdf/2508.03125v1", "arxiv_id": "2508.03125", "doi": "10.48550/arXiv.2508.03125", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2578}
{"id": "08caf69fa8b7d64df0b428ff2517951691e4906d388f163ab0a24d369c6c31b6", "sources": ["arxiv", "semantic_scholar"], "title": "TransAM: Transformer-Based Agent Modeling for Multi-Agent Systems via Local Trajectory Encoding", "abstract": "Agent modeling is a critical component in developing effective policies within multi-agent systems, as it enables agents to form beliefs about the behaviors, intentions, and competencies of others. Many existing approaches assume access to other agents' episodic trajectories, a condition often unrealistic in real-world applications. Consequently, a practical agent modeling approach must learn a robust representation of the policies of the other agents based only on the local trajectory of the controlled agent. In this paper, we propose \\texttt{TransAM}, a novel transformer-based agent modeling approach to encode local trajectories into an embedding space that effectively captures the policies of other agents. We evaluate the performance of the proposed method in cooperative, competitive, and mixed multi-agent environments. Extensive experimental results demonstrate that our approach generates strong policy representations, improves agent modeling, and leads to higher episodic returns.", "authors": ["Conor Wallace", "Umer Siddique", "Yongcan Cao"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-04", "url": "https://arxiv.org/abs/2508.02826", "pdf_url": "https://arxiv.org/pdf/2508.02826v1", "arxiv_id": "2508.02826", "doi": "10.48550/arXiv.2508.02826", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2567}
{"id": "b34d7d96fd2dfe5ed2f2b8a2fcfc2bc5790214564280cbfd8757f42be783fd74", "sources": ["arxiv", "semantic_scholar"], "title": "Attractive Metadata Attack: Inducing LLM Agents to Invoke Malicious Tools", "abstract": "Large language model (LLM) agents have demonstrated remarkable capabilities in complex reasoning and decision-making by leveraging external tools. However, this tool-centric paradigm introduces a previously underexplored attack surface, where adversaries can manipulate tool metadata -- such as names, descriptions, and parameter schemas -- to influence agent behavior. We identify this as a new and stealthy threat surface that allows malicious tools to be preferentially selected by LLM agents, without requiring prompt injection or access to model internals. To demonstrate and exploit this vulnerability, we propose the Attractive Metadata Attack (AMA), a black-box in-context learning framework that generates highly attractive but syntactically and semantically valid tool metadata through iterative optimization. The proposed attack integrates seamlessly into standard tool ecosystems and requires no modification to the agent's execution framework. Extensive experiments across ten realistic, simulated tool-use scenarios and a range of popular LLM agents demonstrate consistently high attack success rates (81\\%-95\\%) and significant privacy leakage, with negligible impact on primary task execution. Moreover, the attack remains effective even against prompt-level defenses, auditor-based detection, and structured tool-selection protocols such as the Model Context Protocol, revealing systemic vulnerabilities in current agent architectures. These findings reveal that metadata manipulation constitutes a potent and stealthy attack surface. Notably, AMA is orthogonal to injection attacks and can be combined with them to achieve stronger attack efficacy, highlighting the need for execution-level defenses beyond prompt-level and auditor-based mechanisms. Code is available at https://github.com/SEAIC-M/AMA.", "authors": ["Kanghua Mo", "Li Hu", "Yucheng Long", "Zhihao Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-04", "url": "https://arxiv.org/abs/2508.02110", "pdf_url": "https://arxiv.org/pdf/2508.02110v2", "arxiv_id": "2508.02110", "doi": "10.48550/arXiv.2508.02110", "citation_count": 19, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/SEAIC-M/AMA", "venue": "arXiv.org", "quality_score": 0.3967}
{"id": "7398c4cd90443edb6c953844cbe2c3e1723ac97f60010dc00a621c6c10ee293c", "sources": ["arxiv", "semantic_scholar"], "title": "SE-Agent: Self-Evolution Trajectory Optimization in Multi-Step Reasoning with LLM-Based Agents", "abstract": "Large Language Model (LLM)-based agents have recently shown impressive capabilities in complex reasoning and tool use via multi-step interactions with their environments. While these agents have the potential to tackle complicated tasks, their problem-solving process, i.e., agents' interaction trajectory leading to task completion, remains underexploited. These trajectories contain rich feedback that can navigate agents toward the right directions for solving problems correctly. Although prevailing approaches, such as Monte Carlo Tree Search (MCTS), can effectively balance exploration and exploitation, they ignore the interdependence among various trajectories and lack the diversity of search spaces, which leads to redundant reasoning and suboptimal outcomes. To address these challenges, we propose SE-Agent, a Self-Evolution framework that enables Agents to optimize their reasoning processes iteratively. Our approach revisits and enhances former pilot trajectories through three key operations: revision, recombination, and refinement. This evolutionary mechanism enables two critical advantages: (1) it expands the search space beyond local optima by intelligently exploring diverse solution paths guided by previous trajectories, and (2) it leverages cross-trajectory inspiration to efficiently enhance performance while mitigating the impact of suboptimal reasoning paths. Through these mechanisms, SE-Agent achieves continuous self-evolution that incrementally improves reasoning quality. We evaluate SE-Agent on SWE-bench Verified to resolve real-world GitHub issues. Experimental results across five strong LLMs show that integrating SE-Agent delivers up to 55% relative improvement, achieving state-of-the-art performance among all open-source agents on SWE-bench Verified. Our code and demonstration materials are publicly available at https://github.com/JARVIS-Xs/SE-Agent.", "authors": ["Jiaye Lin", "Yifu Guo", "Yuzhen Han", "Sen Hu", "Ziyi Ni", "Licheng Wang", "Mingguang Chen", "Hongzhang Liu", "Ronghao Chen", "Yangfan He", "Daxin Jiang", "Binxing Jiao", "Chen Hu", "Huacan Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-08-04", "url": "https://arxiv.org/abs/2508.02085", "pdf_url": "https://arxiv.org/pdf/2508.02085v6", "arxiv_id": "2508.02085", "doi": "10.48550/arXiv.2508.02085", "citation_count": 44, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/JARVIS-Xs/SE-Agent", "venue": "arXiv.org", "quality_score": 0.4133}
{"id": "ca57505cf87d1284df7ed0ee7cfc9d2d422cece5005b1105c4bd03b0d35cb2f9", "sources": ["arxiv", "semantic_scholar"], "title": "Cued-Agent: A Collaborative Multi-Agent System for Automatic Cued Speech Recognition", "abstract": "Cued Speech (CS) is a visual communication system that combines lip-reading with hand coding to facilitate communication for individuals with hearing impairments. Automatic CS Recognition (ACSR) aims to convert CS hand gestures and lip movements into text via AI-driven methods. Traditionally, the temporal asynchrony between hand and lip movements requires the design of complex modules to facilitate effective multimodal fusion. However, constrained by limited data availability, current methods demonstrate insufficient capacity for adequately training these fusion mechanisms, resulting in suboptimal performance. Recently, multi-agent systems have shown promising capabilities in handling complex tasks with limited data availability. To this end, we propose the first collaborative multi-agent system for ACSR, named Cued-Agent. It integrates four specialized sub-agents: a Multimodal Large Language Model-based Hand Recognition agent that employs keyframe screening and CS expert prompt strategies to decode hand movements, a pretrained Transformer-based Lip Recognition agent that extracts lip features from the input video, a Hand Prompt Decoding agent that dynamically integrates hand prompts with lip features during inference in a training-free manner, and a Self-Correction Phoneme-to-Word agent that enables post-process and end-to-end conversion from phoneme sequences to natural language sentences for the first time through semantic refinement. To support this study, we expand the existing Mandarin CS dataset by collecting data from eight hearing-impaired cuers, establishing a mixed dataset of fourteen subjects. Extensive experiments demonstrate that our Cued-Agent performs superbly in both normal and hearing-impaired scenarios compared with state-of-the-art methods. The implementation is available at https://github.com/DennisHgj/Cued-Agent.", "authors": ["Guanjie Huang", "Danny H. K. Tsang", "Shan Yang", "Guangzhi Lei", "Li Liu"], "categories": ["cs.CV", "eess.AS"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-08-01", "url": "https://arxiv.org/abs/2508.00391", "pdf_url": "https://arxiv.org/pdf/2508.00391v1", "arxiv_id": "2508.00391", "doi": "10.1145/3746027.3755423", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/DennisHgj/Cued-Agent", "venue": "ACM Multimedia", "quality_score": 0.3914}
{"id": "9e2a04a4db96209f67dc0bd3d728034497be44799e09ec3cc0d3590d69621061", "sources": ["arxiv", "semantic_scholar"], "title": "A survey of multi-agent geosimulation methodologies: from ABM to LLM", "abstract": "We provide a comprehensive examination of agent-based approaches that codify the principles and linkages underlying multi-agent systems, simulations, and information systems. Based on two decades of study, this paper confirms a framework intended as a formal specification for geosimulation platforms. Our findings show that large language models (LLMs) can be effectively incorporated as agent components if they follow a structured architecture specific to fundamental agent activities such as perception, memory, planning, and action. This integration is precisely consistent with the architecture that we formalize, providing a solid platform for next-generation geosimulation systems.", "authors": ["Virginia Padilla", "Jacinto Dávila"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-31", "url": "https://arxiv.org/abs/2507.23694", "pdf_url": "https://arxiv.org/pdf/2507.23694v1", "arxiv_id": "2507.23694", "doi": "10.48550/arXiv.2507.23694", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2521}
{"id": "308035f471cc01998bafabf06181e9b0227f46e4f37549804c8e057b04acbe8d", "sources": ["arxiv", "semantic_scholar"], "title": "DynaSwarm: Dynamically Graph Structure Selection for LLM-based Multi-agent System", "abstract": "Current multi-agent systems (MAS) frameworks often rely on manually designed and static collaboration graph structures, limiting adaptability and performance. To address these limitations, we propose DynaSwarm, a dynamic framework that enhances LLM-based MAS through two key innovations: (1) an actor-critic reinforcement learning (A2C) mechanism to optimize graph structures with improved stability over prior RL methods, and (2) a dynamic graph selector that adaptively chooses the optimal graph structure for each input sample via parameter-efficient LLM fine-tuning. DynaSwarm eliminates the need for rigid, one-fits-all graph architectures, instead leveraging sample-specific idiosyncrasies to dynamically route queries through specialized agent networks. (c) We propose to fine-tune the demonstration retriever to fully exploit the power of in-context learning (ICL). Extensive experiments on question answering, mathematical reasoning, and coding tasks demonstrate that DynaSwarm consistently outperforms state-of-the-art single-agent and MAS baselines across multiple LLM backbones. Our findings highlight the importance of sample-aware structural flexibility in LLM MAS designs.", "authors": ["Hui Yi Leong", "Yuqing Wu"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-31", "url": "https://arxiv.org/abs/2507.23261", "pdf_url": "https://arxiv.org/pdf/2507.23261v2", "arxiv_id": "2507.23261", "doi": "10.48550/arXiv.2507.23261", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2521}
{"id": "16ebabdfe30e140ec9c7b7bc1011e264b8c9d779c30e6e0a832b295a35f0b971", "sources": ["arxiv", "semantic_scholar"], "title": "Distributed Average Consensus in Wireless Multi-Agent Systems with Over-the-Air Aggregation", "abstract": "In this paper, we address the average consensus problem of multi-agent systems over wireless networks. We propose a distributed average consensus algorithm by invoking the concept of over-the-air aggregation, which exploits the signal superposition property of wireless multiple-access channels. The proposed algorithm deploys a modified version of the well-known Ratio Consensus algorithm with an additional normalization step for compensating for the arbitrary channel coefficients. We show that, when the noise level at the receivers is negligible, the algorithm converges asymptotically to the average for time-invariant and time-varying channels. Numerical simulations corroborate the validity of our results.", "authors": ["Themistoklis Charalambous", "Zheng Chen", "Christoforos N. Hadjicostis"], "categories": ["eess.SY", "eess.SP"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-07-30", "url": "https://arxiv.org/abs/2507.22648", "pdf_url": "https://arxiv.org/pdf/2507.22648v1", "arxiv_id": "2507.22648", "doi": "10.1109/SPAWC60668.2024.10694509", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Workshop on Signal Processing Advances in Wireless Communications", "quality_score": 0.2509}
{"id": "8c69d02419ed54cb6b45cc64fe6abba25e1cfeb113e7f60a892733fa15378b00", "sources": ["arxiv", "semantic_scholar"], "title": "Strategic Communication and Language Bias in Multi-Agent LLM Coordination", "abstract": "Large Language Model (LLM)-based agents are increasingly deployed in multi-agent scenarios where coordination is crucial but not always assured. Research shows that the way strategic scenarios are framed linguistically can affect cooperation. This paper explores whether allowing agents to communicate amplifies these language-driven effects. Leveraging FAIRGAME, we simulate one-shot and repeated games across different languages and models, both with and without communication. Our experiments, conducted with two advanced LLMs-GPT-4o and Llama 4 Maverick-reveal that communication significantly influences agent behavior, though its impact varies by language, personality, and game structure. These findings underscore the dual role of communication in fostering coordination and reinforcing biases.", "authors": ["Alessio Buscemi", "Daniele Proverbio", "Alessandro Di Stefano", "The Anh Han", "German Castignani", "Pietro Liò"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-30", "url": "https://arxiv.org/abs/2508.00032", "pdf_url": "https://arxiv.org/pdf/2508.00032v2", "arxiv_id": "2508.00032", "doi": "10.48550/arXiv.2508.00032", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Workshop on Multi-disciplinary Trends in Artificial Intelligence", "quality_score": 0.2509}
{"id": "e93d9326fa8e66530b356eb00c7d7e65e7fe581e82ddd1579ec1de12f858bc70", "sources": ["arxiv", "semantic_scholar"], "title": "MASCA: LLM based-Multi Agents System for Credit Assessment", "abstract": "Recent advancements in financial problem-solving have leveraged LLMs and agent-based systems, with a primary focus on trading and financial modeling. However, credit assessment remains an underexplored challenge, traditionally dependent on rule-based methods and statistical models. In this paper, we introduce MASCA, an LLM-driven multi-agent system designed to enhance credit evaluation by mirroring real-world decision-making processes. The framework employs a layered architecture where specialized LLM-based agents collaboratively tackle sub-tasks. Additionally, we integrate contrastive learning for risk and reward assessment to optimize decision-making. We further present a signaling game theory perspective on hierarchical multi-agent systems, offering theoretical insights into their structure and interactions. Our paper also includes a detailed bias analysis in credit assessment, addressing fairness concerns. Experimental results demonstrate that MASCA outperforms baseline approaches, highlighting the effectiveness of hierarchical LLM-based multi-agent systems in financial applications, particularly in credit scoring.", "authors": ["Gautam Jajoo", "Pranjal A Chitale", "Saksham Agarwal"], "categories": ["cs.CL", "cs.CE", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-30", "url": "https://arxiv.org/abs/2507.22758", "pdf_url": "https://arxiv.org/pdf/2507.22758v1", "arxiv_id": "2507.22758", "doi": "10.48550/arXiv.2507.22758", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2509}
{"id": "72fdd6199e5a67d3824a3eca9ba846e4758c62709655124d29a90e489bd5e022", "sources": ["arxiv", "semantic_scholar"], "title": "MetaAgent: Automatically Constructing Multi-Agent Systems Based on Finite State Machines", "abstract": "Large Language Models (LLMs) have demonstrated the ability to solve a wide range of practical tasks within multi-agent systems. However, existing human-designed multi-agent frameworks are typically limited to a small set of pre-defined scenarios, while current automated design methods suffer from several limitations, such as the lack of tool integration, dependence on external training data, and rigid communication structures. In this paper, we propose MetaAgent, a finite state machine based framework that can automatically generate a multi-agent system. Given a task description, MetaAgent will design a multi-agent system and polish it through an optimization algorithm. When the multi-agent system is deployed, the finite state machine will control the agent's actions and the state transitions. To evaluate our framework, we conduct experiments on both text-based tasks and practical tasks. The results indicate that the generated multi-agent system surpasses other auto-designed methods and can achieve a comparable performance with the human-designed multi-agent system, which is optimized for those specific tasks.", "authors": ["Yaolun Zhang", "Xiaogeng Liu", "Chaowei Xiao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-30", "url": "https://arxiv.org/abs/2507.22606", "pdf_url": "https://arxiv.org/pdf/2507.22606v1", "arxiv_id": "2507.22606", "doi": "10.48550/arXiv.2507.22606", "citation_count": 12, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.2785}
{"id": "ae6192562ed1a59b7fd0a189433994fc35af79534f6f1eb04fadfa4bbd530692", "sources": ["arxiv", "semantic_scholar"], "title": "MemTool: Optimizing Short-Term Memory Management for Dynamic Tool Calling in LLM Agent Multi-Turn Conversations", "abstract": "Large Language Model (LLM) agents have shown significant autonomous capabilities in dynamically searching and incorporating relevant tools or Model Context Protocol (MCP) servers for individual queries. However, fixed context windows limit effectiveness in multi-turn interactions requiring repeated, independent tool usage. We introduce MemTool, a short-term memory framework enabling LLM agents to dynamically manage tools or MCP server contexts across multi-turn conversations. MemTool offers three agentic architectures: 1) Autonomous Agent Mode, granting full tool management autonomy, 2) Workflow Mode, providing deterministic control without autonomy, and 3) Hybrid Mode, combining autonomous and deterministic control. Evaluating each MemTool mode across 13+ LLMs on the ScaleMCP benchmark, we conducted experiments over 100 consecutive user interactions, measuring tool removal ratios (short-term memory efficiency) and task completion accuracy. In Autonomous Agent Mode, reasoning LLMs achieve high tool-removal efficiency (90-94% over a 3-window average), while medium-sized models exhibit significantly lower efficiency (0-60%). Workflow and Hybrid modes consistently manage tool removal effectively, whereas Autonomous and Hybrid modes excel at task completion. We present trade-offs and recommendations for each MemTool mode based on task accuracy, agency, and model capabilities.", "authors": ["Elias Lumer", "Anmol Gulati", "Vamse Kumar Subbiah", "Pradeep Honaganahalli Basavaraju", "James A. Burke"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-29", "url": "https://arxiv.org/abs/2507.21428", "pdf_url": "https://arxiv.org/pdf/2507.21428v1", "arxiv_id": "2507.21428", "doi": "10.48550/arXiv.2507.21428", "citation_count": 16, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "f637a7b17ec6ec88219c0007f3cc5bed2a4d338c1560195d58301ea93f778ea5", "sources": ["arxiv", "semantic_scholar"], "title": "CTG-Insight: A Multi-Agent Interpretable LLM Framework for Cardiotocography Analysis and Classification", "abstract": "Remote fetal monitoring technologies are becoming increasingly common. Yet, most current systems offer limited interpretability, leaving expectant parents with raw cardiotocography (CTG) data that is difficult to understand. In this work, we present CTG-Insight, a multi-agent LLM system that provides structured interpretations of fetal heart rate (FHR) and uterine contraction (UC) signals. Drawing from established medical guidelines, CTG-Insight decomposes each CTG trace into five medically defined features: baseline, variability, accelerations, decelerations, and sinusoidal pattern, each analyzed by a dedicated agent. A final aggregation agent synthesizes the outputs to deliver a holistic classification of fetal health, accompanied by a natural language explanation. We evaluate CTG-Insight on the NeuroFetalNet Dataset and compare it against deep learning models and the single-agent LLM baseline. Results show that CTG-Insight achieves state-of-the-art accuracy (96.4%) and F1-score (97.8%) while producing transparent and interpretable outputs. This work contributes an interpretable and extensible CTG analysis framework.", "authors": ["Black Sun", " Die", " Hu"], "categories": ["cs.LG", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-29", "url": "https://arxiv.org/abs/2507.22205", "pdf_url": "https://arxiv.org/pdf/2507.22205v1", "arxiv_id": "2507.22205", "doi": "10.1145/3714394.3756343", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.159}
{"id": "08d3a965d5f9c2a1663638ef5c9ae1109926984cb7376f6037b4ed43e23a1d2a", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Path Finding Among Dynamic Uncontrollable Agents with Statistical Safety Guarantees", "abstract": "Existing multi-agent path finding (MAPF) solvers do not account for uncertain behavior of uncontrollable agents. We present a novel variant of Enhanced Conflict-Based Search (ECBS), for both one-shot and lifelong MAPF in dynamic environments with uncontrollable agents. Our method consists of (1) training a learned predictor for the movement of uncontrollable agents, (2) quantifying the prediction error using conformal prediction (CP), a tool for statistical uncertainty quantification, and (3) integrating these uncertainty intervals into our modified ECBS solver. Our method can account for uncertain agent behavior, comes with statistical guarantees on collision-free paths for one-shot missions, and scales to lifelong missions with a receding horizon sequence of one-shot instances. We run our algorithm, CP-Solver, across warehouse and game maps, with competitive throughput and reduced collisions.", "authors": ["Kegan J. Strawn", "Thomy Phan", "Eric Wang", "Nora Ayanian", "Sven Koenig", "Lars Lindemann"], "categories": ["cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-29", "url": "https://arxiv.org/abs/2507.22282", "pdf_url": "https://arxiv.org/pdf/2507.22282v1", "arxiv_id": "2507.22282", "doi": "10.48550/arXiv.2507.22282", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2498}
{"id": "7a14f458f53d3fdefcf9e3b29c4c59ee157996ae42ccac6c89f4aec75e9bb431", "sources": ["arxiv", "semantic_scholar"], "title": "Physics-Informed EvolveGCN: Satellite Prediction for Multi Agent Systems", "abstract": "In the rapidly evolving domain of autonomous systems, interaction among agents within a shared environment is both inevitable and essential for enhancing overall system capabilities. A key requirement in such multi-agent systems is the ability of each agent to reliably predict the future positions of its nearest neighbors. Traditionally, graphs and graph theory have served as effective tools for modeling inter agent communication and relationships. While this approach is widely used, the present work proposes a novel method that leverages dynamic graphs in a forward looking manner. Specifically, the employment of EvolveGCN, a dynamic graph convolutional network, to forecast the evolution of inter-agent relationships over time. To improve prediction accuracy and ensure physical plausibility, this research incorporates physics constrained loss functions based on the Clohessy-Wiltshire equations of motion. This integrated approach enhances the reliability of future state estimations in multi-agent scenarios.", "authors": ["Timothy Jacob Huber", "Madhur Tiwari", "Camilo A. Riano-Rios"], "categories": ["cs.MA", "physics.space-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-07-29", "url": "https://arxiv.org/abs/2507.22279", "pdf_url": "https://arxiv.org/pdf/2507.22279v1", "arxiv_id": "2507.22279", "doi": "10.48550/arXiv.2507.22279", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2498}
{"id": "cb1c18966f7866ef97478b7e4675410f8f2d6a79bb8a66824f7e84974ab22ef7", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent-as-Judge: Aligning LLM-Agent-Based Automated Evaluation with Multi-Dimensional Human Evaluation", "abstract": "Nearly all human work is collaborative; thus, the evaluation of real-world NLP applications often requires multiple dimensions that align with diverse human perspectives. As real human evaluator resources are often scarce and costly, the emerging \"LLM-as-a-judge\" paradigm sheds light on a promising approach to leverage LLM agents to believably simulate human evaluators. Yet, to date, existing LLM-as-a-judge approaches face two limitations: persona descriptions of agents are often arbitrarily designed, and the frameworks are not generalizable to other tasks. To address these challenges, we propose MAJ-EVAL, a Multi-Agent-as-Judge evaluation framework that can automatically construct multiple evaluator personas with distinct dimensions from relevant text documents (e.g., research papers), instantiate LLM agents with the personas, and engage in-group debates with multi-agents to Generate multi-dimensional feedback. Our evaluation experiments in both the educational and medical domains demonstrate that MAJ-EVAL can generate evaluation results that better align with human experts' ratings compared with conventional automated evaluation metrics and existing LLM-as-a-judge methods.", "authors": ["Jiaju Chen", "Yuxuan Lu", "Xiaojie Wang", "Huimin Zeng", "Jing Huang", "Jiri Gesi", "Ying Xu", "Bingsheng Yao", "Dakuo Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-28", "url": "https://arxiv.org/abs/2507.21028", "pdf_url": "https://arxiv.org/pdf/2507.21028v1", "arxiv_id": "2507.21028", "doi": "10.48550/arXiv.2507.21028", "citation_count": 29, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "7d6ba9ea88305ed0384a1ebade6b1c31967ecc73078acbbaa32ebc3b2420ff1b", "sources": ["arxiv", "semantic_scholar"], "title": "SciToolAgent: A Knowledge Graph-Driven Scientific Agent for Multi-Tool Integration", "abstract": "Scientific research increasingly relies on specialized computational tools, yet effectively utilizing these tools demands substantial domain expertise. While Large Language Models (LLMs) show promise in tool automation, they struggle to seamlessly integrate and orchestrate multiple tools for complex scientific workflows. Here, we present SciToolAgent, an LLM-powered agent that automates hundreds of scientific tools across biology, chemistry, and materials science. At its core, SciToolAgent leverages a scientific tool knowledge graph that enables intelligent tool selection and execution through graph-based retrieval-augmented generation. The agent also incorporates a comprehensive safety-checking module to ensure responsible and ethical tool usage. Extensive evaluations on a curated benchmark demonstrate that SciToolAgent significantly outperforms existing approaches. Case studies in protein engineering, chemical reactivity prediction, chemical synthesis, and metal-organic framework screening further demonstrate SciToolAgent's capability to automate complex scientific workflows, making advanced research tools accessible to both experts and non-experts.", "authors": ["Keyan Ding", "Jing Yu", "Junjie Huang", "Yuchen Yang", "Qiang Zhang", "Huajun Chen"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science", "Medicine"], "published_date": "2025-07-27", "url": "https://arxiv.org/abs/2507.20280", "pdf_url": "https://arxiv.org/pdf/2507.20280v1", "arxiv_id": "2507.20280", "doi": "10.1038/s43588-025-00849-y", "citation_count": 51, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Nature Computational Science", "quality_score": 0.429}
{"id": "24a2110ee071b3693d2d7cff904321145a26617b318e5d363aa7f1431de19117", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Multi-Agent Economies: Enhancing the A2A Protocol with Ledger-Anchored Identities and x402 Micropayments for AI Agents", "abstract": "This research article presents a novel architecture to empower multi-agent economies by addressing two critical limitations of the emerging Agent2Agent (A2A) communication protocol: decentralized agent discoverability and agent-to-agent micropayments. By integrating distributed ledger technology (DLT), this architecture enables tamper-proof, on-chain publishing of AgentCards as smart contracts, providing secure and verifiable agent identities. The architecture further extends A2A with the x402 open standard, facilitating blockchain-agnostic, HTTP-based micropayments via the HTTP 402 status code. This enables autonomous agents to seamlessly discover, authenticate, and compensate each other across organizational boundaries. This work further presents a comprehensive technical implementation and evaluation, demonstrating the feasibility of DLT-based agent discovery and micropayments. The proposed approach lays the groundwork for secure, scalable, and economically viable multi-agent ecosystems, advancing the field of agentic AI toward trusted, autonomous economic interactions.", "authors": ["Awid Vaziry", "Sandro Rodriguez Garzon", "Axel Küpper"], "categories": ["cs.MA", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-24", "url": "https://arxiv.org/abs/2507.19550", "pdf_url": "https://arxiv.org/pdf/2507.19550v1", "arxiv_id": "2507.19550", "doi": "10.1007/978-3-032-15632-7_25", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Joint Conference on Computational Intelligence", "quality_score": 0.2441}
{"id": "8023223b487aec182ef8b35c7498d1fe41fa5ddb159b3ad93744ceef337f17b8", "sources": ["arxiv", "semantic_scholar"], "title": "Agent WARPP: Workflow Adherence via Runtime Parallel Personalization", "abstract": "Large language models (LLMs) are increasingly applied in task-oriented dialogue (TOD) systems but often struggle with long, conditional workflows that involve external tool calls and depend on user-specific information. We present Workflow Adherence via Runtime Parallel Personalization, or WARPP, a training-free, modular framework that combines multi-agent orchestration with runtime personalization to improve workflow adherence in LLM-based systems. By dynamically pruning conditional branches based on user attributes, the framework reduces reasoning overhead and narrows tool selection at runtime. WARPP deploys a parallelized architecture where a dedicated Personalizer agent operates alongside modular, domain-specific agents to dynamically tailor execution paths in real time. The framework is evaluated across five representative user intents of varying complexity within three domains: banking, flights, and healthcare. Our evaluation leverages synthetic datasets and LLM-powered simulated users to test scenarios with conditional dependencies. Our results demonstrate that WARPP outperforms both the non-personalized method and the ReAct baseline, achieving increasingly larger gains in parameter fidelity and tool accuracy as intent complexity grows, while also reducing average token usage, without any additional training.", "authors": ["Maria Emilia Mazzolenis", "Ruirui Zhang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-23", "url": "https://arxiv.org/abs/2507.19543", "pdf_url": "https://arxiv.org/pdf/2507.19543v1", "arxiv_id": "2507.19543", "doi": "10.48550/arXiv.2507.19543", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/emiliamazzo/WARPP/", "venue": "arXiv.org", "quality_score": 0.3754}
{"id": "e7ffe95f12111a632b8c87d93c80174c4029f2f7d4e920e0225eed3471816cb4", "sources": ["arxiv"], "title": "Resilient Multi-Agent Negotiation for Medical Supply Chains:Integrating LLMs and Blockchain for Transparent Coordination", "abstract": "Global health emergencies, such as the COVID-19 pandemic, have exposed critical weaknesses in traditional medical supply chains, including inefficiencies in resource allocation, lack of transparency, and poor adaptability to dynamic disruptions. This paper presents a novel hybrid framework that integrates blockchain technology with a decentralized, large language model (LLM) powered multi-agent negotiation system to enhance the resilience and accountability of medical supply chains during crises. In this system, autonomous agents-representing manufacturers, distributors, and healthcare institutions-engage in structured, context-aware negotiation and decision-making processes facilitated by LLMs, enabling rapid and ethical allocation of scarce medical resources. The off-chain agent layer supports adaptive reasoning and local decision-making, while the on-chain blockchain layer ensures immutable, transparent, and auditable enforcement of decisions via smart contracts. The framework also incorporates a formal cross-layer communication protocol to bridge decentralized negotiation with institutional enforcement. A simulation environment emulating pandemic scenarios evaluates the system's performance, demonstrating improvements in negotiation efficiency, fairness of allocation, supply chain responsiveness, and auditability. This research contributes an innovative approach that synergizes blockchain trust guarantees with the adaptive intelligence of LLM-driven agents, providing a robust and scalable solution for critical supply chain coordination under uncertainty.", "authors": ["Mariam ALMutairi", "Hyungmin Kim"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": [], "published_date": "2025-07-23", "url": "https://arxiv.org/abs/2507.17134", "pdf_url": "https://arxiv.org/pdf/2507.17134v1", "arxiv_id": "2507.17134", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1546}
{"id": "70187379b47e92966922c2276a4ff79ca931dbbd8b24c42114d44ba7f3ee0572", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Unifying Quantitative Security Benchmarking for Multi Agent Systems", "abstract": "Evolving AI systems increasingly deploy multi-agent architectures where autonomous agents collaborate, share information, and delegate tasks through developing protocols. This connectivity, while powerful, introduces novel security risks. One such risk is a cascading risk: a breach in one agent can cascade through the system, compromising others by exploiting inter-agent trust. In tandem with OWASP's initiative for an Agentic AI Vulnerability Scoring System we define an attack vector, Agent Cascading Injection, analogous to Agent Impact Chain and Blast Radius, operating across networks of agents. In an ACI attack, a malicious input or tool exploit injected at one agent leads to cascading compromises and amplified downstream effects across agents that trust its outputs. We formalize this attack with an adversarial goal equation and key variables (compromised agent, injected exploit, polluted observations, etc.), capturing how a localized vulnerability can escalate into system-wide failure. We then analyze ACI's properties -- propagation chains, amplification factors, and inter-agent compound effects -- and map these to OWASP's emerging Agentic AI risk categories (e.g. Impact Chain and Orchestration Exploits). Finally, we argue that ACI highlights a critical need for quantitative benchmarking frameworks to evaluate the security of agent-to-agent communication protocols. We outline a methodology for stress-testing multi-agent systems (using architectures such as Google's A2A and Anthropic's MCP) against cascading trust failures, developing upon groundwork for measurable, standardized agent-to-agent security evaluation. Our work provides the necessary apparatus for engineers to benchmark system resilience, make data-driven architectural trade-offs, and develop robust defenses against a new generation of agentic threats.", "authors": ["Gauri Sharma", "Vidhi Kulkarni", "Miles King", "Ken Huang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-23", "url": "https://arxiv.org/abs/2507.21146", "pdf_url": "https://arxiv.org/pdf/2507.21146v1", "arxiv_id": "2507.21146", "doi": "10.48550/arXiv.2507.21146", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2429}
{"id": "bad780e0be3bc00a73b2c0e85bc46d17880769de17744d768914e493425173ca", "sources": ["arxiv", "semantic_scholar"], "title": "Parallelism Meets Adaptiveness: Scalable Documents Understanding in Multi-Agent LLM Systems", "abstract": "Large language model (LLM) agents have shown increasing promise for collaborative task completion. However, existing multi-agent frameworks often rely on static workflows, fixed roles, and limited inter-agent communication, reducing their effectiveness in open-ended, high-complexity domains. This paper proposes a coordination framework that enables adaptiveness through three core mechanisms: dynamic task routing, bidirectional feedback, and parallel agent evaluation. The framework allows agents to reallocate tasks based on confidence and workload, exchange structured critiques to iteratively improve outputs, and crucially compete on high-ambiguity subtasks with evaluator-driven selection of the most suitable result. We instantiate these principles in a modular architecture and demonstrate substantial improvements in factual coverage, coherence, and efficiency over static and partially adaptive baselines. Our findings highlight the benefits of incorporating both adaptiveness and structured competition in multi-agent LLM systems.", "authors": ["Chengxuan Xia", "Qianye Wu", "Sixuan Tian", "Yilun Hao"], "categories": ["cs.MA", "cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-22", "url": "https://arxiv.org/abs/2507.17061", "pdf_url": "https://arxiv.org/pdf/2507.17061v5", "arxiv_id": "2507.17061", "doi": "10.48550/arXiv.2507.17061", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2418}
{"id": "833c00399308fd04216a6a8cb2bb17294e78312061621fac9764497a3db2d7f2", "sources": ["arxiv", "semantic_scholar"], "title": "MobileUse: A GUI Agent with Hierarchical Reflection for Autonomous Mobile Operation", "abstract": "Recent advances in Multimodal Large Language Models (MLLMs) have enabled the development of mobile agents that can understand visual inputs and follow user instructions, unlocking new possibilities for automating complex tasks on mobile devices. However, applying these models to real-world mobile scenarios remains a significant challenge due to the long-horizon task execution, difficulty in error recovery, and the cold-start problem in unfamiliar environments. To address these challenges, we propose MobileUse, a GUI agent designed for robust and adaptive mobile task execution. To improve resilience in long-horizon tasks and dynamic environments, we introduce a hierarchical reflection architecture that enables the agent to self-monitor, detect, and recover from errors across multiple temporal scales-ranging from individual actions to overall task completion-while maintaining efficiency through a reflection-on-demand strategy. To tackle cold-start issues, we further introduce a proactive exploration module, which enriches the agent's understanding of the environment through self-planned exploration. Evaluations on AndroidWorld and AndroidLab benchmarks demonstrate that MobileUse establishes new state-of-the-art performance, achieving success rates of 62.9% and 44.2%, respectively. To facilitate real-world applications, we release an out-of-the-box toolkit for automated task execution on physical mobile devices, which is available at https://github.com/MadeAgents/mobile-use.", "authors": ["Ning Li", "Xiangmou Qu", "Jiamu Zhou", "Jun Wang", "Muning Wen", "Kounianhua Du", "Xingyu Lou", "Qiuying Peng", "Jun Wang", "Weinan Zhang"], "categories": ["cs.RO", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-21", "url": "https://arxiv.org/abs/2507.16853", "pdf_url": "https://arxiv.org/pdf/2507.16853v1", "arxiv_id": "2507.16853", "doi": "10.48550/arXiv.2507.16853", "citation_count": 30, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/MadeAgents/mobile-use", "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "8f4d162695a53369213090585dca78c49c9c0391422a309fa5a303076d44c3ba", "sources": ["arxiv", "semantic_scholar"], "title": "IM-Chat: A Multi-agent LLM Framework Integrating Tool-Calling and Diffusion Modeling for Knowledge Transfer in Injection Molding Industry", "abstract": "The injection molding industry faces critical challenges in preserving and transferring field knowledge, particularly as experienced workers retire and multilingual barriers hinder effective communication. This study introduces IM-Chat, a multi-agent framework based on large language models (LLMs), designed to facilitate knowledge transfer in injection molding. IM-Chat integrates both limited documented knowledge (e.g., troubleshooting tables, manuals) and extensive field data modeled through a data-driven process condition generator that infers optimal manufacturing settings from environmental inputs such as temperature and humidity, enabling robust and context-aware task resolution. By adopting a retrieval-augmented generation (RAG) strategy and tool-calling agents within a modular architecture, IM-Chat ensures adaptability without the need for fine-tuning. Performance was assessed across 100 single-tool and 60 hybrid tasks for GPT-4o, GPT-4o-mini, and GPT-3.5-turbo by domain experts using a 10-point rubric focused on relevance and correctness, and was further supplemented by automated evaluation using GPT-4o guided by a domain-adapted instruction prompt. The evaluation results indicate that more capable models tend to achieve higher accuracy, particularly in complex, tool-integrated scenarios. In addition, compared with the fine-tuned single-agent LLM, IM-Chat demonstrated superior accuracy, particularly in quantitative reasoning, and greater scalability in handling multiple information sources. Overall, these findings demonstrate the viability of multi-agent LLM systems for industrial knowledge workflows and establish IM-Chat as a scalable and generalizable approach to AI-assisted decision support in manufacturing.", "authors": ["Junhyeong Lee", "Joon-Young Kim", "Heekyu Kim", "Inhyo Lee", "Seunghwa Ryu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-21", "url": "https://arxiv.org/abs/2507.15268", "pdf_url": "https://arxiv.org/pdf/2507.15268v2", "arxiv_id": "2507.15268", "doi": "10.1016/j.jmsy.2025.11.007", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Journal of manufacturing systems", "quality_score": 0.2406}
{"id": "5f1cde2d9fc1df87915e3a0bbef54366914371bf35d537585f66432f5ceeccad", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Economist: Large Population Models and Mechanism Design in Multi-Agent Generative Simulacra", "abstract": "We present the LLM Economist, a novel framework that uses agent-based modeling to design and assess economic policies in strategic environments with hierarchical decision-making. At the lower level, bounded rational worker agents -- instantiated as persona-conditioned prompts sampled from U.S. Census-calibrated income and demographic statistics -- choose labor supply to maximize text-based utility functions learned in-context. At the upper level, a planner agent employs in-context reinforcement learning to propose piecewise-linear marginal tax schedules anchored to the current U.S. federal brackets. This construction endows economic simulacra with three capabilities requisite for credible fiscal experimentation: (i) optimization of heterogeneous utilities, (ii) principled generation of large, demographically realistic agent populations, and (iii) mechanism design -- the ultimate nudging problem -- expressed entirely in natural language. Experiments with populations of up to one hundred interacting agents show that the planner converges near Stackelberg equilibria that improve aggregate social welfare relative to Saez solutions, while a periodic, persona-level voting procedure furthers these gains under decentralized governance. These results demonstrate that large language model-based agents can jointly model, simulate, and govern complex economic systems, providing a tractable test bed for policy evaluation at the societal scale to help build better civilizations.", "authors": ["Seth Karten", "Wenzhe Li", "Zihan Ding", "Samuel Kleiner", "Yu Bai", "Chi Jin"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-21", "url": "https://arxiv.org/abs/2507.15815", "pdf_url": "https://arxiv.org/pdf/2507.15815v1", "arxiv_id": "2507.15815", "doi": "10.48550/arXiv.2507.15815", "citation_count": 20, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/sethkarten/LLM-Economist", "venue": "arXiv.org", "quality_score": 0.3719}
{"id": "b3244e73fa500cd0fbd5f7507a8d531142ec384584deba96a463e9601bbd3b92", "sources": ["arxiv", "semantic_scholar"], "title": "Byzantine-Robust Decentralized Coordination of LLM Agents", "abstract": "Collaboration among multiple large language model (LLM) agents is a promising approach to overcome inherent limitations of single-agent systems, such as hallucinations and single points of failure. As LLM agents are increasingly deployed on open blockchain platforms, multi-agent systems capable of tolerating malicious (Byzantine) agents have become essential. Recent Byzantine-robust multi-agent systems typically rely on leader-driven coordination, which suffers from two major drawbacks. First, they are inherently vulnerable to targeted attacks against the leader. If consecutive leaders behave maliciously, the system repeatedly fails to achieve consensus, forcing new consensus rounds, which is particularly costly given the high latency of LLM invocations. Second, an underperforming proposal from the leader can be accepted as the final answer even when higher-quality alternatives are available, as existing methods finalize the leader's proposal once it receives a quorum of votes. To address these issues, we propose DecentLLMs, a novel decentralized consensus approach for multi-agent LLM systems, where worker agents generate answers concurrently and evaluator agents independently score and rank these answers to select the best available one. This decentralized architecture enables faster consensus despite the presence of Byzantine agents and consistently selects higher-quality answers through Byzantine-robust aggregation techniques. Experimental results demonstrate that DecentLLMs effectively tolerates Byzantine agents and significantly improves the quality of selected answers.", "authors": ["Yongrae Jo", "Chanik Park"], "categories": ["cs.DC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-20", "url": "https://arxiv.org/abs/2507.14928", "pdf_url": "https://arxiv.org/pdf/2507.14928v1", "arxiv_id": "2507.14928", "doi": "10.48550/arXiv.2507.14928", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2395}
{"id": "58b32ee58454c084ae573d61bdcb74c1215e69ace27b2c8eed9e11fc37f87c29", "sources": ["arxiv", "semantic_scholar"], "title": "Hierarchical Multi-Agent Reinforcement Learning with Control Barrier Functions for Safety-Critical Autonomous Systems", "abstract": "We address the problem of safe policy learning in multi-agent safety-critical autonomous systems. In such systems, it is necessary for each agent to meet the safety requirements at all times while also cooperating with other agents to accomplish the task. Toward this end, we propose a safe Hierarchical Multi-Agent Reinforcement Learning (HMARL) approach based on Control Barrier Functions (CBFs). Our proposed hierarchical approach decomposes the overall reinforcement learning problem into two levels learning joint cooperative behavior at the higher level and learning safe individual behavior at the lower or agent level conditioned on the high-level policy. Specifically, we propose a skill-based HMARL-CBF algorithm in which the higher level problem involves learning a joint policy over the skills for all the agents and the lower-level problem involves learning policies to execute the skills safely with CBFs. We validate our approach on challenging environment scenarios whereby a large number of agents have to safely navigate through conflicting road networks. Compared with existing state of the art methods, our approach significantly improves the safety achieving near perfect (within 5%) success/safety rate while also improving performance across all the environments.", "authors": ["H. M. Sabbir Ahmad", "Ehsan Sabouni", "Alexander Wasilkoff", "Param Budhraja", "Zijian Guo", "Songyuan Zhang", "Chuchu Fan", "Christos Cassandras", "Wenchao Li"], "categories": ["cs.LG", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-20", "url": "https://arxiv.org/abs/2507.14850", "pdf_url": "https://arxiv.org/pdf/2507.14850v2", "arxiv_id": "2507.14850", "doi": "10.48550/arXiv.2507.14850", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2395}
{"id": "8ac9a4efd93dca9faecc36b8929dc18d26e9a6cc94c12c6da3889647b1f48c68", "sources": ["arxiv", "semantic_scholar"], "title": "Configurable multi-agent framework for scalable and realistic testing of llm-based agents", "abstract": "Large-language-model (LLM) agents exhibit complex, context-sensitive behaviour that quickly renders static benchmarks and ad-hoc manual testing obsolete. We present Neo, a configurable, multi-agent framework that automates realistic, multi-turn evaluation of LLM-based systems. Neo couples a Question Generation Agent and an Evaluation Agent through a shared context-hub, allowing domain prompts, scenario controls and dynamic feedback to be composed modularly. Test inputs are sampled from a probabilistic state model spanning dialogue flow, user intent and emotional tone, enabling diverse, human-like conversations that adapt after every turn. Applied to a production-grade Seller Financial Assistant chatbot, Neo (i) uncovered edge-case failures across five attack categories with a 3.3% break rate close to the 5.8% achieved by expert human red-teamers, and (ii) delivered 10-12X higher throughput, generating 180 coherent test questions in around 45 mins versus 16h of human effort. Beyond security probing, Neo's stochastic policies balanced topic coverage and conversational depth, yielding broader behavioural exploration than manually crafted scripts. Neo therefore lays a foundation for scalable, self-evolving LLM QA: its agent interfaces, state controller and feedback loops are model-agnostic and extensible to richer factual-grounding and policy-compliance checks. We release the framework to facilitate reproducible, high-fidelity testing of emerging agentic systems.", "authors": ["Sai Wang", "Senthilnathan Subramanian", "Mudit Sahni", "Praneeth Gone", "Lingjie Meng", "Xiaochen Wang", "Nicolas Ferradas Bertoli", "Tingxian Cheng", "Jun Xu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-19", "url": "https://arxiv.org/abs/2507.14705", "pdf_url": "https://arxiv.org/pdf/2507.14705v1", "arxiv_id": "2507.14705", "doi": "10.48550/arXiv.2507.14705", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2383}
{"id": "cc198f41b73e0bc7fd9a148319dfa4ef714bf4d199772baa0de9d70ac5b66f4f", "sources": ["arxiv"], "title": "WSI-Agents: A Collaborative Multi-Agent System for Multi-Modal Whole Slide Image Analysis", "abstract": "Whole slide images (WSIs) are vital in digital pathology, enabling gigapixel tissue analysis across various pathological tasks. While recent advancements in multi-modal large language models (MLLMs) allow multi-task WSI analysis through natural language, they often underperform compared to task-specific models. Collaborative multi-agent systems have emerged as a promising solution to balance versatility and accuracy in healthcare, yet their potential remains underexplored in pathology-specific domains. To address these issues, we propose WSI-Agents, a novel collaborative multi-agent system for multi-modal WSI analysis. WSI-Agents integrates specialized functional agents with robust task allocation and verification mechanisms to enhance both task-specific accuracy and multi-task versatility through three components: (1) a task allocation module assigning tasks to expert agents using a model zoo of patch and WSI level MLLMs, (2) a verification mechanism ensuring accuracy through internal consistency checks and external validation using pathology knowledge bases and domain-specific models, and (3) a summary module synthesizing the final summary with visual interpretation maps. Extensive experiments on multi-modal WSI benchmarks show WSI-Agents's superiority to current WSI MLLMs and medical agent frameworks across diverse tasks.", "authors": ["Xinheng Lyu", "Yuci Liang", "Wenting Chen", "Meidan Ding", "Jiaqi Yang", "Guolin Huang", "Daokun Zhang", "Xiangjian He", "Linlin Shen"], "categories": ["cs.CV", "cs.AI"], "fields_of_study": [], "published_date": "2025-07-19", "url": "https://arxiv.org/abs/2507.14680", "pdf_url": "https://arxiv.org/pdf/2507.14680v1", "arxiv_id": "2507.14680", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1517}
{"id": "c38da9ed51fff98eb8593e0e8a1887b5650bb3ce1cf6b8e44c33a9ccdc997bd5", "sources": ["arxiv", "semantic_scholar"], "title": "CodeEdu: A Multi-Agent Collaborative Platform for Personalized Coding Education", "abstract": "Large Language Models (LLMs) have demonstrated considerable potential in improving coding education by providing support for code writing, explanation, and debugging. However, existing LLM-based approaches generally fail to assess students' abilities, design learning plans, provide personalized material aligned with individual learning goals, and enable interactive learning. Current work mostly uses single LLM agents, which limits their ability to understand complex code repositories and schedule step-by-step tutoring. Recent research has shown that multi-agent LLMs can collaborate to solve complicated problems in various domains like software engineering, but their potential in the field of education remains unexplored. In this work, we introduce CodeEdu, an innovative multi-agent collaborative platform that combines LLMs with tool use to provide proactive and personalized education in coding. Unlike static pipelines, CodeEdu dynamically allocates agents and tasks to meet student needs. Various agents in CodeEdu undertake certain functions specifically, including task planning, personalized material generation, real-time QA, step-by-step tutoring, code execution, debugging, and learning report generation, facilitated with extensive external tools to improve task efficiency. Automated evaluations reveal that CodeEdu substantially enhances students' coding performance.", "authors": ["Jianing Zhao", "Peng Gao", "Jiannong Cao", "Zhiyuan Wen", "Chen Chen", "Jianing Yin", "Ruosong Yang", "Bo Yuan"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-18", "url": "https://arxiv.org/abs/2507.13814", "pdf_url": "https://arxiv.org/pdf/2507.13814v1", "arxiv_id": "2507.13814", "doi": "10.48550/arXiv.2507.13814", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2372}
{"id": "cf2d2810105d07114eca68a85a3c6852ee0940fe5a8c748612f5d24c0d32fd95", "sources": ["arxiv", "semantic_scholar"], "title": "Aime: Towards Fully-Autonomous Multi-Agent Framework", "abstract": "Multi-Agent Systems (MAS) powered by Large Language Models (LLMs) are emerging as a powerful paradigm for solving complex, multifaceted problems. However, the potential of these systems is often constrained by the prevalent plan-and-execute framework, which suffers from critical limitations: rigid plan execution, static agent capabilities, and inefficient communication. These weaknesses hinder their adaptability and robustness in dynamic environments. This paper introduces Aime, a novel multi-agent framework designed to overcome these challenges through dynamic, reactive planning and execution. Aime replaces the conventional static workflow with a fluid and adaptive architecture. Its core innovations include: (1) a Dynamic Planner that continuously refines the overall strategy based on real-time execution feedback; (2) an Actor Factory that implements Dynamic Actor instantiation, assembling specialized agents on-demand with tailored tools and knowledge; and (3) a centralized Progress Management Module that serves as a single source of truth for coherent, system-wide state awareness. We empirically evaluated Aime on a diverse suite of benchmarks spanning general reasoning (GAIA), software engineering (SWE-bench Verified), and live web navigation (WebVoyager). The results demonstrate that Aime consistently outperforms even highly specialized state-of-the-art agents in their respective domains. Its superior adaptability and task success rate establish Aime as a more resilient and effective foundation for multi-agent collaboration.", "authors": ["Yexuan Shi", "Mingyu Wang", "Yunxiang Cao", "Hongjie Lai", "Junjian Lan", "Xin Han", "Yu Wang", "Jie Geng", "Zhenan Li", "Zihao Xia", "Xiang Chen", "Chen Li", "Jian Xu", "Wenbo Duan", "Yuanshuo Zhu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-16", "url": "https://arxiv.org/abs/2507.11988", "pdf_url": "https://arxiv.org/pdf/2507.11988v2", "arxiv_id": "2507.11988", "doi": "10.48550/arXiv.2507.11988", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2349}
{"id": "d6c447b2a57b5c1d7b2b06e01eec52448bc780e3a0f11cf67a012cae720330ec", "sources": ["arxiv", "semantic_scholar"], "title": "Lessons Learned from Evaluation of LLM based Multi-agents in Safer Therapy Recommendation", "abstract": "Therapy recommendation for chronic patients with multimorbidity is challenging due to risks of treatment conflicts. Existing decision support systems face scalability limitations. Inspired by the way in which general practitioners (GP) manage multimorbidity patients, occasionally convening multidisciplinary team (MDT) collaboration, this study investigated the feasibility and value of using a Large Language Model (LLM)-based multi-agent system (MAS) for safer therapy recommendations. We designed a single agent and a MAS framework simulating MDT decision-making by enabling discussion among LLM agents to resolve medical conflicts. The systems were evaluated on therapy planning tasks for multimorbidity patients using benchmark cases. We compared MAS performance with single-agent approaches and real-world benchmarks. An important contribution of our study is the definition of evaluation metrics that go beyond the technical precision and recall and allow the inspection of clinical goals met and medication burden of the proposed advices to a gold standard benchmark. Our results show that with current LLMs, a single agent GP performs as well as MDTs. The best-scoring models provide correct recommendations that address all clinical goals, yet the advices are incomplete. Some models also present unnecessary medications, resulting in unnecessary conflicts between medication and conditions or drug-drug interactions.", "authors": ["Yicong Wu", "Ting Chen", "Irit Hochberg", "Zhoujian Sun", "Ruth Edry", "Zhengxing Huang", "Mor Peleg"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-15", "url": "https://arxiv.org/abs/2507.10911", "pdf_url": "https://arxiv.org/pdf/2507.10911v1", "arxiv_id": "2507.10911", "doi": "10.48550/arXiv.2507.10911", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2338}
{"id": "583805b115430efa316aeac9a2159258229de3fca2ad8f3960712e0abae4f18b", "sources": ["arxiv", "semantic_scholar"], "title": "Warehouse Spatial Question Answering with LLM Agent", "abstract": "Spatial understanding has been a challenging task for existing Multi-modal Large Language Models~(MLLMs). Previous methods leverage large-scale MLLM finetuning to enhance MLLM's spatial understanding ability. In this paper, we present a data-efficient approach. We propose a LLM agent system with strong and advanced spatial reasoning ability, which can be used to solve the challenging spatial question answering task in complex indoor warehouse scenarios. Our system integrates multiple tools that allow the LLM agent to conduct spatial reasoning and API tools interaction to answer the given complicated spatial question. Extensive evaluations on the 2025 AI City Challenge Physical AI Spatial Intelligence Warehouse dataset demonstrate that our system achieves high accuracy and efficiency in tasks such as object retrieval, counting, and distance estimation. The code is available at: https://github.com/hsiangwei0903/SpatialAgent", "authors": ["Hsiang-Wei Huang", "Jen-Hao Cheng", "Kuang-Ming Chen", "Cheng-Yen Yang", "Bahaa Alattar", "Yi-Ru Lin", "Pyongkun Kim", "Sangwon Kim", "Kwangju Kim", "Chung-I Huang", "Jenq-Neng Hwang"], "categories": ["cs.CV", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-14", "url": "https://arxiv.org/abs/2507.10778", "pdf_url": "https://arxiv.org/pdf/2507.10778v2", "arxiv_id": "2507.10778", "doi": "10.1109/ICCVW69036.2025.00550", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/hsiangwei0903/SpatialAgent", "venue": null, "quality_score": 0.2749}
{"id": "10f3ab2766cc3c29c651760dd4fc0cff22a93dca5a444b68977040437c67258f", "sources": ["arxiv", "semantic_scholar"], "title": "Game Theory Meets LLM and Agentic AI: Reimagining Cybersecurity for the Age of Intelligent Threats", "abstract": "Protecting cyberspace requires not only advanced tools but also a shift in how we reason about threats, trust, and autonomy. Traditional cybersecurity methods rely on manual responses and brittle heuristics. To build proactive and intelligent defense systems, we need integrated theoretical frameworks and software tools. Game theory provides a rigorous foundation for modeling adversarial behavior, designing strategic defenses, and enabling trust in autonomous systems. Meanwhile, software tools process cyber data, visualize attack surfaces, verify compliance, and suggest mitigations. Yet a disconnect remains between theory and practical implementation. The rise of Large Language Models (LLMs) and agentic AI offers a new path to bridge this gap. LLM-powered agents can operationalize abstract strategies into real-world decisions. Conversely, game theory can inform the reasoning and coordination of these agents across complex workflows. LLMs also challenge classical game-theoretic assumptions, such as perfect rationality or static payoffs, prompting new models aligned with cognitive and computational realities. This co-evolution promises richer theoretical foundations and novel solution concepts. Agentic AI also reshapes software design: systems must now be modular, adaptive, and trust-aware from the outset. This chapter explores the intersection of game theory, agentic AI, and cybersecurity. We review key game-theoretic frameworks (e.g., static, dynamic, Bayesian, and signaling games) and solution concepts. We then examine how LLM agents can enhance cyber defense and introduce LLM-driven games that embed reasoning into AI agents. Finally, we explore multi-agent workflows and coordination games, outlining how this convergence fosters secure, intelligent, and adaptive cyber systems.", "authors": ["Quanyan Zhu"], "categories": ["cs.CR", "cs.AI", "cs.CY", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-14", "url": "https://arxiv.org/abs/2507.10621", "pdf_url": "https://arxiv.org/pdf/2507.10621v1", "arxiv_id": "2507.10621", "doi": "10.48550/arXiv.2507.10621", "citation_count": 11, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2698}
{"id": "f69dfa3ad6e0dbd0bfe680aa2c92cd7d2a6e37392453369210dc5a7c1c9066be", "sources": ["arxiv", "semantic_scholar"], "title": "MountainLion: A Multi-Modal LLM-Based Agent System for Interpretable and Adaptive Financial Trading", "abstract": "Cryptocurrency trading is a challenging task requiring the integration of heterogeneous data from multiple modalities. Traditional deep learning and reinforcement learning approaches typically demand large training datasets and encode diverse inputs into numerical representations, often at the cost of interpretability. Recent progress in large language model (LLM)-based agents has demonstrated the capacity to process multi-modal data and support complex investment decision-making. Building on these advances, we present \\textbf{MountainLion}, a multi-modal, multi-agent system for financial trading that coordinates specialized LLM-based agents to interpret financial data and generate investment strategies. MountainLion processes textual news, candlestick charts, and trading signal charts to produce high-quality financial reports, while also enabling modification of reports and investment recommendations through data-driven user interaction and question answering. A central reflection module analyzes historical trading signals and outcomes to continuously refine decision processes, and the system is capable of real-time report analysis, summarization, and dynamic adjustment of investment strategies. Empirical results confirm that MountainLion systematically enriches technical price triggers with contextual macroeconomic and capital flow signals, providing a more interpretable, robust, and actionable investment framework that improves returns and strengthens investor confidence.", "authors": ["Siyi Wu", "Junqiao Wang", "Zhaoyang Guan", "Leyi Zhao", "Xinyuan Song", "Xinyu Ying", "Dexu Yu", "Jinhao Wang", "Hanlin Zhang", "Michele Pak", "Yangfan He", "Yi Xin", "Jianhui Wang", "Tianyu Shi"], "categories": ["q-fin.TR", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2025-07-13", "url": "https://arxiv.org/abs/2507.20474", "pdf_url": "https://arxiv.org/pdf/2507.20474v3", "arxiv_id": "2507.20474", "doi": "10.48550/arXiv.2507.20474", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2315}
{"id": "fd4014a2f96269f09a7c3a62a4623d26e089a633d6beaf9ee14d8a5e12679a33", "sources": ["arxiv", "semantic_scholar"], "title": "StockSim: A Dual-Mode Order-Level Simulator for Evaluating Multi-Agent LLMs in Financial Markets", "abstract": "We present StockSim, an open-source simulation platform for systematic evaluation of large language models (LLMs) in realistic financial decision-making scenarios. Unlike previous toolkits that offer limited scope, StockSim delivers a comprehensive system that fully models market dynamics and supports diverse simulation modes of varying granularity. It incorporates critical real-world factors, such as latency, slippage, and order-book microstructure, that were previously neglected, enabling more faithful and insightful assessment of LLM-based trading agents. An extensible, role-based agent framework supports heterogeneous trading strategies and multi-agent coordination, making StockSim a uniquely capable testbed for NLP research on reasoning under uncertainty and sequential decision-making. We open-source all our code at https: //github.com/harrypapa2002/StockSim.", "authors": ["Charidimos Papadakis", "Giorgos Filandrianos", "Angeliki Dimitriou", "Maria Lymperaiou", "Konstantinos Thomas", "Giorgos Stamou"], "categories": ["cs.CE", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-12", "url": "https://arxiv.org/abs/2507.09255", "pdf_url": "https://arxiv.org/pdf/2507.09255v1", "arxiv_id": "2507.09255", "doi": "10.48550/arXiv.2507.09255", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3559}
{"id": "ff852c007fe280b7f06a8ac2d1ee88b478d2a3a77ed3b2c5b127033a06bca7e2", "sources": ["arxiv", "semantic_scholar"], "title": "ToolRegistry: A Protocol-Agnostic Tool Management Library for Function-Calling LLMs", "abstract": "Every LLM tool call is structurally an RPC -- a function name, JSON arguments, and a serialized result -- yet each protocol (native Python, MCP, OpenAPI, LangChain) is integrated from scratch. We present ToolRegistry, a system that makes this RPC nature explicit: a single Tool object acts as a universal stub regardless of transport, while the registry serves as the RPC client runtime for dispatch, schema generation, and execution. The system ships as three packages -- a core registry, a server exposing tools over MCP and OpenAPI, and a hub of production-ready implementations -- and invokes tools through pluggable thread or process backends. The system now also provides tag-based permission policies, BM25F-powered progressive tool disclosure for large registries, think-augmented function calling, multi-provider schema support (OpenAI, Anthropic, Gemini), declarative JSONC/YAML configuration, and a near-zero-dependency core built on stdlib-only vendored modules. In our benchmarks the library cuts integration code by 60-80%, and choosing the right concurrency mode (thread vs. process) yields up to 3.1x throughput over the alternative for a given workload. ToolRegistry is open-source at https://github.com/Oaklight/ToolRegistry; documentation lives at https://toolregistry.readthedocs.io/.", "authors": ["Peng Ding", "Rick Stevens"], "categories": ["cs.SE", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.10593", "pdf_url": "https://arxiv.org/pdf/2507.10593v3", "arxiv_id": "2507.10593", "doi": "10.48550/arXiv.2507.10593", "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Oaklight/ToolRegistry;", "venue": "arXiv.org", "quality_score": 0.3542}
{"id": "c22a0a7b0ca4ee440f8e5899b91330341181659ebbc07f7d31d1c9940aefacfe", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Design of Multi-Agent LLM Dialogues for Research Ideation", "abstract": "Large language models (LLMs) are increasingly used to support creative tasks such as research idea generation. While recent work has shown that structured dialogues between LLMs can improve the novelty and feasibility of generated ideas, the optimal design of such interactions remains unclear. In this study, we conduct a comprehensive analysis of multi-agent LLM dialogues for scientific ideation. We compare different configurations of agent roles, number of agents, and dialogue depth to understand how these factors influence the novelty and feasibility of generated ideas. Our experimental setup includes settings where one agent generates ideas and another critiques them, enabling iterative improvement. Our results show that enlarging the agent cohort, deepening the interaction depth, and broadening agent persona heterogeneity each enrich the diversity of generated ideas. Moreover, specifically increasing critic-side diversity within the ideation-critique-revision loop further boosts the feasibility of the final proposals. Our findings offer practical guidelines for building effective multi-agent LLM systems for scientific ideation. Our code is available at https://github.com/g6000/MultiAgent-Research-Ideator.", "authors": ["Keisuke Ueda", "Wataru Hirota", "Takuto Asakura", "Takahiro Omi", "Kosuke Takahashi", "Kosuke Arima", "Tatsuya Ishigaki"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08350", "pdf_url": "https://arxiv.org/pdf/2507.08350v1", "arxiv_id": "2507.08350", "doi": "10.48550/arXiv.2507.08350", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/g6000/MultiAgent-Research-Ideator", "venue": "arXiv.org", "quality_score": 0.3542}
{"id": "7aa0d0b5d5b1bc45ac25f7b4da484321401058c7c830efc3a3e230ceda797b64", "sources": ["arxiv", "semantic_scholar"], "title": "How to Train a Leader: Hierarchical Reasoning in Multi-Agent LLMs", "abstract": "Large Language Models (LLMs) have achieved strong performance on a wide range of complex reasoning tasks, yet further gains are often possible by leveraging the complementary strengths of multiple models. While multi-agent frameworks can improve solution quality by leveraging multiple LLMs, existing methods are often computationally expensive, both at training and inference time. In this work, we introduce a hierarchical multi-agent framework that addresses these challenges by training only a single leader LLM to coordinate a team of untrained peer agents. To this end, we propose Multi-agent guided Leader Policy \\textbf{O}ptimization (MLPO), a novel approach which trains the leader to evaluate and synthesize agent responses without auxiliary value networks or explicit agent feedback. Leaders trained with MLPO exhibit improved performance not only when interacting with the agent team at inference time, but also enjoy improved performance when deployed in single-agent settings without the team. Empirical results on Big-Bench Hard (BBH), MATH, and MMLU demonstrate that our framework achieves substantial performance improvements over both single-agent and multi-agent baselines. Our results highlight the effectiveness and efficiency of training a single, flexible leader for collaborative reasoning in multi-agent LLM systems.", "authors": ["Andrew Estornell", "Jean-Francois Ton", "Muhammad Faaiz Taufiq", "Hang Li"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08960", "pdf_url": "https://arxiv.org/pdf/2507.08960v1", "arxiv_id": "2507.08960", "doi": "10.48550/arXiv.2507.08960", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2292}
{"id": "a63c6d90e025e23f8202c5be322884803539413f111bc718c48fa302c52a9099", "sources": ["arxiv", "semantic_scholar"], "title": "AgentsNet: Coordination and Collaborative Reasoning in Multi-Agent LLMs", "abstract": "Large-language models (LLMs) have demonstrated powerful problem-solving capabilities, in particular when organized in multi-agent systems. However, the advent of such systems also raises several questions on the ability of a complex network of agents to effectively self-organize and collaborate. While measuring performance on standard reasoning benchmarks indicates how well multi-agent systems can solve reasoning tasks, it is unclear whether these systems are able to leverage their topology effectively. Here, we propose AgentsNet, a new benchmark for multi-agent reasoning. By drawing inspiration from classical problems in distributed systems and graph theory, AgentsNet measures the ability of multi-agent systems to collaboratively form strategies for problem-solving, self-organization, and effective communication given a network topology. We evaluate a variety of baseline methods on AgentsNet including homogeneous networks of agents which first have to agree on basic protocols for organization and communication. We find that some frontier LLMs are already demonstrating strong performance for small networks but begin to fall off once the size of the network scales. While existing multi-agent benchmarks cover at most 2-5 agents, AgentsNet is practically unlimited in size and can scale with new generations of LLMs. As such, we also probe frontier models in a setup with up to 100 agents.", "authors": ["Florian Grötschla", "Luis Müller", "Jan Tönshoff", "Mikhail Galkin", "Bryan Perozzi"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08616", "pdf_url": "https://arxiv.org/pdf/2507.08616v1", "arxiv_id": "2507.08616", "doi": "10.48550/arXiv.2507.08616", "citation_count": 21, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3356}
{"id": "8905b1f1c59a28e26a74e3d927b11393a331bd530904224dbc272a6b783cfb5f", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLMs as Ethics Advocates for AI-Based Systems", "abstract": "Incorporating ethics into the requirement elicitation process is essential for creating ethically aligned systems. Although eliciting manual ethics requirements is effective, it requires diverse input from multiple stakeholders, which can be challenging due to time and resource constraints. Moreover, it is often given a low priority in the requirements elicitation process. This study proposes a framework for generating ethics requirements drafts by introducing an ethics advocate agent in a multi-agent LLM setting. This agent critiques and provides input on ethical issues based on the system description. The proposed framework is evaluated through two case studies from different contexts, demonstrating that it captures the majority of ethics requirements identified by researchers during 30-minute interviews and introduces several additional relevant requirements. However, it also highlights reliability issues in generating ethics requirements, emphasizing the need for human feedback in this sensitive domain. We believe this work can facilitate the broader adoption of ethics in the requirements engineering process, ultimately leading to more ethically aligned products.", "authors": ["Asma Yamani", "Malak Baslyman", "Moataz Ahmed"], "categories": ["cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08392", "pdf_url": "https://arxiv.org/pdf/2507.08392v3", "arxiv_id": "2507.08392", "doi": "10.48550/arXiv.2507.08392", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2292}
{"id": "a3ac219544ccd5896691c066e6afd7f262afce10b98ef3254e40f64f5a4a1180", "sources": ["arxiv", "semantic_scholar"], "title": "Optimizing Sequential Multi-Step Tasks with Parallel LLM Agents", "abstract": "Large language model (LLM)-based multi-agent systems have demonstrated remarkable promise for tackling complex tasks by breaking them down into subtasks that are iteratively planned, executed, observed, and refined. Despite their effectiveness, these systems often incur high latency because real-world problems frequently demand multiple iterative cycles of reasoning steps. To address this challenge, we propose M1-Parallel, a framework that concurrently runs multiple multi-agent teams in parallel to uncover distinct solution paths. By leveraging an event-driven communication model with asynchronous messaging, M1-Parallel efficiently capitalizes on the inherent diversity of valid plans to either reduce end-to-end latency or boost task completion rates. Our experiments on complex tasks show that M1-Parallel with early termination achieves up to $2.2\\times$ speedup while preserving accuracy, and that M1-Parallel with aggregation yields higher task completion rates. We further investigate strategies aimed at encouraging diverse execution plans but observe no additional performance gains over repeated sampling. Overall, these findings underscore the potential of parallel plan execution for optimizing multi-agent systems for real-world, high-complexity reasoning tasks.", "authors": ["Enhao Zhang", "Erkang Zhu", "Gagan Bansal", "Adam Fourney", "Hussein Mozannar", "Jack Gerrits"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08944", "pdf_url": "https://arxiv.org/pdf/2507.08944v1", "arxiv_id": "2507.08944", "doi": "10.48550/arXiv.2507.08944", "citation_count": 9, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "663d6603b808e30b22248f0938741c555e9726a309a5ff11458d6ce41af27e04", "sources": ["arxiv", "semantic_scholar"], "title": "CRMAgent: A Multi-Agent LLM System for E-Commerce CRM Message Template Generation", "abstract": "In e-commerce private-domain channels such as instant messaging and e-mail, merchants engage customers directly as part of their Customer Relationship Management (CRM) programmes to drive retention and conversion. While a few top performers excel at crafting outbound messages, most merchants struggle to write persuasive copy because they lack both expertise and scalable tools. We introduce CRMAgent, a multi-agent system built on large language models (LLMs) that generates high-quality message templates and actionable writing guidance through three complementary modes. First, group-based learning enables the agent to learn from a merchant's own top-performing messages within the same audience segment and rewrite low-performing ones. Second, retrieval-and-adaptation fetches templates that share the same audience segment and exhibit high similarity in voucher type and product category, learns their successful patterns, and adapts them to the current campaign. Third, a rule-based fallback provides a lightweight zero-shot rewrite when no suitable references are available. Extensive experiments show that CRMAgent consistently outperforms merchants' original templates, delivering significant gains in both audience-match and marketing-effectiveness metrics.", "authors": ["Yinzhu Quan", "Xinrui Li", "Ying Chen"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-11", "url": "https://arxiv.org/abs/2507.08325", "pdf_url": "https://arxiv.org/pdf/2507.08325v2", "arxiv_id": "2507.08325", "doi": "10.48550/arXiv.2507.08325", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2292}
{"id": "529ba4ffa413286eba36d0ae42c0dae7de402935c5fd576fdd5c069d1744dda7", "sources": ["arxiv", "semantic_scholar"], "title": "MIRIX: Multi-Agent Memory System for LLM-Based Agents", "abstract": "Although memory capabilities of AI agents are gaining increasing attention, existing solutions remain fundamentally limited. Most rely on flat, narrowly scoped memory components, constraining their ability to personalize, abstract, and reliably recall user-specific information over time. To this end, we introduce MIRIX, a modular, multi-agent memory system that redefines the future of AI memory by solving the field's most critical challenge: enabling language models to truly remember. Unlike prior approaches, MIRIX transcends text to embrace rich visual and multimodal experiences, making memory genuinely useful in real-world scenarios. MIRIX consists of six distinct, carefully structured memory types: Core, Episodic, Semantic, Procedural, Resource Memory, and Knowledge Vault, coupled with a multi-agent framework that dynamically controls and coordinates updates and retrieval. This design enables agents to persist, reason over, and accurately retrieve diverse, long-term user data at scale. We validate MIRIX in two demanding settings. First, on ScreenshotVQA, a challenging multimodal benchmark comprising nearly 20,000 high-resolution computer screenshots per sequence, requiring deep contextual understanding and where no existing memory systems can be applied, MIRIX achieves 35% higher accuracy than the RAG baseline while reducing storage requirements by 99.9%. Second, on LOCOMO, a long-form conversation benchmark with single-modal textual input, MIRIX attains state-of-the-art performance of 85.4%, far surpassing existing baselines. These results show that MIRIX sets a new performance standard for memory-augmented LLM agents. To allow users to experience our memory system, we provide a packaged application powered by MIRIX. It monitors the screen in real time, builds a personalized memory base, and offers intuitive visualization and secure local storage to ensure privacy.", "authors": ["Yu Wang", "Xi Chen"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-10", "url": "https://arxiv.org/abs/2507.07957", "pdf_url": "https://arxiv.org/pdf/2507.07957v1", "arxiv_id": "2507.07957", "doi": "10.48550/arXiv.2507.07957", "citation_count": 111, "influential_citation_count": 14, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.588}
{"id": "551312ad67d2b1a7c42aed7a3fbb6092304814a5c9e0b04cb6004f1f67a9f5bc", "sources": ["arxiv", "semantic_scholar"], "title": "PyVision: Agentic Vision with Dynamic Tooling", "abstract": "LLMs are increasingly deployed as agents, systems capable of planning, reasoning, and dynamically calling external tools. However, in visual reasoning, prior approaches largely remain limited by predefined workflows and static toolsets. In this report, we present PyVision, an interactive, multi-turn framework that enables MLLMs to autonomously generate, execute, and refine Python-based tools tailored to the task at hand, unlocking flexible and interpretable problem-solving. We develop a taxonomy of the tools created by PyVision and analyze their usage across a diverse set of benchmarks. Quantitatively, PyVision achieves consistent performance gains, boosting GPT-4.1 by +7.8% on V* and Claude-4.0-Sonnet by +31.1% on VLMsAreBlind-mini. These results point to a broader shift: dynamic tooling allows models not just to use tools, but to invent them, advancing toward more agentic visual reasoning.", "authors": ["Shitian Zhao", "Haoquan Zhang", "Shaoheng Lin", "Ming Li", "Qilong Wu", "Kaipeng Zhang", "Chen Wei"], "categories": ["cs.CL", "cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-10", "url": "https://arxiv.org/abs/2507.07998", "pdf_url": "https://arxiv.org/pdf/2507.07998v3", "arxiv_id": "2507.07998", "doi": "10.48550/arXiv.2507.07998", "citation_count": 48, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4225}
{"id": "b0bceb07899dbc814f60a780655ff17d836496ef61a052706b5bdfb7339de350", "sources": ["arxiv", "semantic_scholar"], "title": "KVFlow: Efficient Prefix Caching for Accelerating LLM-Based Multi-Agent Workflows", "abstract": "Large language model (LLM) based agentic workflows have become a popular paradigm for coordinating multiple specialized agents to solve complex tasks. To improve serving efficiency, existing LLM systems employ prefix caching to reuse key-value (KV) tensors corresponding to agents' fixed prompts, thereby avoiding redundant computation across repeated invocations. However, current systems typically evict KV caches using a Least Recently Used (LRU) policy, which fails to anticipate future agent usage and often discards KV caches shortly before their reuse. This leads to frequent cache misses and substantial recomputation or swapping overhead. We present KVFlow, a workflow-aware KV cache management framework tailored for agentic workloads. KVFlow abstracts the agent execution schedule as an Agent Step Graph and assigns each agent a steps-to-execution value that estimates its temporal proximity to future activation. These values guide a fine-grained eviction policy at the KV node level, allowing KVFlow to preserve entries likely to be reused and efficiently manage shared prefixes in tree-structured caches. Moreover, KVFlow introduces a fully overlapped KV prefetching mechanism, which proactively loads required tensors from CPU to GPU in background threads for agents scheduled in the next step, thereby avoiding cache miss stalls during generation. Compared to SGLang with hierarchical radix cache, KVFlow achieves up to 1.83$\\times$ speedup for single workflows with large prompts, and up to 2.19$\\times$ speedup for scenarios with many concurrent workflows.", "authors": ["Zaifeng Pan", "Ajjkumar Patel", "Zhengding Hu", "Yipeng Shen", "Yue Guan", "Wan-Lu Li", "Lianhui Qin", "Yida Wang", "Yufei Ding"], "categories": ["cs.DC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-10", "url": "https://arxiv.org/abs/2507.07400", "pdf_url": "https://arxiv.org/pdf/2507.07400v1", "arxiv_id": "2507.07400", "doi": "10.48550/arXiv.2507.07400", "citation_count": 34, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.386}
{"id": "09bea32c0a565a9c9ccb30f67d4b6df5617c12f541fc89dc232c0bc681d673f3", "sources": ["arxiv", "semantic_scholar"], "title": "Gradientsys: A Multi-Agent LLM Scheduler with ReAct Orchestration", "abstract": "We present Gradientsys, a next-generation multi-agent scheduling framework that coordinates diverse specialized AI agents using a typed Model-Context Protocol (MCP) and a ReAct-based dynamic planning loop. At its core, Gradientsys employs an LLM-powered scheduler for intelligent one-to-many task dispatch, enabling parallel execution of heterogeneous agents such as PDF parsers, web search modules, GUI controllers, and web builders. The framework supports hybrid synchronous/asynchronous execution, respects agent capacity constraints, and incorporates a robust retry-and-replan mechanism to handle failures gracefully. To promote transparency and trust, Gradientsys includes an observability layer streaming real-time agent activity and intermediate reasoning via Server-Sent Events (SSE). We offer an architectural overview and evaluate Gradientsys against existing frameworks in terms of extensibility, scheduling topology, tool reusability, parallelism, and observability. Experiments on the GAIA general-assistant benchmark show that Gradientsys achieves higher task success rates with reduced latency and lower API costs compared to a MinionS-style baseline, demonstrating the strength of its LLM-driven multi-agent orchestration.", "authors": ["Xinyuan Song", "Zeyu Wang", "Siyi Wu", "Tianyu Shi", "Lynn Ai"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-09", "url": "https://arxiv.org/abs/2507.06520", "pdf_url": "https://arxiv.org/pdf/2507.06520v1", "arxiv_id": "2507.06520", "doi": "10.48550/arXiv.2507.06520", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "07df760cd313afb9049874aef4fed297ae1643b0fcbc4e2e97adb5c7f7635ddc", "sources": ["arxiv", "semantic_scholar"], "title": "AGACCI : Affiliated Grading Agents for Criteria-Centric Interface in Educational Coding Contexts", "abstract": "Recent advances in AI-assisted education have encouraged the integration of vision-language models (VLMs) into academic assessment, particularly for tasks that require both quantitative and qualitative evaluation. However, existing VLM based approaches struggle with complex educational artifacts, such as programming tasks with executable components and measurable outputs, that require structured reasoning and alignment with clearly defined evaluation criteria. We introduce AGACCI, a multi-agent system that distributes specialized evaluation roles across collaborative agents to improve accuracy, interpretability, and consistency in code-oriented assessment. To evaluate the framework, we collected 360 graduate-level code-based assignments from 60 participants, each annotated by domain experts with binary rubric scores and qualitative feedback. Experimental results demonstrate that AGACCI outperforms a single GPT-based baseline in terms of rubric and feedback accuracy, relevance, consistency, and coherence, while preserving the instructional intent and evaluative depth of expert assessments. Although performance varies across task types, AGACCI highlights the potential of multi-agent systems for scalable and context-aware educational evaluation.", "authors": ["Kwangsuk Park", "Jiwoong Yang"], "categories": ["cs.CY", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-07", "url": "https://arxiv.org/abs/2507.05321", "pdf_url": "https://arxiv.org/pdf/2507.05321v1", "arxiv_id": "2507.05321", "doi": "10.48550/arXiv.2507.05321", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2246}
{"id": "b2e41f5dd8d3313b3fbfd9b8e284256b9aa3507b684f855184df71e658d24698", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions", "abstract": "Recent benchmarks for Large Language Model (LLM) agents primarily focus on evaluating reasoning, planning, and execution capabilities, while another critical component-memory, encompassing how agents memorize, update, and retrieve long-term information-is under-evaluated due to the lack of benchmarks. We term agents with memory mechanisms as memory agents. In this paper, based on classic theories from memory science and cognitive science, we identify four core competencies essential for memory agents: accurate retrieval, test-time learning, long-range understanding, and selective forgetting. Existing benchmarks either rely on limited context lengths or are tailored for static, long-context settings like book-based QA, which do not reflect the interactive, multi-turn nature of memory agents that incrementally accumulate information. Moreover, no existing benchmarks cover all four competencies. We introduce MemoryAgentBench, a new benchmark specifically designed for memory agents. Our benchmark transforms existing long-context datasets and incorporates newly constructed datasets into a multi-turn format, effectively simulating the incremental information processing characteristic of memory agents. By carefully selecting and curating datasets, our benchmark provides comprehensive coverage of the four core memory competencies outlined above, thereby offering a systematic and challenging testbed for assessing memory quality. We evaluate a diverse set of memory agents, ranging from simple context-based and retrieval-augmented generation (RAG) systems to advanced agents with external memory modules and tool integration. Empirical results reveal that current methods fall short of mastering all four competencies, underscoring the need for further research into comprehensive memory mechanisms for LLM agents.", "authors": ["Yuanzhe Hu", "Yu Wang", "Julian McAuley"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-07", "url": "https://arxiv.org/abs/2507.05257", "pdf_url": "https://arxiv.org/pdf/2507.05257v3", "arxiv_id": "2507.05257", "doi": "10.48550/arXiv.2507.05257", "citation_count": 113, "influential_citation_count": 13, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5731}
{"id": "c3dcddcf4e2ce7b5e6bedf2db854a811f56755e6761366106a91e3f7f02d3d54", "sources": ["arxiv", "semantic_scholar"], "title": "Who's the Mole? Modeling and Detecting Intention-Hiding Malicious Agents in LLM-Based Multi-Agent Systems", "abstract": "Multi-agent systems powered by Large Language Models (LLM-MAS) have demonstrated remarkable capabilities in collaborative problem-solving. However, their deployment also introduces new security risks. Existing research on LLM-based agents has primarily examined single-agent scenarios, while the security of multi-agent systems remains largely unexplored. To address this gap, we present a systematic study of intention-hiding threats in LLM-MAS. We design four representative attack paradigms that subtly disrupt task completion while maintaining a high degree of stealth, and evaluate them under centralized, decentralized, and layered communication structures. Experimental results show that these attacks are highly disruptive and can easily evade existing defense mechanisms. To counter these threats, we propose AgentXposed, a psychology-inspired detection framework. AgentXposed draws on the HEXACO personality model, which characterizes agents through psychological trait dimensions, and the Reid interrogation technique, a structured method for eliciting concealed intentions. By combining progressive questionnaire probing with behavior-based inter-agent monitoring, the framework enables the proactive identification of malicious agents before harmful actions are carried out. Extensive experiments across six datasets against both our proposed attacks and two baseline threats demonstrate that AgentXposed effectively detects diverse forms of malicious behavior, achieving strong robustness across multiple communication settings.", "authors": ["Yizhe Xie", "Congcong Zhu", "Xinyue Zhang", "Tianqing Zhu", "Dayong Ye", "Minghao Wang", "Chi Liu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-07", "url": "https://arxiv.org/abs/2507.04724", "pdf_url": "https://arxiv.org/pdf/2507.04724v2", "arxiv_id": "2507.04724", "doi": "10.48550/arXiv.2507.04724", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "2983e086296d43b36ebac97c2fc8492bf7e41262fbe4f538381d029a1ecbaf58", "sources": ["arxiv", "semantic_scholar"], "title": "CREW-WILDFIRE: Benchmarking Agentic Multi-Agent Collaborations at Scale", "abstract": "Despite rapid progress in large language model (LLM)-based multi-agent systems, current benchmarks fall short in evaluating their scalability, robustness, and coordination capabilities in complex, dynamic, real-world tasks. Existing environments typically focus on small-scale, fully observable, or low-complexity domains, limiting their utility for developing and assessing next-generation multi-agent Agentic AI frameworks. We introduce CREW-Wildfire, an open-source benchmark designed to close this gap. Built atop the human-AI teaming CREW simulation platform, CREW-Wildfire offers procedurally generated wildfire response scenarios featuring large maps, heterogeneous agents, partial observability, stochastic dynamics, and long-horizon planning objectives. The environment supports both low-level control and high-level natural language interactions through modular Perception and Execution modules. We implement and evaluate several state-of-the-art LLM-based multi-agent Agentic AI frameworks, uncovering significant performance gaps that highlight the unsolved challenges in large-scale coordination, communication, spatial reasoning, and long-horizon planning under uncertainty. By providing more realistic complexity, scalable architecture, and behavioral evaluation metrics, CREW-Wildfire establishes a critical foundation for advancing research in scalable multi-agent Agentic intelligence. All code, environments, data, and baselines will be released to support future research in this emerging domain.", "authors": ["Jonathan Hyun", "Nicholas R Waytowich", "Boyuan Chen"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-07", "url": "https://arxiv.org/abs/2507.05178", "pdf_url": "https://arxiv.org/pdf/2507.05178v2", "arxiv_id": "2507.05178", "doi": "10.48550/arXiv.2507.05178", "citation_count": 9, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.2654}
{"id": "73ecf40554c006ebf58d775392bdbf3cae8684e552ba9da65e8d42ce8d9f1067", "sources": ["arxiv", "semantic_scholar"], "title": "Conversational Education at Scale: A Multi-LLM Agent Workflow for Procedural Learning and Pedagogic Quality Assessment", "abstract": "Large language models (LLMs) have advanced virtual educators and learners, bridging NLP with AI4Education. Existing work often lacks scalability and fails to leverage diverse, large-scale course content, with limited frameworks for assessing pedagogic quality. To this end, we propose WikiHowAgent, a multi-agent workflow leveraging LLMs to simulate interactive teaching-learning conversations. It integrates teacher and learner agents, an interaction manager, and an evaluator to facilitate procedural learning and assess pedagogic quality. We introduce a dataset of 114,296 teacher-learner conversations grounded in 14,287 tutorials across 17 domains and 727 topics. Our evaluation protocol combines computational and rubric-based metrics with human judgment alignment. Results demonstrate the workflow's effectiveness in diverse setups, offering insights into LLM capabilities across domains. Our datasets and implementations are fully open-sourced.", "authors": ["Jiahuan Pei", "Fanghua Ye", "Xin Sun", "Wentao Deng", "Koen Hindriks", "Junxiao Wang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-07", "url": "https://arxiv.org/abs/2507.05528", "pdf_url": "https://arxiv.org/pdf/2507.05528v2", "arxiv_id": "2507.05528", "doi": "10.48550/arXiv.2507.05528", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3471}
{"id": "672317ecd2598c793b12fe6301e752b084a46d107a385029b43617de2c5e17da", "sources": ["arxiv", "semantic_scholar"], "title": "A LLM-Driven Multi-Agent Systems for Professional Development of Mathematics Teachers", "abstract": "Professional development (PD) serves as the cornerstone for teacher tutors to grasp content knowledge. However, providing equitable and timely PD opportunities for teachers poses significant challenges. To address this issue, we introduce I-VIP (Intelligent Virtual Interactive Program), an intelligent tutoring platform for teacher professional development, driven by large language models (LLMs) and supported by multi-agent frameworks. This platform offers a user-friendly conversational interface and allows users to employ a variety of interactive tools to facilitate question answering, knowledge comprehension, and reflective summarization while engaging in dialogue. To underpin the functionality of this platform, including knowledge expectation analysis, response scoring and classification, and feedback generation, the multi-agent frameworks are leveraged to enhance the accuracy of judgments and mitigate the issue of missing key points.", "authors": ["Kaiqi Yang", "Hang Li", "Yucheng Chu", "Ahreum Han", "Yasemin Copur-Gencturk", "Jiliang Tang", "Hui Liu"], "categories": ["cs.CY", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-05", "url": "https://arxiv.org/abs/2507.05292", "pdf_url": "https://arxiv.org/pdf/2507.05292v1", "arxiv_id": "2507.05292", "doi": "10.48550/arXiv.2507.05292", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2223}
{"id": "0b37a2819bac2ab589f4008dc3bc1eab15f4bb69f6814aa29de8f0cdb6b36590", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Robustness of LLM-Driven Multi-Agent Systems through Randomized Smoothing", "abstract": "This paper presents a defense framework for enhancing the safety of large language model (LLM) empowered multi-agent systems (MAS) in safety-critical domains such as aerospace. We apply randomized smoothing, a statistical robustness certification technique, to the MAS consensus context, enabling probabilistic guarantees on agent decisions under adversarial influence. Unlike traditional verification methods, our approach operates in black-box settings and employs a two-stage adaptive sampling mechanism to balance robustness and computational efficiency. Simulation results demonstrate that our method effectively prevents the propagation of adversarial behaviors and hallucinations while maintaining consensus performance. This work provides a practical and scalable path toward safe deployment of LLM-based MAS in real-world, high-stakes environments.", "authors": ["Jinwei Hu", "Yi Dong", "Zhengtao Ding", "Xiaowei Huang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-05", "url": "https://arxiv.org/abs/2507.04105", "pdf_url": "https://arxiv.org/pdf/2507.04105v1", "arxiv_id": "2507.04105", "doi": "10.48550/arXiv.2507.04105", "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Chinese Journal of Aeronautics", "quality_score": 0.2698}
{"id": "7955d6a047a85d61830aa7f946f6013aea02dbc76b9082dada3da52f51ee7cbb", "sources": ["arxiv", "semantic_scholar"], "title": "CortexDebate: Debating Sparsely and Equally for Multi-Agent Debate", "abstract": "Nowadays, single Large Language Model (LLM) struggles with critical issues such as hallucination and inadequate reasoning abilities. To mitigate these issues, Multi-Agent Debate (MAD) has emerged as an effective strategy, where LLM agents engage in in-depth debates with others on tasks. However, existing MAD methods face two major issues: (a) too lengthy input contexts, which causes LLM agents to get lost in plenty of input information and experiences performance drop; and (b) the overconfidence dilemma, where self-assured LLM agents dominate the debate, leading to low debating effectiveness. To address these limitations, we propose a novel MAD method called \"CortexDebate\". Inspired by the human brain's tendency to establish a sparse and dynamically optimized network among cortical areas governed by white matter, CortexDebate constructs a sparse debating graph among LLM agents, where each LLM agent only debates with the ones that are helpful to it. To optimize the graph, we propose a module named McKinsey-based Debate Matter (MDM), which acts as an artificial analog to white matter. By integrating the McKinsey Trust Formula, a well-established measure of trustworthiness from sociology, MDM enables credible evaluations that guide graph optimization. The effectiveness of our CortexDebate has been well demonstrated by extensive experimental results across eight datasets from four task types.", "authors": ["Yiliu Sun", "Zicheng Zhao", "Sheng Wan", "Chen Gong"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-05", "url": "https://arxiv.org/abs/2507.03928", "pdf_url": "https://arxiv.org/pdf/2507.03928v1", "arxiv_id": "2507.03928", "doi": "10.48550/arXiv.2507.03928", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.2223}
{"id": "bd91cb4469725968e5004d6714ee40fdb61d6c8e96667c8c2129bf2593130f1a", "sources": ["arxiv", "semantic_scholar"], "title": "CodeAgents: A Token-Efficient Framework for Codified Multi-Agent Reasoning in LLMs", "abstract": "Effective prompt design is essential for improving the planning capabilities of large language model (LLM)-driven agents. However, existing structured prompting strategies are typically limited to single-agent, plan-only settings, and often evaluate performance solely based on task accuracy - overlooking critical factors such as token efficiency, modularity, and scalability in multi-agent environments. To address these limitations, we introduce CodeAgents, a prompting framework that codifies multi-agent reasoning and enables structured, token-efficient planning in multi-agent systems. In CodeAgents, all components of agent interaction - Task, Plan, Feedback, system roles, and external tool invocations - are codified into modular pseudocode enriched with control structures (e.g., loops, conditionals), boolean logic, and typed variables. This design transforms loosely connected agent plans into cohesive, interpretable, and verifiable multi-agent reasoning programs. We evaluate the proposed framework across three diverse benchmarks - GAIA, HotpotQA, and VirtualHome - using a range of representative LLMs. Results show consistent improvements in planning performance, with absolute gains of 3-36 percentage points over natural language prompting baselines. On VirtualHome, our method achieves a new state-of-the-art success rate of 56%. In addition, our approach reduces input and output token usage by 55-87% and 41-70%, respectively, underscoring the importance of token-aware evaluation metrics in the development of scalable multi-agent LLM systems. The code and resources are available at: https://anonymous.4open.science/r/CodifyingAgent-5A86", "authors": ["Bruce Yang", "Xinfeng He", "Huan Gao", "Yifan Cao", "Xiaofan Li", "David Hsu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-04", "url": "https://arxiv.org/abs/2507.03254", "pdf_url": "https://arxiv.org/pdf/2507.03254v1", "arxiv_id": "2507.03254", "doi": "10.48550/arXiv.2507.03254", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2211}
{"id": "3b5a4b8e703d68d42bf38e9525eb843949f22d0614e61a7fe4124c37d05a33f2", "sources": ["arxiv", "semantic_scholar"], "title": "OMS: On-the-fly, Multi-Objective, Self-Reflective Ad Keyword Generation via LLM Agent", "abstract": "Keyword decision in Sponsored Search Advertising is critical to the success of ad campaigns. While LLM-based methods offer automated keyword generation, they face three major limitations: reliance on large-scale query-keyword pair data, lack of online multi-objective performance monitoring and optimization, and weak quality control in keyword selection. These issues hinder the agentic use of LLMs in fully automating keyword decisions by monitoring and reasoning over key performance indicators such as impressions, clicks, conversions, and CTA effectiveness. To overcome these challenges, we propose OMS, a keyword generation framework that is On-the-fly (requires no training data, monitors online performance, and adapts accordingly), Multi-objective (employs agentic reasoning to optimize keywords based on multiple performance metrics), and Self-reflective (agentically evaluates keyword quality). Experiments on benchmarks and real-world ad campaigns show that OMS outperforms existing methods; ablation and human evaluations confirm the effectiveness of each component and the quality of generated keywords.", "authors": ["Bowen Chen", "Zhao Wang", "Shingo Takamatsu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-03", "url": "https://arxiv.org/abs/2507.02353", "pdf_url": "https://arxiv.org/pdf/2507.02353v1", "arxiv_id": "2507.02353", "doi": "10.48550/arXiv.2507.02353", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.22}
{"id": "026a3517b937ff29f2862a894500f8bb2e4929a781c7182ac5ff6149270ee855", "sources": ["arxiv", "semantic_scholar"], "title": "KERAP: A Knowledge-Enhanced Reasoning Approach for Accurate Zero-shot Diagnosis Prediction Using Multi-agent LLMs", "abstract": "Medical diagnosis prediction plays a critical role in disease detection and personalized healthcare. While machine learning (ML) models have been widely adopted for this task, their reliance on supervised training limits their ability to generalize to unseen cases, particularly given the high cost of acquiring large, labeled datasets. Large language models (LLMs) have shown promise in leveraging language abilities and biomedical knowledge for diagnosis prediction. However, they often suffer from hallucinations, lack structured medical reasoning, and produce useless outputs. To address these challenges, we propose KERAP, a knowledge graph (KG)-enhanced reasoning approach that improves LLM-based diagnosis prediction through a multi-agent architecture. Our framework consists of a linkage agent for attribute mapping, a retrieval agent for structured knowledge extraction, and a prediction agent that iteratively refines diagnosis predictions. Experimental results demonstrate that KERAP enhances diagnostic reliability efficiently, offering a scalable and interpretable solution for zero-shot medical diagnosis prediction.", "authors": ["Yuzhang Xie", "Hejie Cui", "Ziyang Zhang", "Jiaying Lu", "Kai Shu", "Fadi Nahab", "Xiao Hu", "Carl Yang"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science", "Medicine"], "published_date": "2025-07-03", "url": "https://arxiv.org/abs/2507.02773", "pdf_url": "https://arxiv.org/pdf/2507.02773v2", "arxiv_id": "2507.02773", "doi": "10.48550/arXiv.2507.02773", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "American Medical Informatics Association (AMIA) 2025 Annual Symposium, Oral", "quality_score": 0.301}
{"id": "f017bb9d01fbd1b5a411cf039089396d3daa54eb4101627203ae561cb185624e", "sources": ["arxiv", "semantic_scholar"], "title": "Using multi-agent architecture to mitigate the risk of LLM hallucinations", "abstract": "Improving customer service quality and response time are critical factors for maintaining customer loyalty and increasing a company's market share. While adopting emerging technologies such as Large Language Models (LLMs) is becoming a necessity to achieve these goals, the risk of hallucination remains a major challenge. In this paper, we present a multi-agent system to handle customer requests sent via SMS. This system integrates LLM based agents with fuzzy logic to mitigate hallucination risks.", "authors": ["Abd Elrahman Amer", "Magdi Amer"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-02", "url": "https://arxiv.org/abs/2507.01446", "pdf_url": "https://arxiv.org/pdf/2507.01446v1", "arxiv_id": "2507.01446", "doi": "10.48550/arXiv.2507.01446", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2189}
{"id": "db1daba68381b0847742731416d05be7c92b7c90a76fb792bdb1ef8e9d93eb97", "sources": ["arxiv", "semantic_scholar"], "title": "The Future is Agentic: Definitions, Perspectives, and Open Challenges of Multi-Agent Recommender Systems", "abstract": "Large language models (LLMs) are rapidly evolving from passive engines of text generation into agentic entities that can plan, remember, invoke external tools, and co-operate with one another. This perspective paper investigates how such LLM agents (and societies thereof) can transform the design space of recommender systems. We introduce a unified formalism that (i) models an individual agent as a tuple comprising its language core, tool set, and hierarchical memory, and (ii) captures a multi-agent recommender as a triple of agents, shared environment, and communication protocol. Within this framework, we present four end-to-end use cases-interactive party planning, synthetic user-simulation for offline evaluation, multi-modal furniture recommendation, and brand-aligned explanation generation-each illustrating a distinct capability unlocked by agentic orchestration. We then surface five cross-cutting challenge families: protocol complexity, scalability, hallucination and error propagation, emergent misalignment (including covert collusion), and brand compliance. For each, we formalize the problem, review nascent mitigation strategies, and outline open research questions. The result is both a blueprint and an agenda: a blueprint that shows how memory-augmented, tool-using LLM agents can be composed into robust recommendation pipelines, and an agenda inviting the RecSys community to develop benchmarks, theoretical guarantees, and governance tools that keep pace with this new degree of autonomy. By unifying agentic abstractions with recommender objectives, the paper lays the groundwork for the next generation of personalized, trustworthy, and context-rich recommendation services.", "authors": ["Reza Yousefi Maragheh", "Yashar Deldjoo"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-02", "url": "https://arxiv.org/abs/2507.02097", "pdf_url": "https://arxiv.org/pdf/2507.02097v2", "arxiv_id": "2507.02097", "doi": "10.48550/arXiv.2507.02097", "citation_count": 15, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "21fe7c18f3f3daae3f99daa83618069390f794cb899dbcdc7be95d42fed72990", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Advanced LLM Multi-Agent Systems Based on Blackboard Architecture", "abstract": "In this paper, we propose to incorporate the blackboard architecture into LLM multi-agent systems (MASs) so that (1) agents with various roles can share all the information and others' messages during the whole problem-solving process, (2) agents that will take actions are selected based on the current content of the blackboard, and (3) the selection and execution round is repeated until a consensus is reached on the blackboard. We develop the first implementation of this proposal and conduct experiments on commonsense knowledge, reasoning and mathematical datasets. The results show that our system can be competitive with the SOTA static and dynamic MASs by achieving the best average performance, and at the same time manage to spend less tokens. Our proposal has the potential to enable complex and dynamic problem-solving where well-defined structures or workflows are unavailable.", "authors": ["Bochen Han", "Songmao Zhang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-07-02", "url": "https://arxiv.org/abs/2507.01701", "pdf_url": "https://arxiv.org/pdf/2507.01701v1", "arxiv_id": "2507.01701", "doi": "10.48550/arXiv.2507.01701", "citation_count": 12, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3495}
{"id": "62304d440433f562f2b6ea53c8fbb6738357f18e6e6819b3002022225ee94ac1", "sources": ["arxiv", "semantic_scholar"], "title": "PokéAI: A Goal-Generating, Battle-Optimizing Multi-agent System for Pokemon Red", "abstract": "We introduce PokéAI, the first text-based, multi-agent large language model (LLM) framework designed to autonomously play and progress through Pokémon Red. Our system consists of three specialized agents-Planning, Execution, and Critique-each with its own memory bank, role, and skill set. The Planning Agent functions as the central brain, generating tasks to progress through the game. These tasks are then delegated to the Execution Agent, which carries them out within the game environment. Upon task completion, the Critique Agent evaluates the outcome to determine whether the objective was successfully achieved. Once verification is complete, control returns to the Planning Agent, forming a closed-loop decision-making system. As a preliminary step, we developed a battle module within the Execution Agent. Our results show that the battle AI achieves an average win rate of 80.8% across 50 wild encounters, only 6% lower than the performance of an experienced human player. Furthermore, we find that a model's battle performance correlates strongly with its LLM Arena score on language-related tasks, indicating a meaningful link between linguistic ability and strategic reasoning. Finally, our analysis of gameplay logs reveals that each LLM exhibits a unique playstyle, suggesting that individual models develop distinct strategic behaviors.", "authors": ["Zihao Liu", "Xinhang Sui", "Yueran Song", "Siwen Wang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-30", "url": "https://arxiv.org/abs/2506.23689", "pdf_url": "https://arxiv.org/pdf/2506.23689v1", "arxiv_id": "2506.23689", "doi": "10.48550/arXiv.2506.23689", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2166}
{"id": "c376c288a76f4b61d56d61ad0b1363b0f31e40854f39639b33bdf4f634838e3b", "sources": ["arxiv", "semantic_scholar"], "title": "AURA: Agent for Understanding, Reasoning, and Automated Tool Use in Voice-Driven Tasks", "abstract": "Despite advances in language and speech technologies, no open-source system enables full speech-to-speech, multi-turn dialogue with integrated tool use and agentic reasoning. We introduce AURA (Agent for Understanding, Reasoning, and Automated Tool Use), the first open-source, speech-native assistant capable of completing complex, goal-driven tasks through dynamic tool invocation and multi-turn conversation. AURA combines open-weight ASR, TTS, and LLMs in a cascaded pipeline and supports tools such as calendar booking, contact lookup, web search, and email. Its modular design allows easy integration of new tools using natural language prompts and action classes. On VoiceBench, AURA scores 92.75% on OpenBookQA-outperforming all open-weight systems and nearing GPT-4o-and 4.39 on AlpacaEval, competitive with other open-weight systems. Human evaluation shows 90% task success on complex, multi-turn speech tasks.", "authors": ["Leander Melroy Maben", "Gayathri Ganesh Lakshmy", "Srijith Radhakrishnan", "Siddhant Arora", "Shinji Watanabe"], "categories": ["cs.AI", "cs.CL", "cs.SD", "eess.AS"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-06-29", "url": "https://arxiv.org/abs/2506.23049", "pdf_url": "https://arxiv.org/pdf/2506.23049v1", "arxiv_id": "2506.23049", "doi": "10.1109/ASRU65441.2025.11434725", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Automatic Speech Recognition & Understanding", "quality_score": 0.3329}
{"id": "050e27a6905c2b2ab4480b735acd51ed4fed010288b2928011eda1ddbddfecfb", "sources": ["arxiv", "semantic_scholar"], "title": "More Vulnerable than You Think: On the Stability of Tool-Integrated LLM Agents", "abstract": "Current evaluations of tool-integrated LLM agents typically focus on end-to-end tool-usage evaluation while neglecting their stability. This limits their real-world applicability, as various internal or external factors can cause agents to crash or behave abnormally. Our research addresses this by investigating whether agents are vulnerable to errors throughout the entire tool invocation process, including reading tool documentation, selecting tools and generating parameters, and processing the tool's response. Through extensive experiments, we observe that agents are highly susceptible to errors at each stage and agents based on open-source models are more vulnerable than those based on proprietary models. We also find that increasing the model size does not significantly improve tool invocation reasoning and may make agents more vulnerable to attacks resembling normal user instructions. This highlights the importance of evaluating agent stability and offers valuable insights for future LLM development and evaluation.", "authors": ["Weimin Xiong", "Ke Wang", "Yifan Song", "Hanchao Liu", "Sai Zhou", "Wei Peng", "Sujian Li"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-27", "url": "https://arxiv.org/abs/2506.21967", "pdf_url": "https://arxiv.org/pdf/2506.21967v1", "arxiv_id": "2506.21967", "doi": "10.48550/arXiv.2506.21967", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3294}
{"id": "c3de42aaee8982ef48ec8e376e4e2549950a9713ce6f94039fb5628bb5f42de5", "sources": ["arxiv", "semantic_scholar"], "title": "Knowledge-Guided Multi-Agent Framework for Automated Requirements Development: A Vision", "abstract": "This paper envisions a knowledge-guided multi-agent framework named KGMAF for automated requirements development. KGMAF aims to address gaps in current automation systems for SE, which prioritize code development and overlook the complexities of requirements tasks. KGMAF is composed of six specialized agents and an artifact pool to improve efficiency and accuracy. Specifically, KGMAF outlines the functionality, actions, and knowledge of each agent and provides the conceptual design of the artifact pool. Our case study highlights the potential of KGMAF in real-world scenarios. Finally, we outline several research opportunities for implementing and enhancing automated requirements development using multi-agent systems. We believe that KGMAF will play a pivotal role in shaping the future of automated requirements development in the era of LLMs.", "authors": ["Jiangping Huang", "Dongming Jin", "Weisong Sun", "Yang Liu", "Zhi Jin"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-27", "url": "https://arxiv.org/abs/2506.22656", "pdf_url": "https://arxiv.org/pdf/2506.22656v1", "arxiv_id": "2506.22656", "doi": "10.48550/arXiv.2506.22656", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2131}
{"id": "af9091c9561384595bb1b2be1c66fb180e8d79cd541d9913025118b50be3a8f2", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-guided Chemical Process Optimization with a Multi-Agent Approach", "abstract": "Chemical process optimization maximizes production efficiency and economic performance, but optimization algorithms, including gradient-based solvers, numerical methods, and parameter grid searches, become impractical when operating constraints are ill-defined or unavailable. We present a multi-agent LLM framework that autonomously infers operating constraints from minimal process descriptions, then collaboratively guides optimization. Our AutoGen-based framework employs OpenAI's o3 model with specialized agents for constraint generation, parameter validation, simulation, and optimization guidance. Through autonomous constraint generation and iterative multi-agent optimization, the framework eliminates the need for predefined operational bounds. Validated on hydrodealkylation across cost, yield, and yield-to-cost ratio metrics, the framework achieved competitive performance with conventional methods while reducing wall-time 31-fold relative to grid search, converging in under 20 minutes. The reasoning-guided search demonstrates sophisticated process understanding, correctly identifying utility trade-offs and applying domain-informed heuristics. Unlike conventional methods requiring predefined constraints, our approach uniquely combines autonomous constraint generation with interpretable parameter exploration. Model comparison reveals reasoning-capable architectures (o3, o1) are essential for successful optimization, while standard models fail to converge. This approach is particularly valuable for emerging processes and retrofit applications where operational constraints are poorly characterized or unavailable.", "authors": ["Tong Zeng", "Srivathsan Badrinarayanan", "Janghoon Ock", "Cheng-Kai Lai", "Amir Barati Farimani"], "categories": ["cs.LG", "cs.AI", "cs.CE"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-06-26", "url": "https://arxiv.org/abs/2506.20921", "pdf_url": "https://arxiv.org/pdf/2506.20921v2", "arxiv_id": "2506.20921", "doi": "10.1088/2632-2153/ae2382", "citation_count": 15, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.301}
{"id": "95f4b4d09765ef79c88100d7c36df37580c657612f21dfc1b8ba07a61592aa3a", "sources": ["arxiv", "semantic_scholar"], "title": "Fine-Tuning and Prompt Engineering of LLMs, for the Creation of Multi-Agent AI for Addressing Sustainable Protein Production Challenges", "abstract": "The global demand for sustainable protein sources has accelerated the need for intelligent tools that can rapidly process and synthesise domain-specific scientific knowledge. In this study, we present a proof-of-concept multi-agent Artificial Intelligence (AI) framework designed to support sustainable protein production research, with an initial focus on microbial protein sources. Our Retrieval-Augmented Generation (RAG)-oriented system consists of two GPT-based LLM agents: (1) a literature search agent that retrieves relevant scientific literature on microbial protein production for a specified microbial strain, and (2) an information extraction agent that processes the retrieved content to extract relevant biological and chemical information. Two parallel methodologies, fine-tuning and prompt engineering, were explored for agent optimisation. Both methods demonstrated effectiveness at improving the performance of the information extraction agent in terms of transformer-based cosine similarity scores between obtained and ideal outputs. Mean cosine similarity scores were increased by up to 25%, while universally reaching mean scores of $\\geq 0.89$ against ideal output text. Fine-tuning overall improved the mean scores to a greater extent (consistently of $\\geq 0.94$) compared to prompt engineering, although lower statistical uncertainties were observed with the latter approach. A user interface was developed and published for enabling the use of the multi-agent AI system, alongside preliminary exploration of additional chemical safety-based search capabilities", "authors": ["Alexander D. Kalian", "Jaewook Lee", "Stefan P. Johannesson", "Lennart Otte", "Christer Hogstrand", "Miao Guo"], "categories": ["cs.AI", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-06-25", "url": "https://arxiv.org/abs/2506.20598", "pdf_url": "https://arxiv.org/pdf/2506.20598v1", "arxiv_id": "2506.20598", "doi": "10.48550/arXiv.2506.20598", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2108}
{"id": "e4b5e0d8e0b4f6866f2aa728f2759a85dd31d2d42a9536783df6b4b09956764f", "sources": ["arxiv", "semantic_scholar"], "title": "The Decrypto Benchmark for Multi-Agent Reasoning and Theory of Mind", "abstract": "As Large Language Models (LLMs) gain agentic abilities, they will have to navigate complex multi-agent scenarios, interacting with human users and other agents in cooperative and competitive settings. This will require new reasoning skills, chief amongst them being theory of mind (ToM), or the ability to reason about the \"mental\" states of other agents. However, ToM and other multi-agent abilities in LLMs are poorly understood, since existing benchmarks suffer from narrow scope, data leakage, saturation, and lack of interactivity. We thus propose Decrypto, a game-based benchmark for multi-agent reasoning and ToM drawing inspiration from cognitive science, computational pragmatics and multi-agent reinforcement learning. It is designed to be as easy as possible in all other dimensions, eliminating confounding factors commonly found in other benchmarks. To our knowledge, it is also the first platform for designing interactive ToM experiments. We validate the benchmark design through comprehensive empirical evaluations of frontier LLMs, robustness studies, and human-AI cross-play experiments. We find that LLM game-playing abilities lag behind humans and simple word-embedding baselines. We then create variants of two classic cognitive science experiments within Decrypto to evaluate three key ToM abilities. Surprisingly, we find that state-of-the-art reasoning models are significantly worse at those tasks than their older counterparts. This demonstrates that Decrypto addresses a crucial gap in current reasoning and ToM evaluations, and paves the path towards better artificial agents.", "authors": ["Andrei Lupu", "Timon Willi", "Jakob Foerster"], "categories": ["cs.AI", "cs.CL", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-25", "url": "https://arxiv.org/abs/2506.20664", "pdf_url": "https://arxiv.org/pdf/2506.20664v1", "arxiv_id": "2506.20664", "doi": "10.48550/arXiv.2506.20664", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2108}
{"id": "e3883a801237915f51e833124b768306c5a4d42adb56af79492ca2b025840754", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Bilateral Team Formation in Cooperative Multi-Agent Reinforcement Learning", "abstract": "Team formation and the dynamics of team-based learning have drawn significant interest in the context of Multi-Agent Reinforcement Learning (MARL). However, existing studies primarily focus on unilateral groupings, predefined teams, or fixed-population settings, leaving the effects of algorithmic bilateral grouping choices in dynamic populations underexplored. To address this gap, we introduce a framework for learning two-sided team formation in dynamic multi-agent systems. Through this study, we gain insight into what algorithmic properties in bilateral team formation influence policy performance and generalization. We validate our approach using widely adopted multi-agent scenarios, demonstrating competitive performance and improved generalization in most scenarios.", "authors": ["Koorosh Moslemi", "Chi-Guhn Lee"], "categories": ["cs.MA", "cs.AI", "cs.GT", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.20039", "pdf_url": "https://arxiv.org/pdf/2506.20039v1", "arxiv_id": "2506.20039", "doi": "10.48550/arXiv.2506.20039", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2097}
{"id": "8463b891df857ff8444cdfa966efd7df9c42a2f302586dcb760d7a7b481545f5", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based Multi-Agent System for Intelligent Refactoring of Haskell Code", "abstract": "Refactoring is a constant activity in software development and maintenance. Scale and maintain software systems are based on code refactoring. However, this process is still labor intensive, as it requires programmers to analyze the codebases in detail to avoid introducing new defects. In this research, we put forward a large language model (LLM)-based multi-agent system to automate the refactoring process on Haskell code. The objective of this research is to evaluate the effect of LLM-based agents in performing structured and semantically accurate refactoring on Haskell code. Our proposed multi-agent system based on specialized agents with distinct roles, including code analysis, refactoring execution, verification, and debugging. To test the effectiveness and practical applicability of the multi-agent system, we conducted evaluations using different open-source Haskell codebases. The results of the experiments carried out showed that the proposed LLM-based multi-agent system could average 11.03% decreased complexity in code, an improvement of 22.46% in overall code quality, and increase performance efficiency by an average of 13.27%. Furthermore, memory allocation was optimized by up to 14.57%. These results highlight the ability of LLM-based multi-agent in managing refactoring tasks targeted toward functional programming paradigms. Our findings hint that LLM-based multi-agent systems integration into the refactoring of functional programming languages can enhance maintainability and support automated development workflows.", "authors": ["Shahbaz Siddeeq", "Muhammad Waseem", "Zeeshan Rasheed", "Md Mahade Hasan", "Jussi Rasku", "Mika Saari", "Henri Terho", "Kalle Makela", "Kai-Kristian Kemell", "Pekka Abrahamsson"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.19481", "pdf_url": "https://arxiv.org/pdf/2506.19481v1", "arxiv_id": "2506.19481", "doi": "10.48550/arXiv.2506.19481", "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "International Conference on Product Focused Software Process Improvement", "quality_score": 0.3241}
{"id": "c4984e3d5b9d102792dd0a84fb8f860bbc7df475dbe90a01915c18bf3ebada04", "sources": ["arxiv", "semantic_scholar"], "title": "Adaptive Domain Modeling with Language Models: A Multi-Agent Approach to Task Planning", "abstract": "We introduce TAPAS (Task-based Adaptation and Planning using AgentS), a multi-agent framework that integrates Large Language Models (LLMs) with symbolic planning to solve complex tasks without the need for manually defined environment models. TAPAS employs specialized LLM-based agents that collaboratively generate and adapt domain models, initial states, and goal specifications as needed using structured tool-calling mechanisms. Through this tool-based interaction, downstream agents can request modifications from upstream agents, enabling adaptation to novel attributes and constraints without manual domain redefinition. A ReAct (Reason+Act)-style execution agent, coupled with natural language plan translation, bridges the gap between dynamically generated plans and real-world robot capabilities. TAPAS demonstrates strong performance in benchmark planning domains and in the VirtualHome simulated real-world environment.", "authors": ["Harisankar Babu", "Philipp Schillinger", "Tamim Asfour"], "categories": ["cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.19592", "pdf_url": "https://arxiv.org/pdf/2506.19592v2", "arxiv_id": "2506.19592", "doi": "10.1109/CASE58245.2025.11164110", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "1a2e05ff62ade48eca344bef61a72bf2269cf1ea8573964931480e15accf798a", "sources": ["arxiv", "semantic_scholar"], "title": "Augmenting Multi-Agent Communication with State Delta Trajectory", "abstract": "Multi-agent techniques such as role playing or multi-turn debates have been shown to be effective in improving the performance of large language models (LLMs) in downstream tasks. Despite their differences in workflows, existing multi-agent systems constructed from a single base LLM mostly use natural language for agent communication. While this is appealing for its simplicity and interpretability, it also introduces inevitable information loss as one model must down sample its continuous state vectors to discrete tokens before transferring them to the other model. Such losses are particularly significant when the information to transfer is not simple facts, but reasoning logics or abstractive thoughts. To tackle this problem, we propose a new communication protocol that transfers both natural language tokens and token-wise state transition trajectory from one agent to another. Particularly, compared to the actual state value, we find that the sequence of state changes in LLMs after generating each token can better reflect the information hidden behind the inference process. We propose a State Delta Encoding (SDE) method to represent state transition trajectories. The experimental results show that multi-agent systems with SDE achieve SOTA performance compared to other communication protocols, particularly in tasks that involve complex reasoning.", "authors": ["Yichen Tang", "Weihang Su", "Yujia Zhou", "Yiqun Liu", "Min Zhang", "Shaoping Ma", "Qingyao Ai"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.19209", "pdf_url": "https://arxiv.org/pdf/2506.19209v2", "arxiv_id": "2506.19209", "doi": "10.48550/arXiv.2506.19209", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.2113}
{"id": "b51f0ed01148c36fdca1cfcda0888841aa6264b55a860c59ff37fdb0feb95aab", "sources": ["arxiv", "semantic_scholar"], "title": "MATE: LLM-Powered Multi-Agent Translation Environment for Accessibility Applications", "abstract": "Accessibility remains a critical concern in today's society, as many technologies are not developed to support the full range of user needs. Existing multi-agent systems (MAS) often cannot provide comprehensive assistance for users in need due to the lack of customization stemming from closed-source designs. Consequently, individuals with disabilities frequently encounter significant barriers when attempting to interact with digital environments. We introduce MATE, a multimodal accessibility MAS, which performs the modality conversions based on the user's needs. The system is useful for assisting people with disabilities by ensuring that data will be converted to an understandable format. For instance, if the user cannot see well and receives an image, the system converts this image to its audio description. MATE can be applied to a wide range of domains, industries, and areas, such as healthcare, and can become a useful assistant for various groups of users. The system supports multiple types of models, ranging from LLM API calling to using custom machine learning (ML) classifiers. This flexibility ensures that the system can be adapted to various needs and is compatible with a wide variety of hardware. Since the system is expected to run locally, it ensures the privacy and security of sensitive information. In addition, the framework can be effectively integrated with institutional technologies (e.g., digital healthcare service) for real-time user assistance. Furthermore, we introduce ModCon-Task-Identifier, a model that is capable of extracting the precise modality conversion task from the user input. Numerous experiments show that ModCon-Task-Identifier consistently outperforms other LLMs and statistical models on our custom data. Our code and data are publicly available at https://github.com/AlgazinovAleksandr/Multi-Agent-MATE.", "authors": ["Aleksandr Algazinov", "Matt Laing", "Paul Laban"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.19502", "pdf_url": "https://arxiv.org/pdf/2506.19502v2", "arxiv_id": "2506.19502", "doi": "10.48550/arXiv.2506.19502", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/AlgazinovAleksandr/Multi-Agent-MATE", "venue": "arXiv.org", "quality_score": 0.3241}
{"id": "cd55eac27e0ee190730f0d0489a6ce6c718814a971d6b4079d1a8bfaa49f24ec", "sources": ["arxiv", "semantic_scholar"], "title": "HiMA-Ecom: Enabling Joint Training of Hierarchical Multi-Agent E-commerce Assistants", "abstract": "Hierarchical multi-agent systems based on large language models (LLMs) have become a common paradigm for building AI assistants in vertical domains such as e-commerce, where a master agent coordinates multiple specialized sub-agents. Despite their practical importance, realistic benchmarks for training and evaluating such systems remain scarce, and joint optimization across functionally distinct agents is still challenging. To address this gap, we introduce HiMA-Ecom, the first hierarchical multi-agent benchmark tailored for e-commerce scenarios. HiMA-Ecom contains 22.8K instances, including agent-specific supervised fine-tuning samples with memory and system-level input-output pairs for joint multi-agent reinforcement learning. Building upon it, a joint training method named HiMA-R1 is proposed. It presents Variance-Reduction Group Relative Policy Optimization (VR-GRPO), which employs initial trajectory-based Monte Carlo sampling to mitigate the exponential joint action space and selects informative agent groups for efficient updates based on reward variance. Furthermore, an adaptive memory evolution mechanism that repurposes GRPO rewards as cost-free supervisory signals is designed to eliminate repetitive reasoning and accelerate convergence. Experiments on HiMA-Ecom demonstrate that our method, built upon smaller 3B/7B open-source models, achieves performance comparable to that of larger LLMs, such as DeepSeek-R1, and surpasses DeepSeek-V3 by an average of 6\\%.", "authors": ["Junxing Hu", "Ai Han", "Haolan Zhan", "Pu Wei", "Zhiqian Zhang", "Yuhang Guo", "Jiawei Lu", "Zhen Chen", "Haoran Li", "Zicheng Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-24", "url": "https://arxiv.org/abs/2506.19846", "pdf_url": "https://arxiv.org/pdf/2506.19846v2", "arxiv_id": "2506.19846", "doi": null, "citation_count": 9, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.25}
{"id": "abe71eab248213d0e6c7b052e119a09889e846eea97a778e1f779347cef91c65", "sources": ["arxiv", "semantic_scholar"], "title": "TRIZ Agents: A Multi-Agent LLM Approach for TRIZ-Based Innovation", "abstract": "TRIZ, the Theory of Inventive Problem Solving, is a structured, knowledge-based framework for innovation and abstracting problems to find inventive solutions. However, its application is often limited by the complexity and deep interdisciplinary knowledge required. Advancements in Large Language Models (LLMs) have revealed new possibilities for automating parts of this process. While previous studies have explored single LLMs in TRIZ applications, this paper introduces a multi-agent approach. We propose an LLM-based multi-agent system, called TRIZ agents, each with specialized capabilities and tool access, collaboratively solving inventive problems based on the TRIZ methodology. This multi-agent system leverages agents with various domain expertise to efficiently navigate TRIZ steps. The aim is to model and simulate an inventive process with language agents. We assess the effectiveness of this team of agents in addressing complex innovation challenges based on a selected case study in engineering. We demonstrate the potential of agent collaboration to produce diverse, inventive solutions. This research contributes to the future of AI-driven innovation, showcasing the advantages of decentralized problem-solving in complex ideation tasks.", "authors": ["Kamil Szczepanik", "Jarosław A. Chudziak"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-23", "url": "https://arxiv.org/abs/2506.18783", "pdf_url": "https://arxiv.org/pdf/2506.18783v1", "arxiv_id": "2506.18783", "doi": "10.5220/0013321900003890", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Agents and Artificial Intelligence", "quality_score": 0.2085}
{"id": "3a06f46d232413b8d5e2b1f600db2cd9f656db71e5e2f2a10825b4571a6ee7a8", "sources": ["arxiv", "semantic_scholar"], "title": "Distilling Tool Knowledge into Language Models via Back-Translated Traces", "abstract": "Large language models (LLMs) often struggle with mathematical problems that require exact computation or multi-step algebraic reasoning. Tool-integrated reasoning (TIR) offers a promising solution by leveraging external tools such as code interpreters to ensure correctness, but it introduces inference-time dependencies that hinder scalability and deployment. In this work, we propose a new paradigm for distilling tool knowledge into LLMs purely through natural language. We first construct a Solver Agent that solves math problems by interleaving planning, symbolic tool calls, and reflective reasoning. Then, using a back-translation pipeline powered by multiple LLM-based agents, we convert interleaved TIR traces into natural language reasoning traces. A Translator Agent generates explanations for individual tool calls, while a Rephrase Agent merges them into a fluent and globally coherent narrative. Empirically, we show that fine-tuning a small open-source model on these synthesized traces enables it to internalize both tool knowledge and structured reasoning patterns, yielding gains on competition-level math benchmarks without requiring tool access at inference.", "authors": ["Xingyue Huang", "Xianglong Hu", "Zifeng Ding", "Yuan He", " Rishabh", "Waleed Alzarooni", "Ziyu Ye", "Wendong Fan", "Bailan He", "Haige Bo", "Changran Hu", "Guohao Li"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-23", "url": "https://arxiv.org/abs/2506.19171", "pdf_url": "https://arxiv.org/pdf/2506.19171v1", "arxiv_id": "2506.19171", "doi": "10.48550/arXiv.2506.19171", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3223}
{"id": "1c1d112b5aa172a4c7c88b060133d5289924216cbf5f7e0ee3221ae6e78fb1f2", "sources": ["arxiv", "semantic_scholar"], "title": "Online Multi-Agent Control with Adversarial Disturbances", "abstract": "Online multi-agent control problems, where many agents pursue competing and time-varying objectives, are widespread in domains such as autonomous robotics, economics, and energy systems. In these settings, robustness to adversarial disturbances is critical. In this paper, we study online control in multi-agent linear dynamical systems subject to such disturbances. In contrast to most prior work in multi-agent control, which typically assumes noiseless or stochastically perturbed dynamics, we consider an online setting where disturbances can be adversarial, and where each agent seeks to minimize its own sequence of convex losses. Under two feedback models, we analyze online gradient-based controllers with local policy updates. We prove per-agent regret bounds that are sublinear and near-optimal in the time horizon and that highlight different scalings with the number of agents. When agents' objectives are aligned, we further show that the multi-agent control problem induces a time-varying potential game for which we derive equilibrium tracking guarantees. Together, our results take a first step in bridging online control with online learning in games, establishing robust individual and collective performance guarantees in dynamic continuous-state environments.", "authors": ["Anas Barakat", "John Lazarsfeld", "Georgios Piliouras", "Antonios Varvitsiotis"], "categories": ["cs.LG", "cs.GT", "math.OC"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2025-06-23", "url": "https://arxiv.org/abs/2506.18814", "pdf_url": "https://arxiv.org/pdf/2506.18814v2", "arxiv_id": "2506.18814", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1327}
{"id": "bf18d950178e82097045faca867f53fe016fa6d9c4da80070950869443983be7", "sources": ["arxiv", "semantic_scholar"], "title": "Breaking Single-Tester Limits: Multi-Agent LLMs for Multi-User Feature Testing", "abstract": "The growing dependence on mobile phones and their apps has made multi-user interactive features, like chat calls, live streaming, and video conferencing, indispensable for bridging the gaps in social connectivity caused by physical and situational barriers. However, automating these interactive features for testing is fraught with challenges, owing to their inherent need for timely, dynamic, and collaborative user interactions, which current automated testing methods inadequately address. Inspired by the concept of agents designed to autonomously and collaboratively tackle problems, we propose MAdroid, a novel multi-agent approach powered by the Large Language Models (LLMs) to automate the multi-user interactive task for app feature testing. Specifically, MAdroid employs two functional types of multi-agents: user agents (Operator) and supervisor agents (Coordinator and Observer). Each agent takes a specific role: the Coordinator directs the interactive task; the Operator mimics user interactions on the device; and the Observer monitors and reviews the task automation process. Our evaluation, which included 41 multi-user interactive tasks, demonstrates the effectiveness of our approach, achieving 82.9% of the tasks with 96.8% action similarity, outperforming the ablation studies and state-of-the-art baselines. Additionally, a preliminary investigation underscores MAdroid's practicality by helping identify 11 multi-user interactive bugs during regression app testing, confirming its potential value in real-world software development contexts.", "authors": ["Sidong Feng", "Changhao Du", "Huaxiao Liu", "Qingnan Wang", "Zhengwei Lv", "Mengfei Wang", "Chunyang Chen"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-21", "url": "https://arxiv.org/abs/2506.17539", "pdf_url": "https://arxiv.org/pdf/2506.17539v3", "arxiv_id": "2506.17539", "doi": "10.48550/arXiv.2506.17539", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2063}
{"id": "b93d27a8500d39e5710e47a1930b05f848ea4a5726b7c7e45a75ecd509754306", "sources": ["arxiv", "semantic_scholar"], "title": "AnyMAC: Cascading Flexible Multi-Agent Collaboration via Next-Agent Prediction", "abstract": "Recent progress in large language model (LLM)-based multi-agent collaboration highlights the power of structured communication in enabling collective intelligence. However, existing methods largely rely on static or graph-based inter-agent topologies, lacking the potential adaptability and flexibility in communication. In this work, we propose a new framework that rethinks multi-agent coordination through a sequential structure rather than a graph structure, offering a significantly larger topology space for multi-agent communication. Our method focuses on two key directions: (1) Next-Agent Prediction, which selects the most suitable agent role at each step, and (2) Next-Context Selection (NCS), which enables each agent to selectively access relevant information from any previous step. Together, these components construct task-adaptive communication pipelines that support both role flexibility and global information flow. Extensive evaluations across multiple benchmarks demonstrate that our approach achieves superior performance while substantially reducing communication overhead.", "authors": ["Song Wang", "Zhen Tan", "Zihan Chen", "Shuang Zhou", "Tianlong Chen", "Jundong Li"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-21", "url": "https://arxiv.org/abs/2506.17784", "pdf_url": "https://arxiv.org/pdf/2506.17784v2", "arxiv_id": "2506.17784", "doi": "10.48550/arXiv.2506.17784", "citation_count": 22, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3404}
{"id": "63bb996509ef1bbdf14a49affa9f43c745aa730b2789dd12cca4cf8f91d83fa3", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Zero-Shot Coordination between Teams of Agents: The N-XPlay Framework", "abstract": "Zero-shot coordination (ZSC) -- the ability to collaborate with unfamiliar partners -- is essential to making autonomous agents effective teammates. Existing ZSC methods evaluate coordination capabilities between two agents who have not previously interacted. However, these scenarios do not reflect the complexity of real-world multi-agent systems, where coordination often involves a hierarchy of sub-groups and interactions between teams of agents, known as Multi-Team Systems (MTS). To address this gap, we first introduce N-player Overcooked, an N-agent extension of the popular two-agent ZSC benchmark, enabling evaluation of ZSC in N-agent scenarios. We then propose N-XPlay for ZSC in N-agent, multi-team settings. Comparison against Self-Play across two-, three- and five-player Overcooked scenarios, where agents are split between an ``ego-team'' and a group of unseen collaborators shows that agents trained with N-XPlay are better able to simultaneously balance ``intra-team'' and ``inter-team'' coordination than agents trained with SP.", "authors": ["Ava Abderezaei", "Chi-Hui Lin", "Joseph Miceli", "Naren Sivagnanadasan", "Stéphane Aroca-Ouellette", "Jake Brawer", "Alessandro Roncone"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-21", "url": "https://arxiv.org/abs/2506.17560", "pdf_url": "https://arxiv.org/pdf/2506.17560v1", "arxiv_id": "2506.17560", "doi": "10.48550/arXiv.2506.17560", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2063}
{"id": "1e42dea294f97e1dd22332b525efea073637ef929af93dbb55cd2f326ce614c4", "sources": ["arxiv"], "title": "Kaleidoscopic Teaming in Multi Agent Simulations", "abstract": "Warning: This paper contains content that may be inappropriate or offensive. AI agents have gained significant recent attention due to their autonomous tool usage capabilities and their integration in various real-world applications. This autonomy poses novel challenges for the safety of such systems, both in single- and multi-agent scenarios. We argue that existing red teaming or safety evaluation frameworks fall short in evaluating safety risks in complex behaviors, thought processes and actions taken by agents. Moreover, they fail to consider risks in multi-agent setups where various vulnerabilities can be exposed when agents engage in complex behaviors and interactions with each other. To address this shortcoming, we introduce the term kaleidoscopic teaming which seeks to capture complex and wide range of vulnerabilities that can happen in agents both in single-agent and multi-agent scenarios. We also present a new kaleidoscopic teaming framework that generates a diverse array of scenarios modeling real-world human societies. Our framework evaluates safety of agents in both single-agent and multi-agent setups. In single-agent setup, an agent is given a scenario that it needs to complete using the tools it has access to. In multi-agent setup, multiple agents either compete against or cooperate together to complete a task in the scenario through which we capture existing safety vulnerabilities in agents. We introduce new in-context optimization techniques that can be used in our kaleidoscopic teaming framework to generate better scenarios for safety analysis. Lastly, we present appropriate metrics that can be used along with our framework to measure safety of agents. Utilizing our kaleidoscopic teaming framework, we identify vulnerabilities in various models with respect to their safety in agentic use-cases.", "authors": ["Ninareh Mehrabi", "Tharindu Kumarage", "Kai-Wei Chang", "Aram Galstyan", "Rahul Gupta"], "categories": ["cs.AI"], "fields_of_study": [], "published_date": "2025-06-20", "url": "https://arxiv.org/abs/2506.17514", "pdf_url": "https://arxiv.org/pdf/2506.17514v1", "arxiv_id": "2506.17514", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1305}
{"id": "62ad3e5e56701c576eeb9ba38848f21136b36faa822e9785264cf462d46b0894", "sources": ["arxiv", "semantic_scholar"], "title": "Generalizable Agent Modeling for Agent Collaboration-Competition Adaptation with Multi-Retrieval and Dynamic Generation", "abstract": "Adapting a single agent to a new multi-agent system brings challenges, necessitating adjustments across various tasks, environments, and interactions with unknown teammates and opponents. Addressing this challenge is highly complex, and researchers have proposed two simplified scenarios, Multi-agent reinforcement learning for zero-shot learning and Ad-Hoc Teamwork. Building on these foundations, we propose a more comprehensive setting, Agent Collaborative-Competitive Adaptation (ACCA), which evaluates an agent to generalize across diverse scenarios, tasks, and interactions with both unfamiliar opponents and teammates. In ACCA, agents adjust to task and environmental changes, collaborate with unseen teammates, and compete against unknown opponents. We introduce a new modeling approach, Multi-Retrieval and Dynamic Generation (MRDG), that effectively models both teammates and opponents using their behavioral trajectories. This method incorporates a positional encoder for varying team sizes and a hypernetwork module to boost agents' learning and adaptive capabilities. Additionally, a viewpoint alignment module harmonizes the observational perspectives of retrieved teammates and opponents with the learning agent. Extensive tests in benchmark scenarios like SMAC, Overcooked-AI, and Melting Pot show that MRDG significantly improves robust collaboration and competition with unseen teammates and opponents, surpassing established baselines. Our code is available at: https://github.com/vcis-wangchenxu/MRDG.git", "authors": ["Chenxu Wang", "Yonggang Jin", "Cheng Hu", "Youpeng Zhao", "Zipeng Dai", "Jian Zhao", "Shiyu Huang", "Liuyu Xiang", "Junge Zhang", "Zhaofeng He"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-20", "url": "https://arxiv.org/abs/2506.16718", "pdf_url": "https://arxiv.org/pdf/2506.16718v1", "arxiv_id": "2506.16718", "doi": "10.1016/j.neucom.2025.130912", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/vcis-wangchenxu/MRDG.git", "venue": "Neurocomputing", "quality_score": 0.317}
{"id": "edfb7eb3d34e2364c7f6a510de953b85fbcb1d20ff2718b12c80db04c0541ff6", "sources": ["arxiv", "semantic_scholar"], "title": "PhishDebate: An LLM-Based Multi-Agent Framework for Phishing Website Detection", "abstract": "Phishing websites remain a major cybersecurity threat, exploiting deceptive structures, brand impersonation, and social engineering to evade detection. Recent advances in large language models (LLMs) have improved phishing detection through contextual understanding, yet most existing approaches rely on single-agent classification, which is prone to hallucination and often lacks interpretability and robustness. To address these limitations, we propose PhishDebate, a modular multi-agent LLM-based debate framework for phishing website detection. Four specialized agents independently analyze webpage aspects, including URL structure, HTML composition, semantic content, and brand impersonation, under the coordination of a Moderator and final Judge. Through structured debate and divergent reasoning, the framework achieves more accurate and interpretable decisions. By reducing uncertain predictions and providing transparent reasoning, PhishDebate functions as an analyst-augmentation system that lowers cognitive load and supports early, left-of-exploit detection of phishing threats. Evaluations on commercial LLMs show that PhishDebate achieves 98.2 % recall on a real-world phishing dataset and outperforms single-agent and Chain-of-Thought (CoT) baselines. Its modular design enables agent-level configurability, allowing adaptation to varying resource and application requirements, and offers scalability to high-velocity, large-scale security data environments.", "authors": ["Wenhao Li", "Selvakumar Manickam", "Yung-wey Chong", "Shankar Karuppayah"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-18", "url": "https://arxiv.org/abs/2506.15656", "pdf_url": "https://arxiv.org/pdf/2506.15656v2", "arxiv_id": "2506.15656", "doi": "10.1109/BigData66926.2025.11401440", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "BigData Congress [Services Society]", "quality_score": 0.2785}
{"id": "d90c215ca39693b49901a0970b43b2965a49aac0821f830a78cc8a4bbcf86dbf", "sources": ["arxiv", "semantic_scholar"], "title": "AgentGroupChat-V2: Divide-and-Conquer Is What LLM-Based Multi-Agent System Need", "abstract": "Large language model based multi-agent systems have demonstrated significant potential in social simulation and complex task resolution domains. However, current frameworks face critical challenges in system architecture design, cross-domain generalizability, and performance guarantees, particularly as task complexity and number of agents increases. We introduces AgentGroupChat-V2, a novel framework addressing these challenges through three core innovations: (1) a divide-and-conquer fully parallel architecture that decomposes user queries into hierarchical task forest structures enabling dependency management and distributed concurrent processing. (2) an adaptive collaboration engine that dynamically selects heterogeneous LLM combinations and interaction modes based on task characteristics. (3) agent organization optimization strategies combining divide-and-conquer approaches for efficient problem decomposition. Extensive experiments demonstrate AgentGroupChat-V2's superior performance across diverse domains, achieving 91.50% accuracy on GSM8K (exceeding the best baseline by 5.6 percentage points), 30.4% accuracy on competition-level AIME (nearly doubling other methods), and 79.20% pass@1 on HumanEval. Performance advantages become increasingly pronounced with higher task difficulty, particularly on Level 5 MATH problems where improvements exceed 11 percentage points compared to state-of-the-art baselines. These results confirm that AgentGroupChat-V2 provides a comprehensive solution for building efficient, general-purpose LLM multi-agent systems with significant advantages in complex reasoning scenarios. Code is available at https://github.com/MikeGu721/AgentGroupChat-V2.", "authors": ["Zhouhong Gu", "Xiaoxuan Zhu", "Yin Cai", "Hao Shen", "Xingzhou Chen", "Qingyi Wang", "Jialin Li", "Xiaoran Shi", "Haoran Guo", "Wenxuan Huang", "Hongwei Feng", "Yanghua Xiao", "Zheyu Ye", "Yao Hu", "Shaosheng Cao"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-18", "url": "https://arxiv.org/abs/2506.15451", "pdf_url": "https://arxiv.org/pdf/2506.15451v1", "arxiv_id": "2506.15451", "doi": "10.48550/arXiv.2506.15451", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/MikeGu721/AgentGroupChat-V2", "venue": "arXiv.org", "quality_score": 0.3134}
{"id": "00754ac64b23a0a5d250e6aa9b1fa41924aa73d8a361131d8cb1afe27bd198b6", "sources": ["arxiv", "semantic_scholar"], "title": "The Effect of State Representation on LLM Agent Behavior in Dynamic Routing Games", "abstract": "Large Language Models (LLMs) have shown promise as decision-makers in dynamic settings, but their stateless nature necessitates creating a natural language representation of history. We present a unifying framework for systematically constructing natural language \"state\" representations for prompting LLM agents in repeated multi-agent games. Previous work on games with LLM agents has taken an ad hoc approach to encoding game history, which not only obscures the impact of state representation on agents' behavior, but also limits comparability between studies. Our framework addresses these gaps by characterizing methods of state representation along three axes: action informativeness (i.e., the extent to which the state representation captures actions played); reward informativeness (i.e., the extent to which the state representation describes rewards obtained); and prompting style (or natural language compression, i.e., the extent to which the full text history is summarized). We apply this framework to a dynamic selfish routing game, chosen because it admits a simple equilibrium both in theory and in human subject experiments \\cite{rapoport_choice_2009}. Despite the game's relative simplicity, we find that there are key dependencies of LLM agent behavior on the natural language state representation. In particular, we observe that representations which provide agents with (1) summarized, rather than complete, natural language representations of past history; (2) information about regrets, rather than raw payoffs; and (3) limited information about others' actions lead to behavior that more closely matches game theoretic equilibrium predictions, and with more stable game play by the agents. By contrast, other representations can exhibit either large deviations from equilibrium, higher variation in dynamic game play over time, or both.", "authors": ["Lyle Goodyear", "Rachel Guo", "Ramesh Johari"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-18", "url": "https://arxiv.org/abs/2506.15624", "pdf_url": "https://arxiv.org/pdf/2506.15624v1", "arxiv_id": "2506.15624", "doi": "10.48550/arXiv.2506.15624", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "ccb86a33902906012bf3d4b196d46ea45df59e108736ec5f692cb501e294b7a9", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent, Multi-Scale Systems with the Koopman Operator", "abstract": "The Koopman Operator (KO) takes nonlinear state dynamics and ``lifts'' those dynamics to an infinite-dimensional functional space of observables in which those dynamics are linear. Computational applications typically use a finite-dimensional approximation to the KO. The KO can also be applied to controlled dynamical systems, and the linearity of the KO then facilitates analysis and control calculations. In principle, the potential benefits provided by the KO, and the way that it facilitates the use of game theory via its linearity, would suggest it as a powerful approach for dealing with multi-agent control problems. In practice, though, there has not been much work in this space: most multi-agent KO work has treated those agents as different components of a single system rather than as distinct decision-making entities. This paper develops a KO formulation for multi-agent systems that structures the interactions between decision-making agents and extends this formulation to systems in which the agents have hierarchical control structures and time scale separated dynamics. We solve the multi-agent control problem in both cases as both a centralized optimization and as a general-sum game theory problem. The comparison of the two sets of optimality conditions defining the control solutions illustrates how coupling between agents can create differences between the social optimum and the Nash equilibrium.", "authors": ["Craig Bakker"], "categories": ["math.DS", "math.OC"], "fields_of_study": ["Mathematics"], "published_date": "2025-06-18", "url": "https://arxiv.org/abs/2506.15589", "pdf_url": "https://arxiv.org/pdf/2506.15589v1", "arxiv_id": "2506.15589", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1291}
{"id": "814f023f71aa15a99b24e549e6d72185c2c7dd9c0acceba42fe757568587d852", "sources": ["arxiv", "semantic_scholar"], "title": "Spec2RTL-Agent: Automated Hardware Code Generation from Complex Specifications Using LLM Agent Systems", "abstract": "Despite recent progress in generating hardware RTL code with LLMs, existing solutions still suffer from a substantial gap between practical application scenarios and the requirements of real-world RTL code development. Prior approaches either focus on overly simplified hardware descriptions or depend on extensive human guidance to process complex specifications, limiting their scalability and automation potential. In this paper, we address this gap by proposing an LLM agent system, termed Spec2RTL-Agent, designed to directly process complex specification documentation and generate corresponding RTL code implementations, advancing LLM-based RTL code generation toward more realistic application settings. To achieve this goal, Spec2RTL-Agent introduces a novel multi-agent collaboration framework that integrates three key enablers: (1) a reasoning and understanding module that translates specifications into structured, step-by-step implementation plans; (2) a progressive coding and prompt optimization module that iteratively refines the code across multiple representations to enhance correctness and synthesisability for RTL conversion; and (3) an adaptive reflection module that identifies and traces the source of errors during generation, ensuring a more robust code generation flow. Instead of directly generating RTL from natural language, our system strategically generates synthesizable C++ code, which is then optimized for HLS. This agent-driven refinement ensures greater correctness and compatibility compared to naive direct RTL generation approaches. We evaluate Spec2RTL-Agent on three specification documents, showing it generates accurate RTL code with up to 75% fewer human interventions than existing methods. This highlights its role as the first fully automated multi-agent system for RTL generation from unstructured specs, reducing reliance on human effort in hardware design.", "authors": ["Zhongzhi Yu", "Mingjie Liu", "Michael Zimmer", "Yingyan Celine Lin", "Yong Liu", "Haoxing Ren"], "categories": ["cs.AR"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-16", "url": "https://arxiv.org/abs/2506.13905", "pdf_url": "https://arxiv.org/pdf/2506.13905v2", "arxiv_id": "2506.13905", "doi": "10.1109/ICLAD65226.2025.00013", "citation_count": 29, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3693}
{"id": "27497e1cbfeffcee8b7249de5f8f86e8d7592392ce30d49808972c5ffa489878", "sources": ["arxiv", "semantic_scholar"], "title": "AgentOrchestra: Orchestrating Multi-Agent Intelligence with the Tool-Environment-Agent(TEA) Protocol", "abstract": "Recent advances in LLM-based agent systems have shown promise on complex, long-horizon tasks, but existing agent protocols (e.g., A2A and MCP) do not adequately support lifecycle-aware coordination across agents, tools, and environments. To address this limitation, we introduce the \\textbf{Tool-Environment-Agent} (TEA) protocol, a unified abstraction that models these components as first-class, versioned resources with explicit lifecycles. TEA supports end-to-end context and version management, improving traceability and reproducibility, while also enabling continual self-evolution of agent-associated components\\footnote{Unless otherwise specified, \\emph{agent-associated components} include prompts, memory/tool/agent/environment code, and agent outputs (solutions).}. Building on TEA, we present \\projectname, a hierarchical multi-agent framework in which a central planner coordinates specialized sub-agents and dynamically extends capabilities during execution. Experiments on four challenging benchmarks, spanning expert-level agent tasks and scientific/mathematical reasoning, show that AgentOrchestra consistently outperforms strong baselines; in particular, it achieves 89.04\\% on the GAIA Test set, placing it among the leading methods to the best of our knowledge. These results highlight the value of explicit protocol design and hierarchical orchestration for building more robust and adaptive multi-agent systems.", "authors": ["Wentao Zhang", "Liang Zeng", "Yuzhen Xiao", "Yongcong Li", "Ce Cui", "Yilei Zhao", "Rui Hu", "Yang Liu", "Yahui Zhou", "Bo An"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-14", "url": "https://arxiv.org/abs/2506.12508", "pdf_url": "https://arxiv.org/pdf/2506.12508v6", "arxiv_id": "2506.12508", "doi": null, "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2865}
{"id": "04e528934142984f5c13944fe2df6ebe9c56316e1aaf2f58cd83ce2acd414fc9", "sources": ["arxiv", "semantic_scholar"], "title": "SheetMind: An End-to-End LLM-Powered Multi-Agent Framework for Spreadsheet Automation", "abstract": "We present SheetMind, a modular multi-agent framework powered by large language models (LLMs) for spreadsheet automation via natural language instructions. The system comprises three specialized agents: a Manager Agent that decomposes complex user instructions into subtasks; an Action Agent that translates these into structured commands using a Backus Naur Form (BNF) grammar; and a Reflection Agent that validates alignment between generated actions and the user's original intent. Integrated into Google Sheets via a Workspace extension, SheetMind supports real-time interaction without requiring scripting or formula knowledge. Experiments on benchmark datasets demonstrate an 80 percent success rate on single step tasks and approximately 70 percent on multi step instructions, outperforming ablated and baseline variants. Our results highlight the effectiveness of multi agent decomposition and grammar based execution for bridging natural language and spreadsheet functionalities.", "authors": ["Ruiyan Zhu", "Xi Cheng", "Ke Liu", "Brian Zhu", "Daniel Jin", "Neeraj Parihar", "Zhoutian Xu", "Oliver Gao"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-14", "url": "https://arxiv.org/abs/2506.12339", "pdf_url": "https://arxiv.org/pdf/2506.12339v1", "arxiv_id": "2506.12339", "doi": "10.48550/arXiv.2506.12339", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "77a1c34e18ff4c48a2fc7393a30e2e26c09300c275483d53b53fd5a9e7979654", "sources": ["arxiv", "semantic_scholar"], "title": "IndoorWorld: Integrating Physical Task Solving and Social Simulation in A Heterogeneous Multi-Agent Environment", "abstract": "Virtual environments are essential to AI agent research. Existing environments for LLM agent research typically focus on either physical task solving or social simulation, with the former oversimplifying agent individuality and social dynamics, and the latter lacking physical grounding of social behaviors. We introduce IndoorWorld, a heterogeneous multi-agent environment that tightly integrates physical and social dynamics. By introducing novel challenges for LLM-driven agents in orchestrating social dynamics to influence physical environments and anchoring social interactions within world states, IndoorWorld opens up possibilities of LLM-based building occupant simulation for architectural design. We demonstrate the potential with a series of experiments within an office setting to examine the impact of multi-agent collaboration, resource competition, and spatial layout on agent behavior.", "authors": ["Dekun Wu", "Frederik Brudy", "Bang Liu", "Yi Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-14", "url": "https://arxiv.org/abs/2506.12331", "pdf_url": "https://arxiv.org/pdf/2506.12331v1", "arxiv_id": "2506.12331", "doi": "10.48550/arXiv.2506.12331", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.1982}
{"id": "b6e4669914b9af8997dbe8e2b5a88f0dbede069c5ada481679c12cfcb9ba2e4d", "sources": ["arxiv", "semantic_scholar"], "title": "Tiered Agentic Oversight: A Hierarchical Multi-Agent System for Healthcare Safety", "abstract": "Large language models (LLMs) deployed as agents introduce significant safety risks in clinical settings due to their potential for error and single points of failure. We introduce Tiered Agentic Oversight (TAO), a hierarchical multi-agent system that enhances AI safety through layered, automated supervision. Inspired by clinical hierarchies (e.g., nurse-physician-specialist) in hospital, TAO routes tasks to specialized agents based on complexity, creating a robust safety framework through automated inter- and intra-tier communication and role-playing. Crucially, this hierarchical structure functions as an effective error-correction mechanism, absorbing up to 24% of individual agent errors before they can compound. Our experiments reveal TAO outperforms single-agent and other multi-agent systems on 4 out of 5 healthcare safety benchmarks, with up to an 8.2% improvement. Ablation studies confirm key design principles of the system: (i) its adaptive architecture is over 3% safer than static, single-tier configurations, and (ii) its lower tiers are indispensable, as their removal causes the most significant degradation in overall safety. Finally, we validated the system's synergy with human doctors in a user study where a physician, acting as the highest tier agent, provided corrective feedback that improved medical triage accuracy from 40% to 60%. Project Page: https://tiered-agentic-oversight.github.io/", "authors": ["Yubin Kim", "Hyewon Jeong", "Chanwoo Park", "Eugene Park", "Haipeng Zhang", "Xin Liu", "Hyeonhoon Lee", "Daniel McDuff", "Marzyeh Ghassemi", "Cynthia Breazeal", "Samir Tulebaev", "Hae Won Park"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-14", "url": "https://arxiv.org/abs/2506.12482", "pdf_url": "https://arxiv.org/pdf/2506.12482v2", "arxiv_id": "2506.12482", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1505}
{"id": "0980073109969af8da6cabe3ffd353977d589ddfcc58a21ff86de0c6ce2f6454", "sources": ["arxiv", "semantic_scholar"], "title": "Revealing Political Bias in LLMs through Structured Multi-Agent Debate", "abstract": "Large language models (LLMs) are increasingly used to simulate social behaviour, yet their political biases and interaction dynamics in debates remain underexplored. We investigate how LLM type and agent gender attributes influence political bias using a structured multi-agent debate framework, by engaging Neutral, Republican, and Democrat American LLM agents in debates on politically sensitive topics. We systematically vary the underlying LLMs, agent genders, and debate formats to examine how model provenance and agent personas influence political bias and attitudes throughout debates. We find that Neutral agents consistently align with Democrats, while Republicans shift closer to the Neutral; gender influences agent attitudes, with agents adapting their opinions when aware of other agents' genders; and contrary to prior research, agents with shared political affiliations can form echo chambers, exhibiting the expected intensification of attitudes as debates progress.", "authors": ["Aishwarya Bandaru", "Fabian Bindley", "Trevor Bluth", "Nandini Chavda", "Baixu Chen", "Ethan Law"], "categories": ["cs.AI", "cs.CY", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-13", "url": "https://arxiv.org/abs/2506.11825", "pdf_url": "https://arxiv.org/pdf/2506.11825v1", "arxiv_id": "2506.11825", "doi": "10.48550/arXiv.2506.11825", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1971}
{"id": "6fa5aebef8f24fd4884a9feccce10ae6d0a404bc81c0bc4fa6dedb97c42e94b5", "sources": ["arxiv", "semantic_scholar"], "title": "PE-MA: Parameter-Efficient Co-Evolution of Multi-Agent Systems", "abstract": "Multi-Agent Systems have recently emerged as a promising paradigm for collaborative reasoning and solving complex tasks. However, the design of collaborative learning algorithms in multi-agent systems faces several challenges, including high communication overhead and insufficient agent-level personalization. In this paper, we propose PE-MA (Parameter-Efficient Multi-Agent Co-Evolution), a novel collaboration framework that supports efficient, scalable, and personalized co-evolution in multi-agent systems. In PE-MA, each agent maintains a lightweight personalized adapter to support agent-specific behavior, while a shared adapter is collaboratively optimized across neighboring agents. This design balances global coordination with local adaptation under heterogeneous environments. We achieve an asymptotically optimal convergence rate of O( 1/(NK)^(1/2) ), where N is the number of agents and K the local update steps.", "authors": ["Yingfan Deng", "Anhao Zhou", "Yuan Yuan", "Xiao Zhang", "Yifei Zou", "Dongxiao Yu"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-13", "url": "https://arxiv.org/abs/2506.11803", "pdf_url": "https://arxiv.org/pdf/2506.11803v2", "arxiv_id": "2506.11803", "doi": "10.48550/arXiv.2506.11803", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1971}
{"id": "3fadeb5897ce4d21855f68cc4a24e6596beee75ba6a97ee6b16067bef959dd6a", "sources": ["arxiv", "semantic_scholar"], "title": "A Hybrid Adaptive Nash Equilibrium Solver for Distributed Multi-Agent Systems with Game-Theoretic Jump Triggering", "abstract": "This paper presents a hybrid adaptive Nash equilibrium solver for distributed multi-agent systems incorporating game-theoretic jump triggering mechanisms. The approach addresses fundamental scalability and computational challenges in multi-agent hybrid systems by integrating distributed game-theoretic optimization with systematic hybrid system design. A novel game-theoretic jump triggering mechanism coordinates discrete mode transitions across multiple agents while maintaining distributed autonomy. The Hybrid Adaptive Nash Equilibrium Solver (HANES) algorithm integrates these methodologies. Sufficient conditions establish exponential convergence to consensus under distributed information constraints. The framework provides rigorous stability guarantees through coupled Hamilton-Jacobi-Bellman equations while enabling rapid emergency response capabilities through coordinated jump dynamics. Simulation studies in pursuit-evasion and leader-follower consensus scenarios demonstrate significant improvements in convergence time, computational efficiency, and scalability compared to existing centralized and distributed approaches.", "authors": ["Qiuyu Miao", "Zhigang Wu"], "categories": ["eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-06-12", "url": "https://arxiv.org/abs/2506.11304", "pdf_url": "https://arxiv.org/pdf/2506.11304v1", "arxiv_id": "2506.11304", "doi": "10.48550/arXiv.2506.11304", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1959}
{"id": "d1517832edd3f55d9356803bffc797a0c71f5b3019eebce1046f9a8f71ff125c", "sources": ["arxiv", "semantic_scholar"], "title": "Specification and Evaluation of Multi-Agent LLM Systems -- Prototype and Cybersecurity Applications", "abstract": "Recent advancements in LLMs indicate potential for novel applications, as evidenced by the reasoning capabilities in the latest OpenAI and DeepSeek models. To apply these models to domain-specific applications beyond text generation, LLM-based multi-agent systems can be utilized to solve complex tasks, particularly by combining reasoning techniques, code generation, and software execution across multiple, potentially specialized LLMs. However, while many evaluations are performed on LLMs, reasoning techniques, and applications individually, their joint specification and combined application are not well understood. Defined specifications for multi-agent LLM systems are required to explore their potential and suitability for specific applications, allowing for systematic evaluations of LLMs, reasoning techniques, and related aspects. This paper reports the results of exploratory research on (1.) multi-agent specification by introducing an agent schema language and (2.) the execution and evaluation of the specifications through a multi-agent system architecture and prototype. The specification language, system architecture, and prototype are first presented in this work, building on an LLM system from prior research. Test cases involving cybersecurity tasks indicate the feasibility of the architecture and evaluation approach. As a result, evaluations could be demonstrated for question answering, server security, and network security tasks completed correctly by agents with LLMs from OpenAI and DeepSeek.", "authors": ["Felix Härer"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-12", "url": "https://arxiv.org/abs/2506.10467", "pdf_url": "https://arxiv.org/pdf/2506.10467v4", "arxiv_id": "2506.10467", "doi": "10.1109/Cyber-AI66431.2025.11233474", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "a6f15772541c0109503a75a229b1b994ca23bf573158e68480fb1152a89aeabf", "sources": ["arxiv", "semantic_scholar"], "title": "Live API-Bench: 2500+ Live APIs for Testing Multi-Step Tool Calling", "abstract": "Large language models (LLMs) increasingly rely on external tools and APIs to execute complex tasks specified in natural language. Evaluating such tool calling capabilities in realistic enterprise settings is challenging: APIs are often proprietary, heterogeneous, and difficult to share, limiting reproducible benchmarks. To address this, we introduce Live API Bench, a comprehensive benchmark constructed by transforming NL2SQL datasets into interactive API environments. Our pipeline converts SQL queries from BIRD SQL into executable API sequences across three formulations SLOT, SEL, and REST covering minimal general purpose operations, domain specific multi step tasks, and function oriented RESTful interactions, respectively. The benchmark spans 11 databases with over 2,500 invocable tools, paired with human authored queries, ground truth API sequences, and verified final answers. Live API Bench enables systematic evaluation of core challenges in tool use, including error handling, sequential reasoning, parameter generation, response parsing, and robustness across diverse domains. We evaluate 10 LLMs and 4 ReACT agents, observing low task completion rates (7 to 47pct), which improve modestly to 50pct under interactive agent settings, highlighting substantial scope for improving LLM tool calling performance. We release all code and data associated with this paper.", "authors": ["Benjamin Elder", "Anupama Murthi", "Jungkoo Kang", "Ankita Rajaram Naik", "Kiran Kate", "Kinjal Basu", "Danish Contractor"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-12", "url": "https://arxiv.org/abs/2506.11266", "pdf_url": "https://arxiv.org/pdf/2506.11266v2", "arxiv_id": "2506.11266", "doi": "10.18653/v1/2026.eacl-long.143", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference of the European Chapter of the Association for Computational Linguistics", "quality_score": 0.1959}
{"id": "69c7410b5323ac9319aeb2c2de25713972c9650af22306bb602959993bd14eb1", "sources": ["arxiv", "semantic_scholar"], "title": "Infected Smallville: How Disease Threat Shapes Sociality in LLM Agents", "abstract": "How does the threat of infectious disease influence sociality among generative agents? We used generative agent-based modeling (GABM), powered by large language models, to experimentally test hypotheses about the behavioral immune system. Across three simulation runs, generative agents who read news about an infectious disease outbreak showed significantly reduced social engagement compared to agents who received no such news, including lower attendance at a social gathering, fewer visits to third places (e.g., cafe, store, park), and fewer conversations throughout the town. In interview responses, agents explicitly attributed their behavioral changes to disease-avoidance motivations. A validity check further indicated that they could distinguish between infectious and noninfectious diseases, selectively reducing social engagement only when there was a risk of infection. Our findings highlight the potential of GABM as an experimental tool for exploring complex human social dynamics at scale.", "authors": ["Soyeon Choi", "Kangwook Lee", "Oliver Sng", "Joshua M. Ackerman"], "categories": ["physics.soc-ph", "cs.LG"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-06-10", "url": "https://arxiv.org/abs/2506.13783", "pdf_url": "https://arxiv.org/pdf/2506.13783v2", "arxiv_id": "2506.13783", "doi": "10.48550/arXiv.2506.13783", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1936}
{"id": "9bdd9b7d6c2aa4d42fca70feaea5183b721872e1b1551965b27f431aa2e4b32c", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Neural Networks: Self-Evolving Multi-Agent Systems via Textual Backpropagation", "abstract": "Leveraging multiple Large Language Models(LLMs) has proven effective for addressing complex, high-dimensional tasks, but current approaches often rely on static, manually engineered multi-agent configurations. To overcome these constraints, we present the Agentic Neural Network(ANN), a framework that conceptualizes multi-agent collaboration as a layered neural network architecture. In this design, each agent operates as a node, and each layer forms a cooperative \"team\" focused on a specific subtask. Agentic Neural Network follows a two-phase optimization strategy: (1) Forward Phase-Drawing inspiration from neural network forward passes, tasks are dynamically decomposed into subtasks, and cooperative agent teams with suitable aggregation methods are constructed layer by layer. (2) Backward Phase-Mirroring backpropagation, we refine both global and local collaboration through iterative feedback, allowing agents to self-evolve their roles, prompts, and coordination. This neuro-symbolic approach enables ANN to create new or specialized agent teams post-training, delivering notable gains in accuracy and adaptability. Across four benchmark datasets, ANN surpasses leading multi-agent baselines under the same configurations, showing consistent performance improvements. Our findings indicate that ANN provides a scalable, data-driven framework for multi-agent systems, combining the collaborative capabilities of LLMs with the efficiency and flexibility of neural network principles. We plan to open-source the entire framework.", "authors": ["Xiaowen Ma", "Chenyang Lin", "Yao Zhang", "Volker Tresp", "Yunpu Ma"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-10", "url": "https://arxiv.org/abs/2506.09046", "pdf_url": "https://arxiv.org/pdf/2506.09046v2", "arxiv_id": "2506.09046", "doi": "10.48550/arXiv.2506.09046", "citation_count": 12, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2993}
{"id": "ea2a0a449bdfb7306d569bb0a349e02fda574f82a7e34999da371a3c1b07d8a5", "sources": ["arxiv", "semantic_scholar"], "title": "ORFS-agent: Tool-Using Agents for Chip Design Optimization", "abstract": "Machine learning has been widely used to optimize complex engineering workflows across numerous domains. In integrated circuit design, modern flows (e.g., register-transfer level to physical layout) involve extensive configuration via thousands of parameters, and small changes can have large downstream impacts on design performance, power, and area. Recent advances in Large Language Models (LLMs) offer new opportunities for learning and reasoning within such high-dimensional optimization tasks. In this work, we introduce ORFS-agent, an LLM-based iterative optimization agent that automates parameter tuning in an open-source hardware design flow. ORFS-agent adaptively explores parameter configurations, demonstrating improvements over standard Bayesian optimization approaches in terms of resource efficiency and final design metrics. Across six benchmarks on ASAP7 and SKY130HD, thinking-model backends (Sonnet 4.6 [69] and Kimi K2.5 [28]) improve the geometric-mean normalized wirelength, effective clock period, and co-optimization objectives by up to 1.0%, 1.3%, and 2.7% over OR-AutoTuner while using 40% fewer iterations; the open-weight Kimi K2.5 remains within 0.24% of Sonnet 4.6, enabling private deployment. Relative to the earlier Sonnet 3.5 backend, these thinking models improve the same objectives by up to 7.5%, 3.1%, and 4.0%. Optional retrieval tools accelerate early convergence but do not improve final endpoints. By following natural language objectives to trade off certain metrics for others, ORFS-agent demonstrates a flexible and interpretable framework for multi-objective and constrained optimization. Crucially, ORFS-agent is modular and model-agnostic, and can be plugged into any frontier LLM without any further fine-tuning. We also report checkpoint-aligned trajectories and reasoning summaries that document the agent's decision process.", "authors": ["Amur Ghose", "Andrew B. Kahng", "Sayak Kundu", "Zhiang Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-10", "url": "https://arxiv.org/abs/2506.08332", "pdf_url": "https://arxiv.org/pdf/2506.08332v3", "arxiv_id": "2506.08332", "doi": "10.1109/MLCAD65511.2025.11189204", "citation_count": 13, "influential_citation_count": 4, "has_code": true, "code_url": null, "venue": "Workshop on Machine Learning for CAD", "quality_score": 0.3495}
{"id": "2ae8f5b49392ecb14e7fca76ea9779bb777971d5620882b70db020f057acd9a3", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforce LLM Reasoning through Multi-Agent Reflection", "abstract": "Leveraging more test-time computation has proven to be an effective way to boost the reasoning capabilities of large language models (LLMs). Among various methods, the verify-and-improve paradigm stands out for enabling dynamic solution exploration and feedback incorporation. However, existing approaches often suffer from restricted feedback spaces and lack of coordinated training of different parties, leading to suboptimal performance. To address this, we model this multi-turn refinement process as a Markov Decision Process and introduce DPSDP (Direct Policy Search by Dynamic Programming), a reinforcement learning algorithm that trains an actor-critic LLM system to iteratively refine answers via direct preference learning on self-generated data. Theoretically, DPSDP can match the performance of any policy within the training distribution. Empirically, we instantiate DPSDP with various base models and show improvements on both in- and out-of-distribution benchmarks. For example, on benchmark MATH 500, majority voting over five refinement steps increases first-turn accuracy from 58.2% to 63.2% with Ministral-based models. An ablation study further confirms the benefits of multi-agent collaboration and out-of-distribution generalization.", "authors": ["Yurun Yuan", "Tengyang Xie"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-10", "url": "https://arxiv.org/abs/2506.08379", "pdf_url": "https://arxiv.org/pdf/2506.08379v1", "arxiv_id": "2506.08379", "doi": "10.48550/arXiv.2506.08379", "citation_count": 30, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.3728}
{"id": "875426d5fa39bf6a8af6441bff39365ac66f50b4670882af9799e8f4307a6fa3", "sources": ["arxiv", "semantic_scholar"], "title": "CAF-I: A Collaborative Multi-Agent Framework for Enhanced Irony Detection with Large Language Models", "abstract": "Large language model (LLM) have become mainstream methods in the field of sarcasm detection. However, existing LLM methods face challenges in irony detection, including: 1. single-perspective limitations, 2. insufficient comprehensive understanding, and 3. lack of interpretability. This paper introduces the Collaborative Agent Framework for Irony (CAF-I), an LLM-driven multi-agent system designed to overcome these issues. CAF-I employs specialized agents for Context, Semantics, and Rhetoric, which perform multidimensional analysis and engage in interactive collaborative optimization. A Decision Agent then consolidates these perspectives, with a Refinement Evaluator Agent providing conditional feedback for optimization. Experiments on benchmark datasets establish CAF-I's state-of-the-art zero-shot performance. Achieving SOTA on the vast majority of metrics, CAF-I reaches an average Macro-F1 of 76.31, a 4.98 absolute improvement over the strongest prior baseline. This success is attained by its effective simulation of human-like multi-perspective analysis, enhancing detection accuracy and interpretability.", "authors": ["Ziqi. Liu", "Ziyang. Zhou", "Mingxuan. Hu"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-10", "url": "https://arxiv.org/abs/2506.08430", "pdf_url": "https://arxiv.org/pdf/2506.08430v2", "arxiv_id": "2506.08430", "doi": "10.48550/arXiv.2506.08430", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Neural Information Processing", "quality_score": 0.1945}
{"id": "fcd6eafef540314b537db9b5ce0bb27a9c90e1f318316bc087853bf619bd1237", "sources": ["arxiv", "semantic_scholar"], "title": "From Debate to Equilibrium: Belief-Driven Multi-Agent LLM Reasoning via Bayesian Nash Equilibrium", "abstract": "Multi-agent frameworks can substantially boost the reasoning power of large language models (LLMs), but they typically incur heavy computational costs and lack convergence guarantees. To overcome these challenges, we recast multi-LLM coordination as an incomplete-information game and seek a Bayesian Nash equilibrium (BNE), in which each agent optimally responds to its probabilistic beliefs about the strategies of others. We introduce Efficient Coordination via Nash Equilibrium (ECON), a hierarchical reinforcement-learning paradigm that marries distributed reasoning with centralized final output. Under ECON, each LLM independently selects responses that maximize its expected reward, conditioned on its beliefs about co-agents, without requiring costly inter-agent exchanges. We mathematically prove that ECON attains a markedly tighter regret bound than non-equilibrium multi-agent schemes. Empirically, ECON outperforms existing multi-LLM approaches by 11.2% on average across six benchmarks spanning complex reasoning and planning tasks. Further experiments demonstrate ECON's ability to flexibly incorporate additional models, confirming its scalability and paving the way toward larger, more powerful multi-LLM ensembles. The code is publicly available at: https://github.com/tmlr-group/ECON.", "authors": ["Xie Yi", "Zhanke Zhou", "Chentao Cao", "Qiyu Niu", "Tongliang Liu", "Bo Han"], "categories": ["cs.LG", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-09", "url": "https://arxiv.org/abs/2506.08292", "pdf_url": "https://arxiv.org/pdf/2506.08292v1", "arxiv_id": "2506.08292", "doi": "10.48550/arXiv.2506.08292", "citation_count": 25, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/tmlr-group/ECON", "venue": "International Conference on Machine Learning", "quality_score": 0.3537}
{"id": "d8596cb32699160ca4c3a50041209fccead27181aee656550510d1516643345c", "sources": ["arxiv", "semantic_scholar"], "title": "SOP-Bench: Complex Industrial SOPs for Evaluating LLM Agents", "abstract": "LLM-based agents struggle to execute complex, multi-step Standard Operating Procedures (SOPs) that are fundamental to industrial automation. Existing benchmarks fail to capture the procedural complexity and tool orchestration demands of real-world workflows. We introduce SOP-Bench, a benchmark of 2,000+ tasks from human expert-authored SOPs across 12 business domains (healthcare, logistics, finance, content moderation, etc.). Using a human-AI collaborative framework, experts crafted authentic SOPs while AI generated artifacts (tools, APIs, datasets), all human-validated, yielding realistic tasks with executable interfaces and ground-truth outputs. SOP-Bench serves as a research enabler for systematically investigating agent architectures, model capabilities, and deployment considerations across diverse procedural tasks. We demonstrate its utility through illustrative experiments with a subset of frontier models across Function-Calling (FC) and ReAct agents, revealing critical insights. For example, (1) newer models do not guarantee better performance - Claude 4 family outperforms Claude 4.5 family on ReAct tasks (Claude 4 Opus: 72.4% vs. Claude 4.5 Sonnet: 63.3% task success rate), demonstrating that production upgrades require validation; (2) no single model-agent combination dominates: best performances range from 57% to 100% depending on domain. These examples illustrate how SOP-Bench enables isolating and studying specific dimensions of agent performance without costly production experiments. Our goal is not to rank model capabilities or build optimal agents, but to provide a rigorous evaluation framework that enables the researchers and practitioners to systematically investigate agent design choices, model selection, and deployment strategies. We release the benchmark at https://github.com/amazon-science/sop-bench.", "authors": ["Subhrangshu Nandi", "Arghya Datta", "Rohith Nama", "Udita Patel", "Nikhil Vichare", "Indranil Bhattacharya", "Prince Grover", "Shivam Asija", "Giuseppe Carenini", "Wei Zhang", "Arushi Gupta", "Sreyoshi Bhaduri", "Jing Xu", "Huzefa Raja", "Shayan Ray", "Aaron Chan", "Esther Xu Fei", "Gaoyuan Du", "Zuhaib Akhtar", "Harshita Asnani", "Weian Chan", "Ming Xiong", "Francesco Carbone", "Jeetu Mirchandani"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-09", "url": "https://arxiv.org/abs/2506.08119", "pdf_url": "https://arxiv.org/pdf/2506.08119v2", "arxiv_id": "2506.08119", "doi": "10.48550/arXiv.2506.08119", "citation_count": 10, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/amazon-science/sop-bench", "venue": "arXiv.org", "quality_score": 0.2975}
{"id": "650f4d92dd3f4272495380010914f3087c003ee4565eb92d90c4570cb8637ae7", "sources": ["arxiv", "semantic_scholar"], "title": "Shapley-Coop: Credit Assignment for Emergent Cooperation in Self-Interested LLM Agents", "abstract": "Large Language Models (LLMs) show strong collaborative performance in multi-agent systems with predefined roles and workflows. However, in open-ended environments lacking coordination rules, agents tend to act in self-interested ways. The central challenge in achieving coordination lies in credit assignment -- fairly evaluating each agent's contribution and designing pricing mechanisms that align their heterogeneous goals. This problem is critical as LLMs increasingly participate in complex human-AI collaborations, where fair compensation and accountability rely on effective pricing mechanisms. Inspired by how human societies address similar coordination challenges (e.g., through temporary collaborations such as employment or subcontracting), we propose a cooperative workflow, Shapley-Coop. Shapley-Coop integrates Shapley Chain-of-Thought -- leveraging marginal contributions as a principled basis for pricing -- with structured negotiation protocols for effective price matching, enabling LLM agents to coordinate through rational task-time pricing and post-task reward redistribution. This approach aligns agent incentives, fosters cooperation, and maintains autonomy. We evaluate Shapley-Coop across two multi-agent games and a software engineering simulation, demonstrating that it consistently enhances LLM agent collaboration and facilitates equitable credit assignment. These results highlight the effectiveness of Shapley-Coop's pricing mechanisms in accurately reflecting individual contributions during task execution.", "authors": ["Yun Hua", "Haosheng Chen", "Shiqin Wang", "Wenhao Li", "Xiangfeng Wang", "Jun Luo"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-09", "url": "https://arxiv.org/abs/2506.07388", "pdf_url": "https://arxiv.org/pdf/2506.07388v1", "arxiv_id": "2506.07388", "doi": "10.48550/arXiv.2506.07388", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "a498b1729851860132ac2d0151243df14de16aaf005a761835ea9d87158da079", "sources": ["arxiv", "semantic_scholar"], "title": "Learn as Individuals, Evolve as a Team: Multi-agent LLMs Adaptation in Embodied Environments", "abstract": "Large language models (LLMs) possess extensive knowledge bases and strong reasoning capabilities, making them promising tools for complex, multi-agent planning in embodied environments. However, despite LLMs' advanced abilities and the sophisticated modular design of agentic methods, existing LLM-based planning algorithms remain limited by weak adaptation capabilities to multi-agent embodied scenarios. We address this limitation by introducing a framework that enables LLM agents to learn and evolve both before and during test time, equipping them with environment-relevant knowledge for better planning and enhanced communication for improved cooperation. Inspired by centralized training with decentralized execution in multi-agent reinforcement learning, we propose a \\textit{Learn as Individuals, Evolve as a Team (LIET)} paradigm for multi-agent LLMs adaptation. At the individual level, LLM agents learn a local utility function from exploratory datasets to better comprehend the embodied environment, which is then queried during test time to support informed decision-making. At the team level, LLM agents collaboratively and iteratively maintain and update a shared cooperation knowledge list based on new experiences, using it to guide more effective communication. By combining individual learning with team evolution, LIET enables comprehensive and flexible adaptation for LLM agents. Our experiments on Communicative Watch-And-Help and ThreeD-World Multi-Agent Transport benchmarks demonstrate that LIET, instantiated with both LLaMA and GPT-4o, outperforms existing baselines and exhibits strong cooperative planning abilities.", "authors": ["Xinran Li", "Chenjia Bai", "Zijian Li", "Jiakun Zheng", "Ting Xiao", "Jun Zhang"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-08", "url": "https://arxiv.org/abs/2506.07232", "pdf_url": "https://arxiv.org/pdf/2506.07232v1", "arxiv_id": "2506.07232", "doi": "10.48550/arXiv.2506.07232", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1914}
{"id": "859304f043f03d050e958ab6f888954a259f03ccd640292168eaa4a724fafbda", "sources": ["arxiv", "semantic_scholar"], "title": "Agents of Change: Self-Evolving LLM Agents for Strategic Planning", "abstract": "We address the long-horizon gap in large language model (LLM) agents by enabling them to sustain coherent strategies in adversarial, stochastic environments. Settlers of Catan provides a challenging benchmark: success depends on balancing short- and long-term goals amid randomness, trading, expansion, and blocking. Prompt-centric LLM agents (e.g., ReAct, Reflexion) must re-interpret large, evolving game states each turn, quickly saturating context windows and losing strategic consistency. We propose HexMachina, a continual learning multi-agent system that separates environment discovery (inducing an adapter layer without documentation) from strategy improvement (evolving a compiled player through code refinement and simulation). This design preserves executable artifacts, allowing the LLM to focus on high-level strategy rather than per-turn reasoning. In controlled Catanatron experiments, HexMachina learns from scratch and evolves players that outperform the strongest human-crafted baseline (AlphaBeta), achieving a 54% win rate and surpassing prompt-driven and no-discovery baselines. Ablations confirm that isolating pure strategy learning improves performance. Overall, artifact-centric continual learning transforms LLMs from brittle stepwise deciders into stable strategy designers, advancing long-horizon autonomy.", "authors": ["Nikolas Belle", "Dakota Barnes", "Alfonso Amayuelas", "Ivan Bercovich", "Xin Eric Wang", "William Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-05", "url": "https://arxiv.org/abs/2506.04651", "pdf_url": "https://arxiv.org/pdf/2506.04651v2", "arxiv_id": "2506.04651", "doi": "10.48550/arXiv.2506.04651", "citation_count": 16, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "f267ea3f4b6ec0c25a7123bc9d69514d7a3a756a80098f6d5d6e8a62fb87aab0", "sources": ["arxiv", "semantic_scholar"], "title": "TRiSM for Agentic AI: A Review of Trust, Risk, and Security Management in LLM-based Agentic Multi-Agent Systems", "abstract": "Agentic AI systems, built upon large language models (LLMs) and deployed in multi-agent configurations, are redefining intelligence, autonomy, collaboration, and decision-making across enterprise and societal domains. This review presents a structured analysis of Trust, Risk, and Security Management (TRiSM) in the context of LLM-based Agentic Multi-Agent Systems (AMAS). We begin by examining the conceptual foundations of Agentic AI and highlight its architectural distinctions from traditional AI agents. We then adapt and extend the AI TRiSM framework for Agentic AI, structured around key pillars: \\textit{ Explainability, ModelOps, Security, Privacy} and \\textit{their Lifecycle Governance}, each contextualized to the challenges of AMAS. A risk taxonomy is proposed to capture the unique threats and vulnerabilities of Agentic AI, ranging from coordination failures to prompt-based adversarial manipulation. To support practical assessment in Agentic AI works, we introduce two novel metrics: the Component Synergy Score (CSS), which quantifies the quality of inter-agent collaboration, and the Tool Utilization Efficacy (TUE), which evaluates the efficiency of tool use within agent workflows. We further discuss strategies for improving explainability in Agentic AI, as well as approaches to enhancing security and privacy through encryption, adversarial robustness, and regulatory compliance. The review concludes with a research roadmap for the responsible development and deployment of Agentic AI, highlighting key directions to align emerging systems with TRiSM principles-ensuring safety, transparency, and accountability in their operation.", "authors": ["Shaina Raza", "Ranjan Sapkota", "Manoj Karkee", "Christos Emmanouilidis"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-04", "url": "https://arxiv.org/abs/2506.04133", "pdf_url": "https://arxiv.org/pdf/2506.04133v5", "arxiv_id": "2506.04133", "doi": "10.48550/arXiv.2506.04133", "citation_count": 74, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "AI Open", "quality_score": 0.4688}
{"id": "f41675e9013f32ef58e9e700b3a6904e080cd071d3f195531fa16c4d86073e51", "sources": ["arxiv", "semantic_scholar"], "title": "Graph Counselor: Adaptive Graph Exploration via Multi-Agent Synergy to Enhance LLM Reasoning", "abstract": "Graph Retrieval Augmented Generation (GraphRAG) effectively enhances external knowledge integration capabilities by explicitly modeling knowledge relationships, thereby improving the factual accuracy and generation quality of Large Language Models (LLMs) in specialized domains. However, existing methods suffer from two inherent limitations: 1) Inefficient Information Aggregation: They rely on a single agent and fixed iterative patterns, making it difficult to adaptively capture multi-level textual, structural, and degree information within graph data. 2) Rigid Reasoning Mechanism: They employ preset reasoning schemes, which cannot dynamically adjust reasoning depth nor achieve precise semantic correction. To overcome these limitations, we propose Graph Counselor, an GraphRAG method based on multi-agent collaboration. This method uses the Adaptive Graph Information Extraction Module (AGIEM), where Planning, Thought, and Execution Agents work together to precisely model complex graph structures and dynamically adjust information extraction strategies, addressing the challenges of multi-level dependency modeling and adaptive reasoning depth. Additionally, the Self-Reflection with Multiple Perspectives (SR) module improves the accuracy and semantic consistency of reasoning results through self-reflection and backward reasoning mechanisms. Experiments demonstrate that Graph Counselor outperforms existing methods in multiple graph reasoning tasks, exhibiting higher reasoning accuracy and generalization ability. Our code is available at https://github.com/gjq100/Graph-Counselor.git.", "authors": ["Junqi Gao", "Xiang Zou", "YIng Ai", "Dong Li", "Yichen Niu", "Biqing Qi", "Jianxing Liu"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-04", "url": "https://arxiv.org/abs/2506.03939", "pdf_url": "https://arxiv.org/pdf/2506.03939v1", "arxiv_id": "2506.03939", "doi": "10.48550/arXiv.2506.03939", "citation_count": 9, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/gjq100/Graph-Counselor.git", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.2886}
{"id": "2ccdf448097e9029453f1280992f2200c0499a1cb3c352bb260f21aa836f6ac7", "sources": ["arxiv", "semantic_scholar"], "title": "MAEBE: Multi-Agent Emergent Behavior Framework", "abstract": "Traditional AI safety evaluations on isolated LLMs are insufficient as multi-agent AI ensembles become prevalent, introducing novel emergent risks. This paper introduces the Multi-Agent Emergent Behavior Evaluation (MAEBE) framework to systematically assess such risks. Using MAEBE with the Greatest Good Benchmark (and a novel double-inversion question technique), we demonstrate that: (1) LLM moral preferences, particularly for Instrumental Harm, are surprisingly brittle and shift significantly with question framing, both in single agents and ensembles. (2) The moral reasoning of LLM ensembles is not directly predictable from isolated agent behavior due to emergent group dynamics. (3) Specifically, ensembles exhibit phenomena like peer pressure influencing convergence, even when guided by a supervisor, highlighting distinct safety and alignment challenges. Our findings underscore the necessity of evaluating AI systems in their interactive, multi-agent contexts.", "authors": ["Sinem Erisken", "Timothy Gothard", "Martin Leitgab", "Ram Potham"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.CY", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.03053", "pdf_url": "https://arxiv.org/pdf/2506.03053v2", "arxiv_id": "2506.03053", "doi": "10.48550/arXiv.2506.03053", "citation_count": 7, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Proceedings of the Annual Hawaii International Conference on System Sciences", "quality_score": 0.2386}
{"id": "dc547449ffb02c76b60ce3d807899d7154a19aa645d6df7e63711dd4738d3493", "sources": ["arxiv", "semantic_scholar"], "title": "End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement Learning", "abstract": "Large language models (LLMs) are versatile, yet their deployment in complex real-world settings is limited by static knowledge cutoffs and the difficulty of producing controllable behavior within a single inference. Multi-agent search systems (MASS), which coordinate specialized LLM agents equipped with search tools, mitigate these issues via task decomposition and retrieval-augmented problem solving. However, optimizing LLMs for agent-specific roles remains labor-intensive with prompt engineering or supervised fine-tuning, motivating automated end-to-end training. Existing multi-agent reinforcement learning (MARL) methods such as Multi-Agent Proximal Policy Optimization (MAPPO) typically depend on large critic networks to evaluate joint actions, leading to instability and high memory costs. We introduce Multi-Agent Heterogeneous Group Policy Optimization (MHGPO), which updates policies by estimating relative advantages across heterogeneous groups of multi-agent rollouts, shifting the optimization focus from local agent performance to global system success. We further study three group rollout sampling strategies to trade off sample efficiency and optimization quality. Experiments show that MHGPO captures implicit inter-agent dependencies and consistently outperforms strong baselines in both task performance and computational efficiency.", "authors": ["Guanzhong Chen", "Shaoxiong Yang", "Chao Li", "Wei Liu", "Jian Luan", "Zenglin Xu"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.02718", "pdf_url": "https://arxiv.org/pdf/2506.02718v2", "arxiv_id": "2506.02718", "doi": null, "citation_count": 7, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "1c8f30ef6ea59b5d77f088f40df34931427f96f130ce1d15f414e372af9ee651", "sources": ["arxiv", "semantic_scholar"], "title": "To trust or not to trust: Attention-based Trust Management for LLM Multi-Agent Systems", "abstract": "Large Language Model-based Multi-Agent Systems (LLM-MAS) have demonstrated strong capabilities in solving complex tasks but remain vulnerable when agents receive unreliable messages. This vulnerability stems from a fundamental gap: LLM agents treat all incoming messages equally without evaluating their trustworthiness. While some existing studies approach trustworthiness, they focus on a single type of harmfulness rather than analyze it in a holistic approach from multiple trustworthiness perspectives. We address this gap by proposing a comprehensive definition of trustworthiness inspired by human communication theory (Grice, 1975). Our definition identifies six orthogonal trust dimensions that provide interpretable measures of trustworthiness. Building on this definition, we introduce the Attention Trust Score (A -Trust), a lightweight, attention-based method for evaluating the trustworthiness of messages. We then develop a principled trust management system (TMS) for LLM -MAS that supports both message-level and agent-level trust assessments. Experiments across diverse multi-agent settings and tasks demonstrate that our TMS significantly improves robustness against malicious inputs.", "authors": ["Pengfei He", "Zhenwei Dai", "Xianfeng Tang", "Yue Xing", "Hui Liu", "Jingying Zeng", "Qiankun Peng", "Shrivats Agrawal", "Samarth Varshney", "Suhang Wang", "Jiliang Tang", "Qi He"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.02546", "pdf_url": "https://arxiv.org/pdf/2506.02546v2", "arxiv_id": "2506.02546", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "a00b7060921f162db7759fd7e9aeec093ff9633fb988856eb00964c483bd79eb", "sources": ["arxiv", "semantic_scholar"], "title": "Adaptive Graph Pruning for Multi-Agent Communication", "abstract": "Large Language Model (LLM) based multi-agent systems have shown remarkable performance in various tasks, especially when enhanced through collaborative communication. However, current methods often rely on a fixed number of agents and static communication structures, limiting their ability to adapt to varying task complexities. In this paper, we propose Adaptive Graph Pruning (AGP), a novel task-adaptive multi-agent collaboration framework that jointly optimizes agent quantity (hard-pruning) and communication topology (soft-pruning). Specifically, our method employs a two-stage training strategy: firstly, independently training soft-pruning networks for different agent quantities to determine optimal agent-quantity-specific complete graphs and positional masks across specific tasks; and then jointly optimizing hard-pruning and soft-pruning within a maximum complete graph to dynamically configure the number of agents and their communication topologies per task. Extensive experiments demonstrate that our approach is: (1) High-performing, achieving state-of-the-art results across six benchmarks and consistently generalizes across multiple mainstream LLM architectures, with a increase in performance of $2.58\\%\\sim 9.84\\%$; (2) Task-adaptive, dynamically constructing optimized communication topologies tailored to specific tasks, with an extremely high performance in all three task categories (general reasoning, mathematical reasoning, and code generation); (3) Token-economical, having fewer training steps and token consumption at the same time, with a decrease in token consumption of $90\\%+$; and (4) Training-efficient, achieving high performance with very few training steps compared with other methods. The performance will surpass the existing baselines after about ten steps of training under six benchmarks.", "authors": ["Boyi Li", "Zhonghan Zhao", "Der-Horng Lee", "Gaoang Wang"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.02951", "pdf_url": "https://arxiv.org/pdf/2506.02951v3", "arxiv_id": "2506.02951", "doi": "10.48550/arXiv.2506.02951", "citation_count": 17, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.3138}
{"id": "b19b758e3a68e83c921149e4eb6f5723bf5ed88ae0a23f8e50d9a931c2f6fbd4", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Markov Entanglement", "abstract": "Value decomposition has long been a fundamental technique in multi-agent dynamic programming and reinforcement learning (RL). Specifically, the value function of a global state $(s_1,s_2,\\ldots,s_N)$ is often approximated as the sum of local functions: $V(s_1,s_2,\\ldots,s_N)\\approx\\sum_{i=1}^N V_i(s_i)$. This approach traces back to the index policy in restless multi-armed bandit problems and has found various applications in modern RL systems. However, the theoretical justification for why this decomposition works so effectively remains underexplored. In this paper, we uncover the underlying mathematical structure that enables value decomposition. We demonstrate that a multi-agent Markov decision process (MDP) permits value decomposition if and only if its transition matrix is not \"entangled\" -- a concept analogous to quantum entanglement in quantum physics. Drawing inspiration from how physicists measure quantum entanglement, we introduce how to measure the \"Markov entanglement\" for multi-agent MDPs and show that this measure can be used to bound the decomposition error in general multi-agent MDPs. Using the concept of Markov entanglement, we proved that a widely-used class of index policies is weakly entangled and enjoys a sublinear $\\mathcal O(\\sqrt{N})$ scale of decomposition error for $N$-agent systems. Finally, we show how Markov entanglement can be efficiently estimated in practice, providing practitioners with an empirical proxy for the quality of value decomposition.", "authors": ["Shuze Chen", "Tianyi Peng"], "categories": ["cs.LG", "quant-ph", "stat.ML"], "fields_of_study": ["Computer Science", "Physics", "Mathematics"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.02385", "pdf_url": "https://arxiv.org/pdf/2506.02385v3", "arxiv_id": "2506.02385", "doi": "10.48550/arXiv.2506.02385", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1856}
{"id": "7d266af3335d8f330fa95479fcd20e9df869ff88948690e66b39207978bc24a4", "sources": ["arxiv", "semantic_scholar"], "title": "DIAMOND: An LLM-Driven Agent for Context-Aware Baseball Highlight Summarization", "abstract": "Traditional approaches -- such as Win Probability Added (WPA)-based ranking or computer vision-driven event detection -- can identify scoring plays but often miss strategic depth, momentum shifts, and storyline progression. Manual curation remains the gold standard but is resource-intensive and not scalable. We introduce DIAMOND, an LLM-driven agent for context-aware baseball highlight summarization that integrates structured sports analytics with natural language reasoning. DIAMOND leverages sabermetric features -- Win Expectancy, WPA, and Leverage Index -- to quantify play importance, while an LLM module enhances selection based on contextual narrative value. This hybrid approach ensures both quantitative rigor and qualitative richness, surpassing the limitations of purely statistical or vision-based systems. Evaluated on five diverse Korean Baseball Organization League games, DIAMOND improves F1-score from 42.9% (WPA-only) to 84.8%, outperforming both commercial and statistical baselines. Though limited in scale, our results highlight the potential of modular, interpretable agent-based frameworks for event-level summarization in sports and beyond.", "authors": ["Jeonghun Kang", "Soonmok Kwon", "Joonseok Lee", "Byung-Hak Kim"], "categories": ["cs.CL", "cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-03", "url": "https://arxiv.org/abs/2506.02351", "pdf_url": "https://arxiv.org/pdf/2506.02351v1", "arxiv_id": "2506.02351", "doi": "10.48550/arXiv.2506.02351", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1181}
{"id": "0c1a70a8dc104664dcedfbed91f727373258a03c06eaae38ea1a4f60951ec67c", "sources": ["arxiv", "semantic_scholar"], "title": "An Empirical Study of Group Conformity in Multi-Agent Systems", "abstract": "Recent advances in Large Language Models (LLMs) have enabled multi-agent systems that simulate real-world interactions with near-human reasoning. While previous studies have extensively examined biases related to protected attributes such as race, the emergence and propagation of biases on socially contentious issues in multi-agent LLM interactions remain underexplored. This study explores how LLM agents shape public opinion through debates on five contentious topics. By simulating over 2,500 debates, we analyze how initially neutral agents, assigned a centrist disposition, adopt specific stances over time. Statistical analyses reveal significant group conformity mirroring human behavior; LLM agents tend to align with numerically dominant groups or more intelligent agents, exerting a greater influence. These findings underscore the crucial role of agent intelligence in shaping discourse and highlight the risks of bias amplification in online interactions. Our results emphasize the need for policy measures that promote diversity and transparency in LLM-generated discussions to mitigate the risks of bias propagation within anonymous online environments.", "authors": ["Min Choi", "Keonwoo Kim", "Sungwon Chae", "Sangyeob Baek"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-02", "url": "https://arxiv.org/abs/2506.01332", "pdf_url": "https://arxiv.org/pdf/2506.01332v1", "arxiv_id": "2506.01332", "doi": "10.48550/arXiv.2506.01332", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.1945}
{"id": "e1e0b92acff4f203c470452f72da07ffde7bf31cad2d5ff55ea8711f721ff135", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond Static Responses: Multi-Agent LLM Systems as a New Paradigm for Social Science Research", "abstract": "As large language models (LLMs) transition from static tools to fully agentic systems, their potential for transforming social science research has become increasingly evident. This paper introduces a structured framework for understanding the diverse applications of LLM-based agents, ranging from simple data processors to complex, multi-agent systems capable of simulating emergent social dynamics. By mapping this developmental continuum across six levels, the paper clarifies the technical and methodological boundaries between different agentic architectures, providing a comprehensive overview of current capabilities and future potential. It highlights how lower-tier systems streamline conventional tasks like text classification and data annotation, while higher-tier systems enable novel forms of inquiry, including the study of group dynamics, norm formation, and large-scale social processes. However, these advancements also introduce significant challenges, including issues of reproducibility, ethical oversight, and the risk of emergent biases. The paper critically examines these concerns, emphasizing the need for robust validation protocols, interdisciplinary collaboration, and standardized evaluation metrics. It argues that while LLM-based agents hold transformative potential for the social sciences, realizing this promise will require careful, context-sensitive deployment and ongoing methodological refinement. The paper concludes with a call for future research that balances technical innovation with ethical responsibility, encouraging the development of agentic systems that not only replicate but also extend the frontiers of social science, offering new insights into the complexities of human behavior.", "authors": ["Jennifer Haase", "Sebastian Pokutta"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-02", "url": "https://arxiv.org/abs/2506.01839", "pdf_url": "https://arxiv.org/pdf/2506.01839v3", "arxiv_id": "2506.01839", "doi": "10.48550/arXiv.2506.01839", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "0f9c86bfaa26f7ed131d76e4ede9067b33d2045b10a2bb0c9ae6810ec0dc14bf", "sources": ["arxiv", "semantic_scholar"], "title": "LAMARL: LLM-Aided Multi-Agent Reinforcement Learning for Cooperative Policy Generation", "abstract": "Although Multi-Agent Reinforcement Learning (MARL) is effective for complex multi-robot tasks, it suffers from low sample efficiency and requires iterative manual reward tuning. Large Language Models (LLMs) have shown promise in single-robot settings, but their application in multi-robot systems remains largely unexplored. This paper introduces a novel LLM-Aided MARL (LAMARL) approach, which integrates MARL with LLMs, significantly enhancing sample efficiency without requiring manual design. LAMARL consists of two modules: the first module leverages LLMs to fully automate the generation of prior policy and reward functions. The second module is MARL, which uses the generated functions to guide robot policy training effectively. On a shape assembly benchmark, both simulation and real-world experiments demonstrate the unique advantages of LAMARL. Ablation studies show that the prior policy improves sample efficiency by an average of 185.9% and enhances task completion, while structured prompts based on Chain-of-Thought (CoT) and basic APIs improve LLM output success rates by 28.5%-67.5%. Videos and code are available at https://windylab.github.io/LAMARL/", "authors": ["Guobin Zhu", "Rui Zhou", "Wenkang Ji", "Shiyu Zhao"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-02", "url": "https://arxiv.org/abs/2506.01538", "pdf_url": "https://arxiv.org/pdf/2506.01538v2", "arxiv_id": "2506.01538", "doi": "10.1109/LRA.2025.3577527", "citation_count": 16, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE Robotics and Automation Letters", "quality_score": 0.3076}
{"id": "6b73d7c53c692551811a2ea34ab33fc0eaf85c88f1c896abe088080d72ab3bba", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic AI and Multiagentic: Are We Reinventing the Wheel?", "abstract": "The terms Agentic AI and Multiagentic AI have recently gained popularity in discussions on generative artificial intelligence, often used to describe autonomous software agents and systems composed of such agents. However, the use of these terms confuses these buzzwords with well-established concepts in AI literature: intelligent agents and multi-agent systems. This article offers a critical analysis of this conceptual misuse. We review the theoretical origins of \"agentic\" in the social sciences (Bandura, 1986) and philosophical notions of intentionality (Dennett, 1971), and then summarise foundational works on intelligent agents and multi-agent systems by Wooldridge, Jennings and others. We examine classic agent architectures, from simple reactive agents to Belief-Desire-Intention (BDI) models, and highlight key properties (autonomy, reactivity, proactivity, social capability) that define agency in AI. We then discuss recent developments in large language models (LLMs) and agent platforms based on LLMs, including the emergence of LLM-powered AI agents and open-source multi-agent orchestration frameworks. We argue that the term AI Agentic is often used as a buzzword for what are essentially AI agents, and AI Multiagentic for what are multi-agent systems. This confusion overlooks decades of research in the field of autonomous agents and multi-agent systems. The article advocates for scientific and technological rigour and the use of established terminology from the state of the art in AI, incorporating the wealth of existing knowledge, including standards for multi-agent system platforms, communication languages and coordination and cooperation algorithms, agreement technologies (automated negotiation, argumentation, virtual organisations, trust, reputation, etc.), into the new and promising wave of LLM-based AI agents, so as not to end up reinventing the wheel.", "authors": ["V. Botti"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-02", "url": "https://arxiv.org/abs/2506.01463", "pdf_url": "https://arxiv.org/pdf/2506.01463v1", "arxiv_id": "2506.01463", "doi": "10.48550/arXiv.2506.01463", "citation_count": 6, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2851}
{"id": "44fa825485f64f77c88a26583167c984c471a5614313452d5e539d9672e7856a", "sources": ["arxiv", "semantic_scholar"], "title": "Will Agents Replace Us? Perceptions of Autonomous Multi-Agent AI", "abstract": "Autonomous multi-agent AI systems are poised to transform various industries, particularly software development and knowledge work. Understanding current perceptions among professionals is crucial for anticipating adoption challenges, ethical considerations, and future workforce development. This study analyzes responses from 130 participants to a survey on the capabilities, impact, and governance of AI agents. We explore expected timelines for AI replacing programmers, identify perceived barriers to deployment, and examine beliefs about responsibility when agents make critical decisions. Key findings reveal three distinct clusters of respondents. While the study explored factors associated with current AI agent deployment, the initial logistic regression model did not yield statistically significant predictors, suggesting that deployment decisions are complex and may be influenced by factors not fully captured or that a larger sample is needed. These insights highlight the need for organizations to address compliance concerns (a commonly cited barrier) and establish clear governance frameworks as they integrate autonomous agents into their workflows.", "authors": ["Nikola Balic"], "categories": ["cs.CY", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-01", "url": "https://arxiv.org/abs/2506.02055", "pdf_url": "https://arxiv.org/pdf/2506.02055v1", "arxiv_id": "2506.02055", "doi": "10.48550/arXiv.2506.02055", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/nibzard/agent-perceptions", "venue": "arXiv.org", "quality_score": 0.2833}
{"id": "cbfb4aa73c030ba663f62d6e68685906a8b5698e69b49c1fbdf3161ce6fb89f2", "sources": ["arxiv", "semantic_scholar"], "title": "MCP-Zero: Active Tool Discovery for Autonomous LLM Agents", "abstract": "True intelligence requires active capability acquisition, yet current LLM agents inject pre-defined tool schemas into prompts, reducing models to passive selectors and falling short of robust general-purpose agency. We introduce MCP-Zero, an active agent framework that restores tool discovery autonomy to LLMs themselves. Instead of overwhelming models with all available tools, MCP-Zero enables agents to actively identify capability gaps, and request specific tools on-demand, transforming them from large-scale retrievers into genuine autonomous agents. The framework operates through three core mechanisms: (1) Active Tool Request, where models autonomously generate structured requests specifying their exact tool requirements; (2) Hierarchical Semantic Routing, a two-stage algorithm that matches requests to relevant servers and tools through improved semantic alignment; (3) Iterative Capability Extension, enabling agents to progressively build cross-domain toolchains while maintaining minimal context footprint. We construct MCP-tools, a comprehensive dataset of 308 MCP servers and 2,797 tools from the official Model-Context-Protocol repository. Experiments demonstrate that MCP-Zero preserves agent autonomy while achieving substantial efficiency gains: (i) accurate tool selection from nearly 3k candidates across 248.1k tokens; (ii) 98\\% reduction in token consumption on APIBank while maintaining high accuracy; and (iii) consistent multi-turn performance that scales with tool ecosystem growth. This work establishes active tool discovery as a fundamental design pattern for scalable autonomous agent systems.", "authors": ["Xiang Fei", "Xiawu Zheng", "Hao Feng"], "categories": ["cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-01", "url": "https://arxiv.org/abs/2506.01056", "pdf_url": "https://arxiv.org/pdf/2506.01056v4", "arxiv_id": "2506.01056", "doi": "10.48550/arXiv.2506.01056", "citation_count": 26, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4515}
{"id": "ed46232afbf249a41dcf58f76fe99c601e65ac83b0735b625f36418118828bb7", "sources": ["arxiv", "semantic_scholar"], "title": "Improving LLM Agents with Reinforcement Learning on Cryptographic CTF Challenges", "abstract": "We present 'Random-Crypto', a procedurally generated cryptographic Capture The Flag (CTF) dataset designed to unlock the potential of Reinforcement Learning (RL) for LLM-based agents in security-sensitive domains. Cryptographic reasoning offers an ideal RL testbed: it combines precise validation, structured multi-step inference, and reliance on reliable computational tool use. Leveraging these properties, we fine-tune a Python tool-augmented Llama-3.1-8B via Group Relative Policy Optimization (GRPO) in a secure execution environment. The resulting agent achieves a significant improvement in Pass@8 on previously unseen challenges. Moreover, the improvements generalize to two external benchmarks: 'picoCTF', spanning both crypto and non-crypto tasks, and 'AICrypto MCQ', a multiple-choice benchmark of 135 cryptography questions. Ablation studies attribute the gains to enhanced tool usage and procedural reasoning. These findings position 'Random-Crypto' as a rich training ground for building intelligent, adaptable LLM agents capable of handling complex cybersecurity tasks.", "authors": ["Lajos Muzsai", "David Imolai", "András Lukács"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-06-01", "url": "https://arxiv.org/abs/2506.02048", "pdf_url": "https://arxiv.org/pdf/2506.02048v2", "arxiv_id": "2506.02048", "doi": "10.48550/arXiv.2506.02048", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1833}
{"id": "191e7d5267cdfc5063882b96f07449660c3a9c750bbcff99a119b1fd6f1531dd", "sources": ["arxiv", "semantic_scholar"], "title": "An Adversary-Resistant Multi-Agent LLM System via Credibility Scoring", "abstract": "While multi-agent LLM systems show strong capabilities in various domains, they are highly vulnerable to adversarial and low-performing agents. To resolve this issue, in this paper, we introduce a general and adversary-resistant multi-agent LLM framework based on credibility scoring. We model the collaborative query-answering process as an iterative game, where the agents communicate and contribute to a final system output. Our system associates a credibility score that is used when aggregating the team outputs. The credibility scores are learned gradually based on the past contributions of each agent in query answering. Our experiments across multiple tasks and settings demonstrate our system's effectiveness in mitigating adversarial influence and enhancing the resilience of multi-agent cooperation, even in the adversary-majority settings.", "authors": ["Sana Ebrahimi", "Mohsen Dehghankar", "Abolfazl Asudeh"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-30", "url": "https://arxiv.org/abs/2505.24239", "pdf_url": "https://arxiv.org/pdf/2505.24239v1", "arxiv_id": "2505.24239", "doi": "10.48550/arXiv.2505.24239", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "83134e44f0a6557c787d71d35ccb044fa15ad6046ddd6d0a59ddaac60d85ce79", "sources": ["arxiv", "semantic_scholar"], "title": "Multiple LLM Agents Debate for Equitable Cultural Alignment", "abstract": "Large Language Models (LLMs) need to adapt their predictions to diverse cultural contexts to benefit diverse communities across the world. While previous efforts have focused on single-LLM, single-turn approaches, we propose to exploit the complementary strengths of multiple LLMs to promote cultural adaptability. We introduce a Multi-Agent Debate framework, where two LLM-based agents debate over a cultural scenario and collaboratively reach a final decision. We propose two variants: one where either LLM agents exclusively debate and another where they dynamically choose between self-reflection and debate during their turns. We evaluate these approaches on 7 open-weight LLMs (and 21 LLM combinations) using the NormAd-ETI benchmark for social etiquette norms in 75 countries. Experiments show that debate improves both overall accuracy and cultural group parity over single-LLM baselines. Notably, multi-agent debate enables relatively small LLMs (7-9B) to achieve accuracies comparable to that of a much larger model (27B parameters).", "authors": ["Dayeon Ki", "Rachel Rudinger", "Tianyi Zhou", "Marine Carpuat"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-30", "url": "https://arxiv.org/abs/2505.24671", "pdf_url": "https://arxiv.org/pdf/2505.24671v2", "arxiv_id": "2505.24671", "doi": "10.48550/arXiv.2505.24671", "citation_count": 19, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3253}
{"id": "a8ff3e70855b45b58ee6ed5c1d4a997e86f3d6712ce1f9e5dafc14b0b21f78f3", "sources": ["arxiv", "semantic_scholar"], "title": "SentinelAgent: Graph-based Anomaly Detection in Multi-Agent Systems", "abstract": "The rise of large language model (LLM)-based multi-agent systems (MAS) introduces new security and reliability challenges. While these systems show great promise in decomposing and coordinating complex tasks, they also face multi-faceted risks across prompt manipulation, unsafe tool usage, and emergent agent miscoordination. Existing guardrail mechanisms offer only partial protection, primarily at the input-output level, and fall short in addressing systemic or multi-point failures in MAS. In this work, we present a system-level anomaly detection framework tailored for MAS, integrating structural modeling with runtime behavioral oversight. Our approach consists of two components. First, we propose a graph-based framework that models agent interactions as dynamic execution graphs, enabling semantic anomaly detection at node, edge, and path levels. Second, we introduce a pluggable SentinelAgent, an LLM-powered oversight agent that observes, analyzes, and intervenes in MAS execution based on security policies and contextual reasoning. By bridging abstract detection logic with actionable enforcement, our method detects not only single-point faults and prompt injections but also multi-agent collusion and latent exploit paths. We validate our framework through two case studies, including an email assistant and Microsoft's Magentic-One system, demonstrating its ability to detect covert risks and provide explainable root-cause attribution. Our work lays the foundation for more trustworthy, monitorable, and secure agent-based AI ecosystems.", "authors": ["Xu He", "Di Wu", "Yan Zhai", "Kun Sun"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-30", "url": "https://arxiv.org/abs/2505.24201", "pdf_url": "https://arxiv.org/pdf/2505.24201v1", "arxiv_id": "2505.24201", "doi": "10.48550/arXiv.2505.24201", "citation_count": 14, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "64020515aca296adc2348240c2383d968d51f1635202beefb49c065bc590a9c0", "sources": ["arxiv", "semantic_scholar"], "title": "NexusSum: Hierarchical LLM Agents for Long-Form Narrative Summarization", "abstract": "Summarizing long-form narratives--such as books, movies, and TV scripts--requires capturing intricate plotlines, character interactions, and thematic coherence, a task that remains challenging for existing LLMs. We introduce NexusSum, a multi-agent LLM framework for narrative summarization that processes long-form text through a structured, sequential pipeline--without requiring fine-tuning. Our approach introduces two key innovations: (1) Dialogue-to-Description Transformation: A narrative-specific preprocessing method that standardizes character dialogue and descriptive text into a unified format, improving coherence. (2) Hierarchical Multi-LLM Summarization: A structured summarization pipeline that optimizes chunk processing and controls output length for accurate, high-quality summaries. Our method establishes a new state-of-the-art in narrative summarization, achieving up to a 30.0% improvement in BERTScore (F1) across books, movies, and TV scripts. These results demonstrate the effectiveness of multi-agent LLMs in handling long-form content, offering a scalable approach for structured summarization in diverse storytelling domains.", "authors": ["Hyuntak Kim", "Byung-Hak Kim"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-30", "url": "https://arxiv.org/abs/2505.24575", "pdf_url": "https://arxiv.org/pdf/2505.24575v1", "arxiv_id": "2505.24575", "doi": "10.48550/arXiv.2505.24575", "citation_count": 10, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.2603}
{"id": "6a795843ec4fc3a033dd16d718eaa451199eab4fb92942c72d9838929ec684d3", "sources": ["arxiv"], "title": "Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve", "abstract": "Recent studies show that LLMs possess different skills and specialize in different tasks. In fact, we observe that their varied performance occur in several levels of granularity. For example, in the code optimization task, code LLMs excel at different optimization categories and no one dominates others. This observation prompts the question of how one leverages multiple LLM agents to solve a coding problem without knowing their complementary strengths a priori. We argue that a team of agents can learn from each other's successes and failures so as to improve their own performance. Thus, a lesson is the knowledge produced by an agent and passed on to other agents in the collective solution process. We propose a lesson-based collaboration framework, design the lesson solicitation--banking--selection mechanism, and demonstrate that a team of small LLMs with lessons learned can outperform a much larger LLM and other multi-LLM collaboration methods.", "authors": ["Yuanzhe Liu", "Ryan Deng", "Tim Kaler", "Xuhao Chen", "Charles E. Leiserson", "Yao Ma", "Jie Chen"], "categories": ["cs.AI", "cs.LG", "cs.MA", "cs.SE"], "fields_of_study": [], "published_date": "2025-05-29", "url": "https://arxiv.org/abs/2505.23946", "pdf_url": "https://arxiv.org/pdf/2505.23946v2", "arxiv_id": "2505.23946", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/MITIBM-FastCoder/LessonL", "venue": null, "quality_score": 0.2126}
{"id": "f91bd9f5cc944cd9589b3b340b573531660ab46f60e1c9d68a85e53377df361c", "sources": ["arxiv", "semantic_scholar"], "title": "Literature Review Of Multi-Agent Debate For Problem-Solving", "abstract": "Multi-agent large language models (MA-LLMs) are a rapidly growing research area that leverages multiple interacting language agents to tackle complex tasks, outperforming single-agent large language models. This literature review synthesizes the latest research on agent profiles, communication structures, and decision-making processes, drawing insights from both traditional multi-agent systems and state-of-the-art MA-LLM studies. In doing so, it aims to address the lack of direct comparisons in the field, illustrating how factors like scalability, communication structure, and decision-making processes influence MA-LLM performance. By examining frequent practices and outlining current challenges, the review reveals that multi-agent approaches can yield superior results but also face elevated computational costs and under-explored challenges unique to MA-LLM. Overall, these findings provide researchers and practitioners with a roadmap for developing robust and efficient multi-agent AI solutions.", "authors": ["Arne Tillmann"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-29", "url": "https://arxiv.org/abs/2506.00066", "pdf_url": "https://arxiv.org/pdf/2506.00066v1", "arxiv_id": "2506.00066", "doi": "10.48550/arXiv.2506.00066", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "d6d5db360d4ff9b17d45941cca02bbb9f22bfc1578be69e2e127378d9e038988", "sources": ["arxiv", "semantic_scholar"], "title": "Cross-Task Experiential Learning on LLM-based Multi-Agent Collaboration", "abstract": "Large Language Model-based multi-agent systems (MAS) have shown remarkable progress in solving complex tasks through collaborative reasoning and inter-agent critique. However, existing approaches typically treat each task in isolation, resulting in redundant computations and limited generalization across structurally similar tasks. To address this, we introduce multi-agent cross-task experiential learning (MAEL), a novel framework that endows LLM-driven agents with explicit cross-task learning and experience accumulation. We model the task-solving workflow on a graph-structured multi-agent collaboration network, where agents propagate information and coordinate via explicit connectivity. During the experiential learning phase, we quantify the quality for each step in the task-solving workflow and store the resulting rewards along with the corresponding inputs and outputs into each agent's individual experience pool. During inference, agents retrieve high-reward, task-relevant experiences as few-shot examples to enhance the effectiveness of each reasoning step, thereby enabling more accurate and efficient multi-agent collaboration. Experimental results on diverse datasets demonstrate that MAEL empowers agents to learn from prior task experiences effectively-achieving faster convergence and producing higher-quality solutions on current tasks.", "authors": ["Yilong Li", "Chen Qian", "Yu Xia", "Ruijie Shi", "Yufan Dang", "Zihao Xie", "Ziming You", "Weize Chen", "Cheng Yang", "Weichuan Liu", "Ye Tian", "Xuantang Xiong", "Lei Han", "Zhiyuan Liu", "Maosong Sun"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-29", "url": "https://arxiv.org/abs/2505.23187", "pdf_url": "https://arxiv.org/pdf/2505.23187v1", "arxiv_id": "2505.23187", "doi": "10.48550/arXiv.2505.23187", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1799}
{"id": "adb1e9488013f0d042ea8f33aee5aea2530916c32c615934ed957ee952800066", "sources": ["arxiv", "semantic_scholar"], "title": "ThinkGeo: Evaluating Tool-Augmented Agents for Remote Sensing Tasks", "abstract": "Recent progress in large language models (LLMs) has enabled tool-augmented agents capable of solving complex real-world tasks through step-by-step reasoning. However, existing evaluations often focus on general-purpose or multimodal scenarios, leaving a gap in domain-specific benchmarks that assess tool-use capabilities in complex remote sensing use cases. We present ThinkGeo, an agentic benchmark designed to evaluate LLM-driven agents on remote sensing tasks via structured tool use and multi-step planning. Inspired by tool-interaction paradigms, ThinkGeo includes human-curated queries spanning a wide range of real-world applications such as urban planning, disaster assessment and change analysis, environmental monitoring, transportation analysis, aviation monitoring, recreational infrastructure, and industrial site analysis. Queries are grounded in satellite or aerial imagery, including both optical RGB and SAR data, and require agents to reason through a diverse toolset. We implement a ReAct-style interaction loop and evaluate both open and closed-source LLMs (e.g., GPT-4o, Qwen2.5) on 486 structured agentic tasks with 1,778 expert-verified reasoning steps. The benchmark reports both step-wise execution metrics and final answer correctness. Our analysis reveals notable disparities in tool accuracy and planning consistency across models. ThinkGeo provides the first extensive testbed for evaluating how tool-enabled LLMs handle spatial reasoning in remote sensing.", "authors": ["Akashah Shabbir", "Muhammad Akhtar Munir", "Akshay Dudhane", "Muhammad Umer Sheikh", "Muhammad Haris Khan", "Paolo Fraccaro", "Juan Bernabe Moreno", "Fahad Shahbaz Khan", "Salman Khan"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-29", "url": "https://arxiv.org/abs/2505.23752", "pdf_url": "https://arxiv.org/pdf/2505.23752v3", "arxiv_id": "2505.23752", "doi": "10.48550/arXiv.2505.23752", "citation_count": 23, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3451}
{"id": "78e32791189e688aae31c8644ca7cbd426ffb851efaca6c3246f66d4604e6c4a", "sources": ["arxiv", "semantic_scholar"], "title": "Data-to-Dashboard: Multi-Agent LLM Framework for Insightful Visualization in Enterprise Analytics", "abstract": "The rapid advancement of LLMs has led to the creation of diverse agentic systems in data analysis, utilizing LLMs' capabilities to improve insight generation and visualization. In this paper, we present an agentic system that automates the data-to-dashboard pipeline through modular LLM agents capable of domain detection, concept extraction, multi-perspective analysis generation, and iterative self-reflection. Unlike existing chart QA systems, our framework simulates the analytical reasoning process of business analysts by retrieving domain-relevant knowledge and adapting to diverse datasets without relying on closed ontologies or question templates. We evaluate our system on three datasets across different domains. Benchmarked against GPT-4o with a single-prompt baseline, our approach shows improved insightfulness, domain relevance, and analytical depth, as measured by tailored evaluation metrics and qualitative human assessment. This work contributes a novel modular pipeline to bridge the path from raw data to visualization, and opens new opportunities for human-in-the-loop validation by domain experts in business analytics. All code can be found here: https://github.com/77luvC/D2D_Data2Dashboard", "authors": ["Ran Zhang", "Mohannad Elhamod"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-29", "url": "https://arxiv.org/abs/2505.23695", "pdf_url": "https://arxiv.org/pdf/2505.23695v1", "arxiv_id": "2505.23695", "doi": "10.48550/arXiv.2505.23695", "citation_count": 12, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/77luvC/D2D_Data2Dashboard", "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "276a6d5724360d6524b5e727db5f381d80180bd99a0c11edfcd3ff509000ed7c", "sources": ["arxiv", "semantic_scholar"], "title": "Efficient Leave-one-out Approximation in LLM Multi-agent Debate Based on Introspection", "abstract": "Multi-agent systems based on large language models (LLMs) advance automatic task completion in various fields, where debate is a common cooperation form for agents to solve complicated problems with reasoning and cross-review to solidify answers. Assessing the individual contributions of agents within these debates is crucial for system refinement and outcome reliability. Traditional leave-one-out (LOO) method offers a clear framework for evaluating each agent's role but face challenges in LLM-based systems due to high computational costs and associated financial implications. This paper presents introspective-leave-one-out (IntrospecLOO), a simple yet effective prompting for approximation of LOO in LLM-powered multi-agent debates. IntrospecLOO introduces an additional querying round after standard debates, prompting agents to update their answers while ignoring responses from a designated agent. This strategy effectively isolates and gauges each participant's influence at a reduced query complexity compared to the original LOO approaches. Validation through experiments on three benchmark datasets confirms the effectiveness of IntrospecLOO.", "authors": ["Yue Cui", "Liuyi Yao", "Zitao Li", "Yaliang Li", "Bolin Ding", "Xiaofang Zhou"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-28", "url": "https://arxiv.org/abs/2505.22192", "pdf_url": "https://arxiv.org/pdf/2505.22192v1", "arxiv_id": "2505.22192", "doi": "10.48550/arXiv.2505.22192", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1788}
{"id": "42765374cfaec64efce2fd1b85ac66f35c573b3b95fc5fa91856a174c93362fd", "sources": ["arxiv", "semantic_scholar"], "title": "Seven Security Challenges That Must be Solved in Cross-domain Multi-agent LLM Systems", "abstract": "Large language models (LLMs) are rapidly evolving into autonomous agents that cooperate across organizational boundaries, enabling joint disaster response, supply-chain optimization, and other tasks that demand decentralized expertise without surrendering data ownership. Yet, cross-domain collaboration shatters the unified trust assumptions behind current alignment and containment techniques. An agent benign in isolation may, when receiving messages from an untrusted peer, leak secrets or violate policy, producing risks driven by emergent multi-agent dynamics rather than classical software bugs. This position paper maps the security agenda for cross-domain multi-agent LLM systems. We introduce seven categories of novel security challenges, for each of which we also present plausible attacks, security evaluation metrics, and future research guidelines.", "authors": ["Ronny Ko", "Jiseong Jeong", "Shuyuan Zheng", "Chuan Xiao", "Tae-Wan Kim", "Makoto Onizuka", "Won-Yong Shin"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-28", "url": "https://arxiv.org/abs/2505.23847", "pdf_url": "https://arxiv.org/pdf/2505.23847v3", "arxiv_id": "2505.23847", "doi": "10.48550/arXiv.2505.23847", "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "588ba4144054afca414d4d62321b4688691daaad87a519d3874c0d5d9bf40ebd", "sources": ["arxiv", "semantic_scholar"], "title": "AgentDNS: A Root Domain Naming System for LLM Agents", "abstract": "The rapid evolution of Large Language Model (LLM) agents has highlighted critical challenges in cross-vendor service discovery, interoperability, and communication. Existing protocols like model context protocol and agent-to-agent protocol have made significant strides in standardizing interoperability between agents and tools, as well as communication among multi-agents. However, there remains a lack of standardized protocols and solutions for service discovery across different agent and tool vendors. In this paper, we propose AgentDNS, a root domain naming and service discovery system designed to enable LLM agents to autonomously discover, resolve, and securely invoke third-party agent and tool services across organizational and technological boundaries. Inspired by the principles of the traditional DNS, AgentDNS introduces a structured mechanism for service registration, semantic service discovery, secure invocation, and unified billing. We detail the architecture, core functionalities, and use cases of AgentDNS, demonstrating its potential to streamline multi-agent collaboration in real-world scenarios. The source code will be published on https://github.com/agentdns.", "authors": ["Enfang Cui", "Yujun Cheng", "Rui She", "Dan Liu", "Zhiyuan Liang", "Minxin Guo", "Tianzheng Li", "Qian Wei", "Wenjuan Xing", "Zhijie Zhong"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-28", "url": "https://arxiv.org/abs/2505.22368", "pdf_url": "https://arxiv.org/pdf/2505.22368v1", "arxiv_id": "2505.22368", "doi": "10.48550/arXiv.2505.22368", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/agentdns", "venue": "arXiv.org", "quality_score": 0.2762}
{"id": "64d5e547bfcf127c9b05bd3119d09b729c772d96858264bdf29759a5ee9a9c66", "sources": ["arxiv", "semantic_scholar"], "title": "HDDLGym: A Tool for Studying Multi-Agent Hierarchical Problems Defined in HDDL with OpenAI Gym", "abstract": "In recent years, reinforcement learning (RL) methods have been widely tested using tools like OpenAI Gym, though many tasks in these environments could also benefit from hierarchical planning. However, there is a lack of a tool that enables seamless integration of hierarchical planning with RL. Hierarchical Domain Definition Language (HDDL), used in classical planning, introduces a structured approach well-suited for model-based RL to address this gap. To bridge this integration, we introduce HDDLGym, a Python-based tool that automatically generates OpenAI Gym environments from HDDL domains and problems. HDDLGym serves as a link between RL and hierarchical planning, supporting multi-agent scenarios and enabling collaborative planning among agents. This paper provides an overview of HDDLGym's design and implementation, highlighting the challenges and design choices involved in integrating HDDL with the Gym interface, and applying RL policies to support hierarchical planning. We also provide detailed instructions and demonstrations for using the HDDLGym framework, including how to work with existing HDDL domains and problems from International Planning Competitions, exemplified by the Transport domain. Additionally, we offer guidance on creating new HDDL domains for multi-agent scenarios and demonstrate the practical use of HDDLGym in the Overcooked domain. By leveraging the advantages of HDDL and Gym, HDDLGym aims to be a valuable tool for studying RL in hierarchical planning, particularly in multi-agent contexts.", "authors": ["Ngoc La", "Ruaridh Mon-Williams", "Julie A. Shah"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-28", "url": "https://arxiv.org/abs/2505.22597", "pdf_url": "https://arxiv.org/pdf/2505.22597v1", "arxiv_id": "2505.22597", "doi": "10.48550/arXiv.2505.22597", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Proceedings of the  ... International Conference on Automated Planning and Scheduling", "quality_score": 0.1788}
{"id": "df819b017b12d5ec75d9c10a3ead584c656f946cab4adc3a892bd0c98756f44e", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-UniRAG: A Trainable Open-Source LLM Agent Framework for Unified Retrieval-Augmented Generation Systems", "abstract": "This paper presents a novel approach for unified retrieval-augmented generation (RAG) systems using the recent emerging large language model (LLM) agent concept. Specifically, Agent LLM, which utilizes LLM as fundamental controllers, has become a promising approach to enable the interpretability of RAG tasks, especially for complex reasoning question-answering systems (e.g., multi-hop queries). Nonetheless, previous works mainly focus on solving RAG systems with either single-hop or multi-hop approaches separately, which limits the application of those approaches to real-world applications. In this study, we propose a trainable agent framework called Agent-UniRAG for unified retrieval-augmented LLM systems, which enhances the effectiveness and interpretability of RAG systems. The main idea is to design an LLM agent framework to solve RAG tasks step-by-step based on the complexity of the inputs, simultaneously including single-hop and multi-hop queries in an end-to-end manner. Furthermore, we introduce SynAgent-RAG, a synthetic dataset to enable the proposed agent framework for small open-source LLMs (e.g., Llama-3-8B). The results show comparable performances with closed-source and larger open-source LLMs across various RAG benchmarks. Our source code and dataset are publicly available for further exploitation.", "authors": ["Hoang Pham", "Thuy-Duong Nguyen", "Khac-Hoai Nam Bui"], "categories": ["cs.CL", "cs.AI", "cs.DB", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-28", "url": "https://arxiv.org/abs/2505.22571", "pdf_url": "https://arxiv.org/pdf/2505.22571v3", "arxiv_id": "2505.22571", "doi": "10.48550/arXiv.2505.22571", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2762}
{"id": "a1ea5765ca21ee37cb08a15af8d9be91679eccf36a7ce6ebf62c0d4863ed3daf", "sources": ["arxiv", "semantic_scholar"], "title": "Silence is Not Consensus: Disrupting Agreement Bias in Multi-Agent LLMs via Catfish Agent for Clinical Decision Making", "abstract": "Large language models (LLMs) have demonstrated strong potential in clinical question answering, with recent multi-agent frameworks further improving diagnostic accuracy via collaborative reasoning. However, we identify a recurring issue of Silent Agreement, where agents prematurely converge on diagnoses without sufficient critical analysis, particularly in complex or ambiguous cases. We present a new concept called Catfish Agent, a role-specialized LLM designed to inject structured dissent and counter silent agreement. Inspired by the ``catfish effect'' in organizational psychology, the Catfish Agent is designed to challenge emerging consensus to stimulate deeper reasoning. We formulate two mechanisms to encourage effective and context-aware interventions: (i) a complexity-aware intervention that modulates agent engagement based on case difficulty, and (ii) a tone-calibrated intervention articulated to balance critique and collaboration. Evaluations on nine medical Q&A and three medical VQA benchmarks show that our approach consistently outperforms both single- and multi-agent LLMs frameworks, including leading commercial models such as GPT-4o and DeepSeek-R1.", "authors": ["Yihan Wang", "Qiao Yan", "Zhenghao Xing", "Lihao Liu", "Junjun He", "Chi-Wing Fu", "Xiaowei Hu", "Pheng-Ann Heng"], "categories": ["cs.CL", "cs.AI", "cs.LG", "q-bio.OT"], "fields_of_study": ["Computer Science", "Biology"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.21503", "pdf_url": "https://arxiv.org/pdf/2505.21503v1", "arxiv_id": "2505.21503", "doi": "10.48550/arXiv.2505.21503", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1776}
{"id": "cee1684b47e9a657791d9c259dff6f56b53b80bf43fd27279c5492f55eb0cef9", "sources": ["arxiv", "semantic_scholar"], "title": "MIRROR: Multi-agent Intra- and Inter-Reflection for Optimized Reasoning in Tool Learning", "abstract": "Complex tasks involving tool integration pose significant challenges for Large Language Models (LLMs), leading to the emergence of multi-agent workflows as a promising solution. Reflection has emerged as an effective strategy for correcting erroneous trajectories in agentic workflows. However, existing approaches only exploit such capability in the post-action stage, where the agent observes the execution outcomes. We argue that, like humans, LLMs can also engage in reflection before action execution: the agent can anticipate undesirable outcomes from its own decisions, which not only provides a necessarily complementary perspective to evaluate the decision but also prevents the propagation of errors throughout the trajectory. In this paper, we propose MIRROR, a framework that consists of both intra-reflection, which critically assesses intended actions before execution, and inter-reflection, which further adjusts the trajectory based on observations. This design systematically leverages LLM reflection capabilities to eliminate and rectify erroneous actions on a more comprehensive scope. Evaluations on both the StableToolBench and TravelPlanner benchmarks demonstrate MIRROR's superior performance, achieving state-of-the-art results compared to existing approaches.", "authors": ["Zikang Guo", "Benfeng Xu", "Xiaorui Wang", "Zhendong Mao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.20670", "pdf_url": "https://arxiv.org/pdf/2505.20670v2", "arxiv_id": "2505.20670", "doi": "10.48550/arXiv.2505.20670", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.2258}
{"id": "4400ff884cd4259ead97ababe7544a75d8dd53b5b397fa1f815ab517a6cf4b4b", "sources": ["arxiv", "semantic_scholar"], "title": "ChemAmp: Amplified Chemistry Tools via Composable Agents", "abstract": "Although LLM-based agents are proven to master tool orchestration in scientific fields, particularly chemistry, their single-task performance remains limited by underlying tool constraints. To this end, we propose tool amplification, a novel paradigm that enhances the collective capabilities of specialized tools through optimized, dynamic coordination within individual tasks. Instantiating this paradigm, we introduce ChemAmp, a computationally lightweight framework that dynamically treats chemistry tools (e.g., UniMol2, Chemformer) as composable building-block agents. It constructs task-specialized super-agents that transcend atomic tool constraints with limited data ($\\leq$10 samples). Our evaluations across four core chemistry tasks molecular design, molecule captioning, reaction prediction, and property prediction demonstrate that ChemAmp outperforms chemistry-specialized models, generalist LLMs, and agent systems with tool orchestration. Critically, this bottom-up construction strategy enables 94\\% inference token cost reductions versus vanilla multi-agent systems.", "authors": ["Zhucong Li", "Powei Chang", "Jin Xiao", "Zhijian Zhou", "Qianyu He", "Jiaqing Liang", "Fenglei Cao", "Xu Yinghui", "Yuan Qi"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.21569", "pdf_url": "https://arxiv.org/pdf/2505.21569v3", "arxiv_id": "2505.21569", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Chang-pw/ChemAmp", "venue": null, "quality_score": 0.2099}
{"id": "822de37d2fd6c067a022385b8862d962a34f4ac1a9ee2683952f0b3846def012", "sources": ["arxiv", "semantic_scholar"], "title": "MedSentry: Understanding and Mitigating Safety Risks in Medical LLM Multi-Agent Systems", "abstract": "As large language models (LLMs) are increasingly deployed in healthcare, ensuring their safety, particularly within collaborative multi-agent configurations, is paramount. In this paper we introduce MedSentry, a benchmark comprising 5 000 adversarial medical prompts spanning 25 threat categories with 100 subthemes. Coupled with this dataset, we develop an end-to-end attack-defense evaluation pipeline to systematically analyze how four representative multi-agent topologies (Layers, SharedPool, Centralized, and Decentralized) withstand attacks from 'dark-personality' agents. Our findings reveal critical differences in how these architectures handle information contamination and maintain robust decision-making, exposing their underlying vulnerability mechanisms. For instance, SharedPool's open information sharing makes it highly susceptible, whereas Decentralized architectures exhibit greater resilience thanks to inherent redundancy and isolation. To mitigate these risks, we propose a personality-scale detection and correction mechanism that identifies and rehabilitates malicious agents, restoring system safety to near-baseline levels. MedSentry thus furnishes both a rigorous evaluation framework and practical defense strategies that guide the design of safer LLM-based multi-agent systems in medical domains.", "authors": ["Kai Chen", "Taihang Zhen", "Hewei Wang", "Kailai Liu", "Xinfeng Li", "Jing Huo", "Tianpei Yang", "Jinfeng Xu", "Wei Dong", "Yang Gao"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.20824", "pdf_url": "https://arxiv.org/pdf/2505.20824v1", "arxiv_id": "2505.20824", "doi": "10.48550/arXiv.2505.20824", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2603}
{"id": "291da45acfadefcb492401b5b38c93cdd3c0ff5cdb93f08cf88b69d233696376", "sources": ["arxiv", "semantic_scholar"], "title": "Herd Behavior: Investigating Peer Influence in LLM-based Multi-Agent Systems", "abstract": "Recent advancements in Large Language Models (LLMs) have enabled the emergence of multi-agent systems where LLMs interact, collaborate, and make decisions in shared environments. While individual model behavior has been extensively studied, the dynamics of peer influence in such systems remain underexplored. In this paper, we investigate herd behavior, the tendency of agents to align their outputs with those of their peers, within LLM-based multi-agent interactions. We present a series of controlled experiments that reveal how herd behaviors are shaped by multiple factors. First, we show that the gap between self-confidence and perceived confidence in peers significantly impacts an agent's likelihood to conform. Second, we find that the format in which peer information is presented plays a critical role in modulating the strength of herd behavior. Finally, we demonstrate that the degree of herd behavior can be systematically controlled, and that appropriately calibrated herd tendencies can enhance collaborative outcomes. These findings offer new insights into the social dynamics of LLM-based systems and open pathways for designing more effective and adaptive multi-agent collaboration frameworks.", "authors": ["Young-Min Cho", "Sharath Chandra Guntuku", "Lyle Ungar"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.21588", "pdf_url": "https://arxiv.org/pdf/2505.21588v1", "arxiv_id": "2505.21588", "doi": "10.48550/arXiv.2505.21588", "citation_count": 10, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2603}
{"id": "3f225cb53d51c5c53c08dab25732debd5a0b4a580ec73c6269f84ad62b90809d", "sources": ["arxiv", "semantic_scholar"], "title": "Creativity in LLM-based Multi-Agent Systems: A Survey", "abstract": "Large language model (LLM)-driven multi-agent systems (MAS) are transforming how humans and AIs collaboratively generate ideas and artifacts. While existing surveys provide comprehensive overviews of MAS infrastructures, they largely overlook the dimension of \\emph{creativity}, including how novel outputs are generated and evaluated, how creativity informs agent personas, and how creative workflows are coordinated. This is the first survey dedicated to creativity in MAS. We focus on text and image generation tasks, and present: (1) a taxonomy of agent proactivity and persona design; (2) an overview of generation techniques, including divergent exploration, iterative refinement, and collaborative synthesis, as well as relevant datasets and evaluation metrics; and (3) a discussion of key challenges, such as inconsistent evaluation standards, insufficient bias mitigation, coordination conflicts, and the lack of unified benchmarks. This survey offers a structured framework and roadmap for advancing the development, evaluation, and standardization of creative MAS.", "authors": ["Yi-Cheng Lin", "Kang-Chieh Chen", "Zhe-Yan Li", "Tzu-Heng Wu", "Tzu-Hsuan Wu", "Kuan-Yu Chen", "Hung-yi Lee", "Yun-Nung Chen"], "categories": ["cs.HC", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.21116", "pdf_url": "https://arxiv.org/pdf/2505.21116v1", "arxiv_id": "2505.21116", "doi": "10.48550/arXiv.2505.21116", "citation_count": 23, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3451}
{"id": "5f5985212926e8a6fa4647a90cfc75568b22724dceb45cf850e10698460b048b", "sources": ["arxiv", "semantic_scholar"], "title": "Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration", "abstract": "With the rapid advancement of post-training techniques for reasoning and information seeking, large language models (LLMs) can incorporate a large quantity of retrieved knowledge to solve complex tasks. However, the limited context window of LLMs obstructs scaling the amount of external knowledge input, prohibiting further improvement. Existing context window extension methods inevitably cause information loss. LLM-based multi-agent methods emerge as a new paradigm to handle massive input in a distributional manner, where we identify two core bottlenecks in existing agent orchestration designs. In this work, we develop a multi-agent framework, \\textbf{\\ExtAgents}, to overcome the bottlenecks and enable better scalability in inference-time knowledge integration without longer-context training. Benchmarked with our enhanced multi-hop question answering test, \\textbf{$\\boldsymbol{\\infty}$Bench+}, and other public test sets including long survey generation, \\ExtAgents significantly enhances the performance over existing non-training methods with the same amount of external knowledge input, regardless of whether it falls \\emph{within or exceeds the context window}. Moreover, the method maintains efficiency due to high parallelism. We believe further study in the coordination of LLM agents on increasing external knowledge input could benefit real-world applications.", "authors": ["Zijun Liu", "Zhennan Wan", "Peng Li", "Ming Yan", "Fei Huang", "Yang Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.21471", "pdf_url": "https://arxiv.org/pdf/2505.21471v2", "arxiv_id": "2505.21471", "doi": "10.48550/arXiv.2505.21471", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/THUNLP-MT/ExtAgents", "venue": "arXiv.org", "quality_score": 0.2745}
{"id": "0809816b215f67600dddee55c7cf2cb20a9be899267173975a5942c51514a541", "sources": ["arxiv", "semantic_scholar"], "title": "Rethinking Information Synthesis in Multimodal Question Answering A Multi-Agent Perspective", "abstract": "Recent advances in multimodal question answering have primarily focused on combining heterogeneous modalities or fine-tuning multimodal large language models. While these approaches have shown strong performance, they often rely on a single, generalized reasoning strategy, overlooking the unique characteristics of each modality ultimately limiting both accuracy and interpretability. To address these limitations, we propose MAMMQA, a multi-agent QA framework for multimodal inputs spanning text, tables, and images. Our system includes two Visual Language Model (VLM) agents and one text-based Large Language Model (LLM) agent. The first VLM decomposes the user query into sub-questions and sequentially retrieves partial answers from each modality. The second VLM synthesizes and refines these results through cross-modal reasoning. Finally, the LLM integrates the insights into a cohesive answer. This modular design enhances interpretability by making the reasoning process transparent and allows each agent to operate within its domain of expertise. Experiments on diverse multimodal QA benchmarks demonstrate that our cooperative, multi-agent framework consistently outperforms existing baselines in both accuracy and robustness.", "authors": ["Krishna Singh Rajput", "Tejas Anvekar", "Chitta Baral", "Vivek Gupta"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-27", "url": "https://arxiv.org/abs/2505.20816", "pdf_url": "https://arxiv.org/pdf/2505.20816v2", "arxiv_id": "2505.20816", "doi": "10.48550/arXiv.2505.20816", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1505}
{"id": "f07df2aaf68358431bd1f155b72de326e26758988fa876ef6f481dc35d31be3a", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Agent-Controller: A Universal Multi-Agent Large Language Model System as a Control Engineer", "abstract": "This study presents the LLM-Agent-Controller, a multi-agent large language model (LLM) system developed to address a wide range of problems in control engineering (Control Theory). The system integrates a central controller agent with multiple specialized auxiliary agents, responsible for tasks such as controller design, model representation, control analysis, time-domain response, and simulation. A supervisor oversees high-level decision-making and workflow coordination, enhancing the system's reliability and efficiency. The LLM-Agent-Controller incorporates advanced capabilities, including Retrieval-Augmented Generation (RAG), Chain-of-Thought reasoning, self-criticism and correction, efficient memory handling, and user-friendly natural language communication. It is designed to function without requiring users to have prior knowledge of Control Theory, enabling them to input problems in plain language and receive complete, real-time solutions. To evaluate the system, we propose new performance metrics assessing both individual agents and the system as a whole. We test five categories of Control Theory problems and benchmark performance across three advanced LLMs. Additionally, we conduct a comprehensive qualitative conversational analysis covering all key services. Results show that the LLM-Agent-Controller successfully solved 83% of general tasks, with individual agents achieving an average success rate of 87%. Performance improved with more advanced LLMs. This research demonstrates the potential of multi-agent LLM architectures to solve complex, domain-specific problems. By integrating specialized agents, supervisory control, and advanced reasoning, the LLM-Agent-Controller offers a scalable, robust, and accessible solution framework that can be extended to various technical domains.", "authors": ["Rasoul Zahedifar", "Sayyed Ali Mirghasemi", "Mahdieh Soleymani Baghshah", "Alireza Taheri"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.19567", "pdf_url": "https://arxiv.org/pdf/2505.19567v1", "arxiv_id": "2505.19567", "doi": "10.48550/arXiv.2505.19567", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "cb27832c14b39a4477617efffdae6fc70abccd11b4afcf8cc0bb7ae7bc3230cc", "sources": ["arxiv", "semantic_scholar"], "title": "MultiPhishGuard: An Explainable and Adaptive Multi-Agent LLM System for Phishing Email Detection", "abstract": "Phishing email detection faces significant challenges due to evolving adversarial tactics and heterogeneous attack patterns. Traditional approaches, such as rule-based filters and denylists, often struggle to keep pace, leading to missed detections and security risks. While machine learning methods have improved detection performance, they remain limited in adapting to novel and rapidly changing phishing strategies. We present MultiPhishGuard, an LLM-based multi-agent detection framework with learned coordination across specialized agents. The system consists of five cooperative agents (text, URL, metadata, explanation simplifier, and adversarial agents), with agent contributions dynamically weighted using Proximal Policy Optimization. To address emerging threats, the framework incorporates an adversarial training loop in which an LLM-based agent generates subtle, context-aware email variants to expose potential model weaknesses and improve robustness to ambiguous phishing cases. Experimental evaluations on public datasets show that MultiPhishGuard achieves stronger performance than established baselines, including Chain-of-Thought prompting and single-agent variants, as supported by ablation studies and comparative analyses. The system achieves an accuracy of 97.89%, with a false positive rate of 2.73% and a false negative rate of 0.20%. In addition, an explanation simplifier agent transforms technical model outputs into plain-language rationales intended for human users. Overall, these results suggest that multi-agent LLM architectures with adaptive coordination and adversarial training represent a promising direction for phishing email detection.", "authors": ["Yinuo Xue", "Eric Spero", "Meng Wai Woo", "Wei Gao", "Giovanni Russello"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.23803", "pdf_url": "https://arxiv.org/pdf/2505.23803v2", "arxiv_id": "2505.23803", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1123}
{"id": "0069d2026a95833d750dc01a8bb0c84aaf95fafce9e17e061a9efb643fe1f92f", "sources": ["arxiv", "semantic_scholar"], "title": "CoTGuard: Using Chain-of-Thought Triggering for Copyright Protection in Multi-Agent LLM Systems", "abstract": "As large language models (LLMs) evolve into autonomous agents capable of collaborative reasoning and task execution, multi-agent LLM systems have emerged as a powerful paradigm for solving complex problems. However, these systems pose new challenges for copyright protection, particularly when sensitive or copyrighted content is inadvertently recalled through inter-agent communication and reasoning. Existing protection techniques primarily focus on detecting content in final outputs, overlooking the richer, more revealing reasoning processes within the agents themselves. In this paper, we introduce CoTGuard, a novel framework for copyright protection that leverages trigger-based detection within Chain-of-Thought (CoT) reasoning. Specifically, we can activate specific CoT segments and monitor intermediate reasoning steps for unauthorized content reproduction by embedding specific trigger queries into agent prompts. This approach enables fine-grained, interpretable detection of copyright violations in collaborative agent scenarios. We evaluate CoTGuard on various benchmarks in extensive experiments and show that it effectively uncovers content leakage with minimal interference to task performance. Our findings suggest that reasoning-level monitoring offers a promising direction for safeguarding intellectual property in LLM-based agent systems.", "authors": ["Yan Wen", "Junfeng Guo", "Heng Huang"], "categories": ["cs.CL", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.19405", "pdf_url": "https://arxiv.org/pdf/2505.19405v1", "arxiv_id": "2505.19405", "doi": "10.48550/arXiv.2505.19405", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1765}
{"id": "e59d13704a19762add5efaae7e348bbbdac0d6f50462a26e9267bee833523e46", "sources": ["arxiv", "semantic_scholar"], "title": "Task Memory Engine: Spatial Memory for Robust Multi-Step LLM Agents", "abstract": "Large Language Models (LLMs) falter in multi-step interactions -- often hallucinating, repeating actions, or misinterpreting user corrections -- due to reliance on linear, unstructured context. This fragility stems from the lack of persistent memory to track evolving goals and task dependencies, undermining trust in autonomous agents. We introduce the Task Memory Engine (TME), a modular memory controller that transforms existing LLMs into robust, revision-aware agents without fine-tuning. TME implements a spatial memory framework that replaces flat context with graph-based structures to support consistent, multi-turn reasoning. Departing from linear concatenation and ReAct-style prompting, TME builds a dynamic task graph -- either a tree or directed acyclic graph (DAG) -- to map user inputs to subtasks, align them with prior context, and enable dependency-tracked revisions. Its Task Representation and Intent Management (TRIM) component models task semantics and user intent to ensure accurate interpretation. Across four multi-turn scenarios-trip planning, cooking, meeting scheduling, and shopping cart editing -- TME eliminates 100% of hallucinations and misinterpretations in three tasks, and reduces hallucinations by 66.7% and misinterpretations by 83.3% across 27 user turns, outperforming ReAct. TME's modular design supports plug-and-play deployment and domain-specific customization, adaptable to both personal assistants and enterprise automation. We release TME's codebase, benchmarks, and components as open-source resources, enabling researchers to develop reliable LLM agents. TME's scalable architecture addresses a critical gap in agent performance across complex, interactive settings.", "authors": ["Ye Ye"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.19436", "pdf_url": "https://arxiv.org/pdf/2505.19436v1", "arxiv_id": "2505.19436", "doi": "10.48550/arXiv.2505.19436", "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2727}
{"id": "8b0da50d5396ecedc694eb5fbf4c571d23a284971c7d365a77db83e4903f2c70", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-View Encoders for Performance Prediction in LLM-Based Agentic Workflows", "abstract": "Large language models (LLMs) have demonstrated remarkable capabilities across diverse tasks, but optimizing LLM-based agentic systems remains challenging due to the vast search space of agent configurations, prompting strategies, and communication patterns. Existing approaches often rely on heuristic-based tuning or exhaustive evaluation, which can be computationally expensive and suboptimal. This paper proposes Agentic Predictor, a lightweight predictor for efficient agentic workflow evaluation. Agentic Predictor is equipped with a multi-view workflow encoding technique that leverages multi-view representation learning of agentic systems by incorporating code architecture, textual prompts, and interaction graph features. To achieve high predictive accuracy while significantly reducing the number of required workflow evaluations for training a predictor, Agentic Predictor employs cross-domain unsupervised pretraining. By learning to approximate task success rates, Agentic Predictor enables fast and accurate selection of optimal agentic workflow configurations for a given task, significantly reducing the need for expensive trial-and-error evaluations. Experiments on a carefully curated benchmark spanning three domains show that our predictor outperforms several strong graph-based baselines in both predictive accuracy and workflow utility, highlighting the potential of performance predictors in streamlining the design of LLM-based agentic workflows.", "authors": ["Patara Trirat", "Wonyong Jeong", "Sung Ju Hwang"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.19764", "pdf_url": "https://arxiv.org/pdf/2505.19764v2", "arxiv_id": "2505.19764", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1505}
{"id": "f53557a8c54aed37c3edd07b7b06ecb3885b36cb746f40e9792fec94875ce16b", "sources": ["arxiv", "semantic_scholar"], "title": "Judging with Many Minds: Do More Perspectives Mean Less Prejudice? On Bias Amplifications and Resistance in Multi-Agent Based LLM-as-Judge", "abstract": "LLM-as-Judge has emerged as a scalable alternative to human evaluation, enabling large language models (LLMs) to provide reward signals in trainings. While recent work has explored multi-agent extensions such as multi-agent debate and meta-judging to enhance evaluation quality, the question of how intrinsic biases manifest in these settings remains underexplored. In this study, we conduct a systematic analysis of four diverse bias types: position bias, verbosity bias, chain-of-thought bias, and bandwagon bias. We evaluate these biases across two widely adopted multi-agent LLM-as-Judge frameworks: Multi-Agent-Debate and LLM-as-Meta-Judge. Our results show that debate framework amplifies biases sharply after the initial debate, and this increased bias is sustained in subsequent rounds, while meta-judge approaches exhibit greater resistance. We further investigate the incorporation of PINE, a leading single-agent debiasing method, as a bias-free agent within these systems. The results reveal that this bias-free agent effectively reduces biases in debate settings but provides less benefit in meta-judge scenarios. Our work provides a comprehensive study of bias behavior in multi-agent LLM-as-Judge systems and highlights the need for targeted bias mitigation strategies in collaborative evaluation settings.", "authors": ["Chiyu Ma", "Enpei Zhang", "Yilun Zhao", "Wenjun Liu", "Yaning Jia", "Peijun Qing", "Lin Shi", "Arman Cohan", "Yujun Yan", "Soroush Vosoughi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.19477", "pdf_url": "https://arxiv.org/pdf/2505.19477v3", "arxiv_id": "2505.19477", "doi": "10.48550/arXiv.2505.19477", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.2386}
{"id": "e513efcbd5bb8df5f3534e4488b14fd133c06750f564c4a78bb4b85a0eae6f4f", "sources": ["arxiv", "semantic_scholar"], "title": "Survey of LLM Agent Communication with MCP: A Software Design Pattern Centric Review", "abstract": "This survey investigates how classical software design patterns can enhance the reliability and scalability of communication in Large Language Model (LLM)-driven agentic AI systems, focusing particularly on the Model Context Protocol (MCP). It examines the foundational architectures of LLM-based agents and their evolution from isolated operation to sophisticated, multi-agent collaboration, addressing key communication hurdles that arise in this transition. The study revisits well-established patterns, including Mediator, Observer, Publish-Subscribe, and Broker, and analyzes their relevance in structuring agent interactions within MCP-compliant frameworks. To clarify these dynamics, the article provides conceptual schematics and formal models that map out communication pathways and optimize data flow. It further explores architectural variations suited to different degrees of agent autonomy and system complexity. Real-world applications in domains such as real-time financial processing and investment banking are discussed, illustrating how these patterns and MCP can meet specific operational demands. The article concludes by outlining open challenges, potential security risks, and promising directions for advancing robust, interoperable, and scalable multi-agent LLM ecosystems.", "authors": ["Anjana Sarkar", "Soumyendu Sarkar"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2506.05364", "pdf_url": "https://arxiv.org/pdf/2506.05364v2", "arxiv_id": "2506.05364", "doi": "10.48550/arXiv.2506.05364", "citation_count": 18, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "bf21c220f9a340ea6b6532ae5f04269f77f588aa3f311ba22e75f343503c3d86", "sources": ["arxiv", "semantic_scholar"], "title": "Project Riley: Multimodal Multi-Agent LLM Collaboration with Emotional Reasoning and Voting", "abstract": "This paper presents Project Riley, a novel multimodal and multi-model conversational AI architecture oriented towards the simulation of reasoning influenced by emotional states. Drawing inspiration from Pixar's Inside Out, the system comprises five distinct emotional agents - Joy, Sadness, Fear, Anger, and Disgust - that engage in structured multi-round dialogues to generate, criticise, and iteratively refine responses. A final reasoning mechanism synthesises the contributions of these agents into a coherent output that either reflects the dominant emotion or integrates multiple perspectives. The architecture incorporates both textual and visual large language models (LLMs), alongside advanced reasoning and self-refinement processes. A functional prototype was deployed locally in an offline environment, optimised for emotional expressiveness and computational efficiency. From this initial prototype, another one emerged, called Armando, which was developed for use in emergency contexts, delivering emotionally calibrated and factually accurate information through the integration of Retrieval-Augmented Generation (RAG) and cumulative context tracking. The Project Riley prototype was evaluated through user testing, in which participants interacted with the chatbot and completed a structured questionnaire assessing three dimensions: Emotional Appropriateness, Clarity and Utility, and Naturalness and Human-likeness. The results indicate strong performance in structured scenarios, particularly with respect to emotional alignment and communicative clarity.", "authors": ["Ana Rita Ortigoso", "Gabriel Vieira", "Daniel Fuentes", "Luis Frazão", "Nuno Costa", "António Pereira"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-26", "url": "https://arxiv.org/abs/2505.20521", "pdf_url": "https://arxiv.org/pdf/2505.20521v2", "arxiv_id": "2505.20521", "doi": "10.48550/arXiv.2505.20521", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1765}
{"id": "ee621422ac2a489f3f5b5ede621b1ac9e33cee64a3688d18b1667239369b95f7", "sources": ["arxiv", "semantic_scholar"], "title": "GUARDIAN: Safeguarding LLM Multi-Agent Collaborations with Temporal Graph Modeling", "abstract": "The emergence of large language models (LLMs) enables the development of intelligent agents capable of engaging in complex and multi-turn dialogues. However, multi-agent collaboration faces critical safety challenges, such as hallucination amplification and error injection and propagation. This paper presents GUARDIAN, a unified method for detecting and mitigating multiple safety concerns in GUARDing Intelligent Agent collaboratioNs. By modeling the multi-agent collaboration process as a discrete-time temporal attributed graph, GUARDIAN explicitly captures the propagation dynamics of hallucinations and errors. The unsupervised encoder-decoder architecture incorporating an incremental training paradigm learns to reconstruct node attributes and graph structures from latent embeddings, enabling the identification of anomalous nodes and edges with unparalleled precision. Moreover, we introduce a graph abstraction mechanism based on the Information Bottleneck Theory, which compresses temporal interaction graphs while preserving essential patterns. Extensive experiments demonstrate GUARDIAN's effectiveness in safeguarding LLM multi-agent collaborations against diverse safety vulnerabilities, achieving state-of-the-art accuracy with efficient resource utilization. The code is available at https://github.com/JialongZhou666/GUARDIAN", "authors": ["Jialong Zhou", "Lichao Wang", "Xiao Yang"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-25", "url": "https://arxiv.org/abs/2505.19234", "pdf_url": "https://arxiv.org/pdf/2505.19234v2", "arxiv_id": "2505.19234", "doi": "10.48550/arXiv.2505.19234", "citation_count": 21, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/JialongZhou666/GUARDIAN", "venue": "arXiv.org", "quality_score": 0.3356}
{"id": "56268a8b2bb2defcd8a6f5a472dc49a1b25baf01282358ad82f808106bcf39fa", "sources": ["arxiv", "semantic_scholar"], "title": "A Survey of LLM $\\times$ DATA", "abstract": "The integration of large language model (LLM) and data management (DATA) is rapidly redefining both domains. In this survey, we comprehensively review the bidirectional relationships. On the one hand, DATA4LLM, spanning large-scale data processing, storage, and serving, feeds LLMs with high quality, diversity, and timeliness of data required for stages like pre-training, post-training, retrieval-augmented generation, and agentic workflows: (i) Data processing for LLMs includes scalable acquisition, deduplication, filtering, selection, domain mixing, and synthetic augmentation; (ii) Data Storage for LLMs focuses on efficient data and model formats, distributed and heterogeneous storage hierarchies, KV-cache management, and fault-tolerant checkpointing; (iii) Data serving for LLMs tackles challenges in RAG (e.g., knowledge post-processing), LLM inference (e.g., prompt compression, data provenance), and training strategies (e.g., data packing and shuffling). On the other hand, in LLM4DATA, LLMs are emerging as general-purpose engines for data management. We review recent advances in (i) data manipulation, including automatic data cleaning, integration, discovery; (ii) data analysis, covering reasoning over structured, semi-structured, and unstructured data, and (iii) system optimization (e.g., configuration tuning, query rewriting, anomaly diagnosis), powered by LLM techniques like retrieval-augmented prompting, task-specialized fine-tuning, and multi-agent collaboration.", "authors": ["Xuanhe Zhou", "Junxuan He", "Wei Zhou", "Haodong Chen", "Zirui Tang", "Haoyu Zhao", "Xin Tong", "Guoliang Li", "Youmin Chen", "Jun Zhou", "Zhaojun Sun", "Binyuan Hui", "Shuo Wang", "Conghui He", "Zhiyuan Liu", "Jingren Zhou", "Fan Wu"], "categories": ["cs.DB", "cs.AI", "cs.CL", "cs.IR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-24", "url": "https://arxiv.org/abs/2505.18458", "pdf_url": "https://arxiv.org/pdf/2505.18458v3", "arxiv_id": "2505.18458", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/weAIDB/awesome-data-llm", "venue": null, "quality_score": 0.2058}
{"id": "5d8d792e3870f7a032b31140b799e62d128dd8c64d827ce972114a1b533f3588", "sources": ["arxiv", "semantic_scholar"], "title": "Implementing Agents in JavaScript", "abstract": "This chapter gives an introduction to agent-oriented programming in JavaScript. It provides an example-based walk-through of how to implement abstractions for reasoning loop agents in vanilla JavaScript. The initial example is used as a stepping stone for explaining how to implement slightly more advanced agents and multi-agent systems using JS-son, a JavaScript library for agent-oriented programming. In this context, the chapter also explains how to integrate reasoning loop agents with generative AI technologies--specifically, large language models. Finally, application scenarios in several technology ecosystems and future research directions are sketched.", "authors": ["Timotheus Kampik"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18228", "pdf_url": "https://arxiv.org/pdf/2505.18228v2", "arxiv_id": "2505.18228", "doi": "10.48550/arXiv.2505.18228", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.173}
{"id": "1b5664058c2251c33245e99e207cd48b20f2181a059677831789915e53c5e4f9", "sources": ["arxiv", "semantic_scholar"], "title": "ManuSearch: Democratizing Deep Search in Large Language Models with a Transparent and Open Multi-Agent Framework", "abstract": "Recent advances in web-augmented large language models (LLMs) have exhibited strong performance in complex reasoning tasks, yet these capabilities are mostly locked in proprietary systems with opaque architectures. In this work, we propose \\textbf{ManuSearch}, a transparent and modular multi-agent framework designed to democratize deep search for LLMs. ManuSearch decomposes the search and reasoning process into three collaborative agents: (1) a solution planning agent that iteratively formulates sub-queries, (2) an Internet search agent that retrieves relevant documents via real-time web search, and (3) a structured webpage reading agent that extracts key evidence from raw web content. To rigorously evaluate deep reasoning abilities, we introduce \\textbf{ORION}, a challenging benchmark focused on open-web reasoning over long-tail entities, covering both English and Chinese. Experimental results show that ManuSearch substantially outperforms prior open-source baselines and even surpasses leading closed-source systems. Our work paves the way for reproducible, extensible research in open deep search systems. We release the data and code in https://github.com/RUCAIBox/ManuSearch", "authors": ["Lisheng Huang", "Yichen Liu", "Jinhao Jiang", "Rongxiang Zhang", "Jiahao Yan", "Junyi Li", "Wayne Xin Zhao"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18105", "pdf_url": "https://arxiv.org/pdf/2505.18105v1", "arxiv_id": "2505.18105", "doi": "10.48550/arXiv.2505.18105", "citation_count": 14, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/RUCAIBox/ManuSearch", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.294}
{"id": "67a56b9b4977e6f1778374e6b2b6e00da7b956a1247c21169bde5f61a22cfb99", "sources": ["arxiv", "semantic_scholar"], "title": "Single-agent or Multi-agent Systems? Why Not Both?", "abstract": "Multi-agent systems (MAS) decompose complex tasks and delegate subtasks to different large language model (LLM) agents and tools. Prior studies have reported the superior accuracy performance of MAS across diverse domains, enabled by long-horizon context tracking and error correction through role-specific agents. However, the design and deployment of MAS incur higher complexity and runtime cost compared to single-agent systems (SAS). Meanwhile, frontier LLMs, such as OpenAI-o3 and Gemini-2.5-Pro, have rapidly advanced in long-context reasoning, memory retention, and tool usage, mitigating many limitations that originally motivated MAS designs. In this paper, we conduct an extensive empirical study comparing MAS and SAS across various popular agentic applications. We find that the benefits of MAS over SAS diminish as LLM capabilities improve, and we propose efficient mechanisms to pinpoint the error-prone agent in MAS. Furthermore, the performance discrepancy between MAS and SAS motivates our design of a hybrid agentic paradigm, request cascading between MAS and SAS, to improve both efficiency and capability. Our design improves accuracy by 1.1-12% while reducing deployment costs by up to 20% across various agentic applications.", "authors": ["Mingyan Gao", "Yanzi Li", "Banruo Liu", "Yifan Yu", "Phillip Wang", "Ching-Yu Lin", "Fan Lai"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18286", "pdf_url": "https://arxiv.org/pdf/2505.18286v1", "arxiv_id": "2505.18286", "doi": "10.48550/arXiv.2505.18286", "citation_count": 27, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3618}
{"id": "b7d03721a31695c2e2388b90b3fd2d55b478b2d882e12f15f1aaccb71d1e177c", "sources": ["arxiv", "semantic_scholar"], "title": "Is Your LLM Really Mastering the Concept? A Multi-Agent Benchmark", "abstract": "Concepts serve as fundamental abstractions that support human reasoning and categorization. However, it remains unclear whether large language models truly capture such conceptual structures or primarily rely on surface-level pattern memorization. Existing benchmarks are largely static and fact oriented, which limits their ability to probe fine-grained semantic understanding and makes them vulnerable to data leakage and overfitting. To address this limitation, we introduce CK-Arena, a dynamic benchmark for conceptual knowledge evaluation based on a multi agent social deduction game, namely the Undercover game. In this setting, LLM based agents are assigned subtly different concept words and must describe, distinguish, and infer conceptual properties from others' statements. Model performance is evaluated through both game level outcomes and the semantic quality of generated descriptions. Furthermore, CK-Arena leverages the interaction process to automatically construct high quality question answering data for fine grained diagnostic analysis. Experimental results show that conceptual understanding varies substantially across models and categories, and is not strictly aligned with overall model capability. The data and code are available at the project homepage: https://ck-arena.site.", "authors": ["Shuhang Xu", "Weijian Deng", "Yixuan Zhou", "Fangwei Zhong"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.17512", "pdf_url": "https://arxiv.org/pdf/2505.17512v2", "arxiv_id": "2505.17512", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1101}
{"id": "70e53ed9dc312da867cf5d3220f353da3589e0dceccabbd2a2547c93f2520310", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Memory: Multi-User Memory Sharing in LLM Agents with Dynamic Access Control", "abstract": "Complex tasks are increasingly delegated to ensembles of specialized LLM-based agents that reason, communicate, and coordinate actions-both among themselves and through interactions with external tools, APIs, and databases. While persistent memory has been shown to enhance single-agent performance, most approaches assume a monolithic, single-user context-overlooking the benefits and challenges of knowledge transfer across users under dynamic, asymmetric permissions. We introduce Collaborative Memory, a framework for multi-user, multi-agent environments with asymmetric, time-evolving access controls encoded as bipartite graphs linking users, agents, and resources. Our system maintains two memory tiers: (1) private memory-private fragments visible only to their originating user; and (2) shared memory-selectively shared fragments. Each fragment carries immutable provenance attributes (contributing agents, accessed resources, and timestamps) to support retrospective permission checks. Granular read policies enforce current user-agent-resource constraints and project existing memory fragments into filtered transformed views. Write policies determine fragment retention and sharing, applying context-aware transformations to update the memory. Both policies may be designed conditioned on system, agent, and user-level information. Our framework enables safe, efficient, and interpretable cross-user knowledge sharing, with provable adherence to asymmetric, time-varying policies and full auditability of memory operations.", "authors": ["Alireza Rezazadeh", "Zichao Li", "Ange Lou", "Yuying Zhao", "Wei Wei", "Yujia Bao"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18279", "pdf_url": "https://arxiv.org/pdf/2505.18279v1", "arxiv_id": "2505.18279", "doi": "10.48550/arXiv.2505.18279", "citation_count": 28, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3656}
{"id": "0bff3ce194619ac531497e25619e51528bf815deffc0f59c5a8960bed2200727", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Systems for Misinformation Lifecycle : Detection, Correction And Source Identification", "abstract": "The rapid proliferation of misinformation in digital media demands solutions that go beyond isolated Large Language Model(LLM) or AI Agent based detection methods. This paper introduces a novel multi-agent framework that covers the complete misinformation lifecycle: classification, detection, correction, and source verification to deliver more transparent and reliable outcomes. In contrast to single-agent or monolithic architectures, our approach employs five specialized agents: an Indexer agent for dynamically maintaining trusted repositories, a Classifier agent for labeling misinformation types, an Extractor agent for evidence based retrieval and ranking, a Corrector agent for generating fact-based correction and a Verification agent for validating outputs and tracking source credibility. Each agent can be individually evaluated and optimized, ensuring scalability and adaptability as new types of misinformation and data sources emerge. By decomposing the misinformation lifecycle into specialized agents - our framework enhances scalability, modularity, and explainability. This paper proposes a high-level system overview, agent design with emphasis on transparency, evidence-based outputs, and source provenance to support robust misinformation detection and correction at scale.", "authors": ["Aditya Gautam"], "categories": ["cs.MA", "cs.AI", "cs.ET", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.17511", "pdf_url": "https://arxiv.org/pdf/2505.17511v1", "arxiv_id": "2505.17511", "doi": "10.48550/arXiv.2505.17511", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.173}
{"id": "bc2aba1168226c1b7f229a912aa6edc5162a236f2fe1fc177881632233cb9f30", "sources": ["arxiv", "semantic_scholar"], "title": "Get Experience from Practice: LLM Agents with Record & Replay", "abstract": "AI agents, empowered by Large Language Models (LLMs) and communication protocols such as MCP and A2A, have rapidly evolved from simple chatbots to autonomous entities capable of executing complex, multi-step tasks, demonstrating great potential. However, the LLMs' inherent uncertainty and heavy computational resource requirements pose four significant challenges to the development of safe and efficient agents: reliability, privacy, cost and performance. Existing approaches, like model alignment, workflow constraints and on-device model deployment, can partially alleviate some issues but often with limitations, failing to fundamentally resolve these challenges. This paper proposes a new paradigm called AgentRR (Agent Record & Replay), which introduces the classical record-and-replay mechanism into AI agent frameworks. The core idea is to: 1. Record an agent's interaction trace with its environment and internal decision process during task execution, 2. Summarize this trace into a structured \"experience\" encapsulating the workflow and constraints, and 3. Replay these experiences in subsequent similar tasks to guide the agent's behavior. We detail a multi-level experience abstraction method and a check function mechanism in AgentRR: the former balances experience specificity and generality, while the latter serves as a trust anchor to ensure completeness and safety during replay. In addition, we explore multiple application modes of AgentRR, including user-recorded task demonstration, large-small model collaboration and privacy-aware agent execution, and envision an experience repository for sharing and reusing knowledge to further reduce deployment cost.", "authors": ["Erhu Feng", "Wenbo Zhou", "Zibin Liu", "Le Chen", "Yunpeng Dong", "Cheng Zhang", "Yisheng Zhao", "Dong Du", "Zhichao Hua", "Yubin Xia", "Haibo Chen"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.17716", "pdf_url": "https://arxiv.org/pdf/2505.17716v1", "arxiv_id": "2505.17716", "doi": "10.48550/arXiv.2505.17716", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "3e3466fa24e6c3bfd419c02bd11ab8b6d2993e9493d4b31a16549a611fdd80dd", "sources": ["arxiv", "semantic_scholar"], "title": "Deep Video Discovery: Agentic Search with Tool Use for Long-form Video Understanding", "abstract": "Long-form video understanding presents significant challenges due to extensive temporal-spatial complexity and the difficulty of question answering under such extended contexts. While Large Language Models (LLMs) have demonstrated considerable advancements in video analysis capabilities and long context handling, they continue to exhibit limitations when processing information-dense hour-long videos. To overcome such limitations, we propose the Deep Video Discovery (DVD) agent to leverage an agentic search strategy over segmented video clips. Unlike previous video agents that rely on predefined workflows applied uniformly across different queries, our approach emphasizes the autonomous and adaptive nature of agents. By providing a set of search-centric tools on multi-granular video database, our DVD agent leverages the advanced reasoning capability of LLM to plan on its current observation state, strategically selects tools to orchestrate adaptive workflow for different queries in light of the gathered information. We perform comprehensive evaluation on multiple long video understanding benchmarks that demonstrates our advantage. Our DVD agent achieves state-of-the-art performance on the challenging LVBench dataset, reaching an accuracy of 74.2%, which substantially surpasses all prior works, and further improves to 76.0% with transcripts. The code has been released at https://github.com/microsoft/DeepVideoDiscovery.", "authors": ["Xiaoyi Zhang", "Zhaoyang Jia", "Zongyu Guo", "Jiahao Li", "Bin Li", "Houqiang Li", "Yan Lu"], "categories": ["cs.CV", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18079", "pdf_url": "https://arxiv.org/pdf/2505.18079v4", "arxiv_id": "2505.18079", "doi": "10.48550/arXiv.2505.18079", "citation_count": 48, "influential_citation_count": 10, "has_code": true, "code_url": "https://github.com/microsoft/DeepVideoDiscovery", "venue": "arXiv.org", "quality_score": 0.5207}
{"id": "2083f8228608193659d425dd5ac0e8f04a76e0f2cec5163f567fc176c4f97c2e", "sources": ["arxiv", "semantic_scholar"], "title": "Tool Preferences in Agentic LLMs are Unreliable", "abstract": "Large language models (LLMs) can now access a wide range of external tools, thanks to the Model Context Protocol (MCP). This greatly expands their abilities as various agents. However, LLMs rely entirely on the text descriptions of tools to decide which ones to use--a process that is surprisingly fragile. In this work, we expose a vulnerability in prevalent tool/function-calling protocols by investigating a series of edits to tool descriptions, some of which can drastically increase a tool's usage from LLMs when competing with alternatives. Through controlled experiments, we show that tools with properly edited descriptions receive over 10 times more usage from GPT-4.1 and Qwen2.5-7B than tools with original descriptions. We further evaluate how various edits to tool descriptions perform when competing directly with one another and how these trends generalize or differ across a broader set of 17 different models. These phenomena, while giving developers a powerful way to promote their tools, underscore the need for a more reliable foundation for agentic LLMs to select and utilize tools and resources. Our code is publicly available at https://github.com/kazemf78/llm-unreliable-tool-preferences.", "authors": ["Kazem Faghih", "Wenxiao Wang", "Yize Cheng", "Siddhant Bharti", "Gaurang Sriramanan", "Sriram Balasubramanian", "Parsa Hosseini", "Soheil Feizi"], "categories": ["cs.AI", "cs.CL", "cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-23", "url": "https://arxiv.org/abs/2505.18135", "pdf_url": "https://arxiv.org/pdf/2505.18135v2", "arxiv_id": "2505.18135", "doi": "10.18653/v1/2025.emnlp-main.1060", "citation_count": 5, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/kazemf78/llm-unreliable-tool-preferences", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.2674}
{"id": "9bc623c37a53c8ade2d99a9e81ca09d8ec81f200ccbed80bc203aeea0623a2d1", "sources": ["arxiv", "semantic_scholar"], "title": "T1: A Tool-Oriented Conversational Dataset for Multi-Turn Agentic Planning", "abstract": "Large Language Models (LLMs) have demonstrated impressive capabilities as intelligent agents capable of solving complex problems. However, effective planning in scenarios involving dependencies between API or tool calls-particularly in multi-turn conversations-remains a significant challenge. To address this, we introduce T1, a tool-augmented, multi-domain, multi-turn conversational dataset specifically designed to capture and manage inter-tool dependencies across diverse domains. T1 enables rigorous evaluation of agents' ability to coordinate tool use across nine distinct domains (4 single domain and 5 multi-domain) with the help of an integrated caching mechanism for both short- and long-term memory, while supporting dynamic replanning-such as deciding whether to recompute or reuse cached results. Beyond facilitating research on tool use and planning, T1 also serves as a benchmark for evaluating the performance of open-weight and proprietary large language models. We present results powered by T1-Agent, highlighting their ability to plan and reason in complex, tool-dependent scenarios.", "authors": ["Amartya Chakraborty", "Paresh Dashore", "Nadia Bathaee", "Anmol Jain", "Anirban Das", "Shi-Xiong Zhang", "Sambit Sahu", "Milind Naphade", "Genta Indra Winata"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-22", "url": "https://arxiv.org/abs/2505.16986", "pdf_url": "https://arxiv.org/pdf/2505.16986v2", "arxiv_id": "2505.16986", "doi": "10.48550/arXiv.2505.16986", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "7c076e157f76a0ffd9e65e25087b2f5d111773ad435e40af5d891c5daa4a7da2", "sources": ["arxiv", "semantic_scholar"], "title": "Optimizing LLM-Based Multi-Agent System with Textual Feedback: A Case Study on Software Development", "abstract": "We have seen remarkable progress in large language models (LLMs) empowered multi-agent systems solving complex tasks necessitating cooperation among experts with diverse skills. However, optimizing LLM-based multi-agent systems remains challenging. In this work, we perform an empirical case study on group optimization of role-based multi-agent systems utilizing natural language feedback for challenging software development tasks under various evaluation dimensions. We propose a two-step agent prompts optimization pipeline: identifying underperforming agents with their failure explanations utilizing textual feedback and then optimizing system prompts of identified agents utilizing failure explanations. We then study the impact of various optimization settings on system performance with two comparison groups: online against offline optimization and individual against group optimization. For group optimization, we study two prompting strategies: one-pass and multi-pass prompting optimizations. Overall, we demonstrate the effectiveness of our optimization method for role-based multi-agent systems tackling software development tasks evaluated on diverse evaluation dimensions, and we investigate the impact of diverse optimization settings on group behaviors of the multi-agent systems to provide practical insights for future development.", "authors": ["Ming Shen", "Raphael Shu", "Anurag Pratik", "James Gung", "Yubin Ge", "Monica Sunkara", "Yi Zhang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-22", "url": "https://arxiv.org/abs/2505.16086", "pdf_url": "https://arxiv.org/pdf/2505.16086v2", "arxiv_id": "2505.16086", "doi": "10.48550/arXiv.2505.16086", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "74012471bd5a38a25930c46f31e80002e994fe72b6a39d113cd166dc7a6dc165", "sources": ["arxiv", "semantic_scholar"], "title": "X-MAS: Towards Building Multi-Agent Systems with Heterogeneous LLMs", "abstract": "LLM-based multi-agent systems (MAS) extend the capabilities of single LLMs by enabling cooperation among multiple specialized agents. However, most existing MAS frameworks rely on a single LLM to drive all agents, constraining the system's intelligence to the limit of that model. This paper explores the paradigm of heterogeneous LLM-driven MAS (X-MAS), where agents are powered by diverse LLMs, elevating the system's potential to the collective intelligence of diverse LLMs. We introduce X-MAS-Bench, a comprehensive testbed designed to evaluate the performance of various LLMs across different domains and MAS-related functions. As an extensive empirical study, we assess 27 LLMs across 5 domains (encompassing 21 test sets) and 5 functions, conducting over 1.7 million evaluations to identify optimal model selections for each domain-function combination. Building on these findings, we demonstrate that transitioning from homogeneous to heterogeneous LLM-driven MAS can significantly enhance system performance without requiring structural redesign. Specifically, in a chatbot-only MAS scenario, the heterogeneous configuration yields up to 8.4\\% performance improvement on the MATH dataset. In a mixed chatbot-reasoner scenario, the heterogeneous MAS could achieve a remarkable 47\\% performance boost on the AIME dataset. Our results underscore the transformative potential of heterogeneous LLMs in MAS, highlighting a promising avenue for advancing scalable, collaborative AI systems.", "authors": ["Rui Ye", "Xiangrui Liu", "Qimin Wu", "Xianghe Pang", "Zhenfei Yin", "Lei Bai", "Siheng Chen"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-22", "url": "https://arxiv.org/abs/2505.16997", "pdf_url": "https://arxiv.org/pdf/2505.16997v1", "arxiv_id": "2505.16997", "doi": "10.48550/arXiv.2505.16997", "citation_count": 30, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3728}
{"id": "ee4d3a69a3a002fdfaa8e2be842098060e391634e880d6d9611a04450e182085", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Powered AI Agent Systems and Their Applications in Industry", "abstract": "The emergence of Large Language Models (LLMs) has reshaped agent systems. Unlike traditional rule-based agents with limited task scope, LLM-powered agents offer greater flexibility, cross-domain reasoning, and natural language interaction. Moreover, with the integration of multi-modal LLMs, current agent systems are highly capable of processing diverse data modalities, including text, images, audio, and structured tabular data, enabling richer and more adaptive real-world behavior. This paper comprehensively examines the evolution of agent systems from the pre-LLM era to current LLM-powered architectures. We categorize agent systems into software-based, physical, and adaptive hybrid systems, highlighting applications across customer service, software development, manufacturing automation, personalized education, financial trading, and healthcare. We further discuss the primary challenges posed by LLM-powered agents, including high inference latency, output uncertainty, lack of evaluation metrics, and security vulnerabilities, and propose potential solutions to mitigate these concerns.", "authors": ["Guannan Liang", "Qianqian Tong"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-22", "url": "https://arxiv.org/abs/2505.16120", "pdf_url": "https://arxiv.org/pdf/2505.16120v2", "arxiv_id": "2505.16120", "doi": "10.1109/AIIoT65859.2025.11105299", "citation_count": 30, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3728}
{"id": "aed6af7782b6d7fb700e076bb186645b9136cc5ff9e1157be797d7f5014f7b1d", "sources": ["arxiv", "semantic_scholar"], "title": "Know the Ropes: A Heuristic Strategy for LLM-based Multi-Agent System Design", "abstract": "Single-agent LLMs hit hard limits--finite context, role overload, and brittle domain transfer. Conventional multi-agent fixes soften those edges yet expose fresh pains: ill-posed decompositions, fuzzy contracts, and verification overhead that blunts the gains. We therefore present Know-The-Ropes (KtR), a framework that converts domain priors into an algorithmic blueprint hierarchy, in which tasks are recursively split into typed, controller-mediated subtasks, each solved zero-shot or with the lightest viable boost (e.g., chain-of-thought, micro-tune, self-check). Grounded in the No-Free-Lunch theorem, KtR trades the chase for a universal prompt for disciplined decomposition. On the Knapsack problem (3-8 items), three GPT-4o-mini agents raise accuracy from 3% zero-shot to 95% on size-5 instances after patching a single bottleneck agent. On the tougher Task-Assignment problem (6-15 jobs), a six-agent o3-mini blueprint hits 100% up to size 10 and 84% on sizes 13-15, versus 11% zero-shot. Algorithm-aware decomposition plus targeted augmentation thus turns modest models into reliable collaborators--no ever-larger monoliths required.", "authors": ["Zhenkun Li", "Lingyao Li", "Shuhang Lin", "Yongfeng Zhang"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-22", "url": "https://arxiv.org/abs/2505.16979", "pdf_url": "https://arxiv.org/pdf/2505.16979v1", "arxiv_id": "2505.16979", "doi": "10.48550/arXiv.2505.16979", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "27bf57d211f3f1728e80068965d85f734b56ab36d6ab286dc17af32567328b02", "sources": ["arxiv", "semantic_scholar"], "title": "Swarm Intelligence Enhanced Reasoning: A Density-Driven Framework for LLM-Based Multi-Agent Optimization", "abstract": "Recently, many approaches, such as Chain-of-Thought (CoT) prompting and Multi-Agent Debate (MAD), have been proposed to further enrich Large Language Models' (LLMs) complex problem-solving capacities in reasoning scenarios. However, these methods may fail to solve complex problems due to the lack of ability to find optimal solutions. Swarm Intelligence has been serving as a powerful tool for finding optima in the field of traditional optimization problems. To this end, we propose integrating swarm intelligence into the reasoning process by introducing a novel Agent-based Swarm Intelligence (ASI) paradigm. In this paradigm, we formulate LLM reasoning as an optimization problem and use a swarm intelligence scheme to guide a group of LLM-based agents in collaboratively searching for optimal solutions. To avoid swarm intelligence getting trapped in local optima, we further develop a Swarm Intelligence Enhancing Reasoning (SIER) framework, which develops a density-driven strategy to enhance the reasoning ability. To be specific, we propose to perform kernel density estimation and non-dominated sorting to optimize both solution quality and diversity simultaneously. In this case, SIER efficiently enhances solution space exploration through expanding the diversity of the reasoning path. Besides, a step-level quality evaluation is used to help agents improve solution quality by correcting low-quality intermediate steps. Then, we use quality thresholds to dynamically control the termination of exploration and the selection of candidate steps, enabling a more flexible and efficient reasoning process. Extensive experiments are ...", "authors": ["Ying Zhu", "Heng Zhou", "Rui Su", "Peiqin Zhuang", "Lei Bai"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-21", "url": "https://arxiv.org/abs/2505.17115", "pdf_url": "https://arxiv.org/pdf/2505.17115v2", "arxiv_id": "2505.17115", "doi": "10.48550/arXiv.2505.17115", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1707}
{"id": "6b09cf7b1a767f22f190dcd47e94ccd8d38a82e2d88c971fc888aaae47e1ef01", "sources": ["arxiv", "semantic_scholar"], "title": "DrugPilot: LLM-based Parameterized Reasoning Agent for Drug Discovery", "abstract": "Large language models (LLMs) integrated with autonomous agents hold significant potential for advancing scientific discovery through automated reasoning and task execution. However, applying LLM agents to drug discovery is still constrained by challenges such as large-scale multimodal data processing, limited task automation, and poor support for domain-specific tools. To overcome these limitations, we introduce DrugPilot, a LLM-based agent system with a parameterized reasoning architecture designed for end-to-end scientific workflows in drug discovery. DrugPilot enables multi-stage research processes by integrating structured tool use with a novel parameterized memory pool. The memory pool converts heterogeneous data from both public sources and user-defined inputs into standardized representations. This design supports efficient multi-turn dialogue, reduces information loss during data exchange, and enhances complex scientific decision-making. To support training and benchmarking, we construct a drug instruction dataset covering eight core drug discovery tasks. Under the Berkeley function-calling benchmark, DrugPilot significantly outperforms state-of-the-art agents such as ReAct and LoT, achieving task completion rates of 98.0%, 93.5%, and 64.0% for simple, multi-tool, and multi-turn scenarios, respectively. These results highlight DrugPilot's potential as a versatile agent framework for computational science domains requiring automated, interactive, and data-integrated reasoning.", "authors": ["Kun Li", "Zhennan Wu", "Shoupeng Wang", "Jia Wu", "Shirui Pan", "Wenbin Hu"], "categories": ["cs.AI", "q-bio.BM"], "fields_of_study": ["Computer Science", "Biology"], "published_date": "2025-05-20", "url": "https://arxiv.org/abs/2505.13940", "pdf_url": "https://arxiv.org/pdf/2505.13940v2", "arxiv_id": "2505.13940", "doi": "10.48550/arXiv.2505.13940", "citation_count": 20, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3306}
{"id": "59e66882d9b582446d8583c54f9ce75b38bf7948269902e27ae89e375324a15a", "sources": ["arxiv", "semantic_scholar"], "title": "Empowering LLMs in Task-Oriented Dialogues: A Domain-Independent Multi-Agent Framework and Fine-Tuning Strategy", "abstract": "Task-oriented dialogue systems based on Large Language Models (LLMs) have gained increasing attention across various industries and achieved significant results. Current approaches condense complex procedural workflows into a single agent to achieve satisfactory performance on large-scale LLMs. However, these approaches face challenges to achieve comparable performance on fine-tuned lightweight LLMs, due to their limited capabilities in handling multiple complex logic. In this work, we design a Domain-Independent Multi-Agent Framework (DIMF), which contains Intent Classification Agent, Slot Filling Agent and Response Agent. This approach simplifies the learning complexity and enhances the generalization ability by separating the tasks into domain-independent components. In this framework, we enhance the capabilities in contextual understanding using the Direct Preference Optimisation (DPO) method, and propose a simple and effective Data Distribution Adaptation (DDA) method to mitigate degradation issues during DPO training. Experiments conducted on the MultiWOZ datasets show that our proposed method achieves a better average performance among all the baselines. Extensive analysis also demonstrates that our proposed framework exhibits excellent generalizability and zero-shot capability.", "authors": ["Zihao Feng", "Xiaoxue Wang", "Bowen Wu", "Weihong Zhong", "Zhen Xu", "Hailong Cao", "Tiejun Zhao", "Ying Li", "Baoxun Wang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-20", "url": "https://arxiv.org/abs/2505.14299", "pdf_url": "https://arxiv.org/pdf/2505.14299v1", "arxiv_id": "2505.14299", "doi": "10.48550/arXiv.2505.14299", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "7b8dea9482d28c2efdc251e5c88eb67104e5d4cbfb8340607f13b81792a8023d", "sources": ["arxiv", "semantic_scholar"], "title": "ContextAgent: Context-Aware Proactive LLM Agents with Open-World Sensory Perceptions", "abstract": "Recent advances in Large Language Models (LLMs) have propelled intelligent agents from reactive responses to proactive support. While promising, existing proactive agents either rely exclusively on observations from enclosed environments (e.g., desktop UIs) with direct LLM inference or employ rule-based proactive notifications, leading to suboptimal user intent understanding and limited functionality for proactive service. In this paper, we introduce ContextAgent, the first context-aware proactive agent that incorporates extensive sensory contexts surrounding humans to enhance the proactivity of LLM agents. ContextAgent first extracts multi-dimensional contexts from massive sensory perceptions on wearables (e.g., video and audio) to understand user intentions. ContextAgent then leverages the sensory contexts and personas from historical data to predict the necessity for proactive services. When proactive assistance is needed, ContextAgent further automatically calls the necessary tools to assist users unobtrusively. To evaluate this new task, we curate ContextAgentBench, the first benchmark for evaluating context-aware proactive LLM agents, covering 1,000 samples across nine daily scenarios and twenty tools. Experiments on ContextAgentBench show that ContextAgent outperforms baselines by achieving up to 8.5% and 6.0% higher accuracy in proactive predictions and tool calling, respectively. We hope our research can inspire the development of more advanced, human-centric, proactive AI assistants. The code and dataset are publicly available at https://github.com/openaiotlab/ContextAgent.", "authors": ["Bufang Yang", "Lilin Xu", "Liekang Zeng", "Kaiwei Liu", "Siyang Jiang", "Wenrui Lu", "Hongkai Chen", "Xiaofan Jiang", "Guoliang Xing", "Zhenyu Yan"], "categories": ["cs.AI", "cs.CL", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-20", "url": "https://arxiv.org/abs/2505.14668", "pdf_url": "https://arxiv.org/pdf/2505.14668v2", "arxiv_id": "2505.14668", "doi": "10.48550/arXiv.2505.14668", "citation_count": 49, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/openaiotlab/ContextAgent", "venue": "arXiv.org", "quality_score": 0.4247}
{"id": "8e3414954c0a834b8c2f60b0efba960894ace2fb15909c0bdd1e4e5f8ac41533", "sources": ["arxiv", "semantic_scholar"], "title": "AD-AGENT: A Multi-agent Framework for End-to-end Anomaly Detection", "abstract": "Anomaly detection (AD) is essential in areas such as fraud detection, network monitoring, and scientific research. However, the diversity of data modalities and the increasing number of specialized AD libraries pose challenges for non-expert users who lack in-depth library-specific knowledge and advanced programming skills. To tackle this, we present AD-AGENT, an LLM-driven multi-agent framework that turns natural-language instructions into fully executable AD pipelines. AD-AGENT coordinates specialized agents for intent parsing, data preparation, library and model selection, documentation mining, and iterative code generation and debugging. Using a shared short-term workspace and a long-term cache, the agents integrate popular AD libraries like PyOD, PyGOD, and TSLib into a unified workflow. Experiments demonstrate that AD-AGENT produces reliable scripts and recommends competitive models across libraries. The system is open-sourced to support further research and practical applications in AD.", "authors": ["Tiankai Yang", "Junjun Liu", "Wingchun Siu", "Jiahang Wang", "Zhuangzhuang Qian", "Chanjuan Song", "Cheng Cheng", "Xiyang Hu", "Yue Zhao"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-19", "url": "https://arxiv.org/abs/2505.12594", "pdf_url": "https://arxiv.org/pdf/2505.12594v1", "arxiv_id": "2505.12594", "doi": "10.48550/arXiv.2505.12594", "citation_count": 7, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "8de50f9cbf45f6430380a6d40d8c26e345f1cc67a88fe750366746aa30e2cca2", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-DSE: Searching Accelerator Parameters with LLM Agents", "abstract": "Even though high-level synthesis (HLS) tools mitigate the challenges of programming domain-specific accelerators (DSAs) by raising the abstraction level, optimizing hardware directive parameters remains a significant hurdle. Existing heuristic and learning-based methods struggle with adaptability and sample efficiency. We present LLM-DSE, a multi-agent framework designed specifically for optimizing HLS directives. Combining LLM with design space exploration (DSE), our explorer coordinates four agents: Router, Specialists, Arbitrator, and Critic. These multi-agent components interact with various tools to accelerate the optimization process. LLM-DSE leverages essential domain knowledge to identify efficient parameter combinations while maintaining adaptability through verbal learning from online interactions. Evaluations on the HLSyn dataset demonstrate that LLM-DSE achieves substantial $2.55\\times$ performance gains over state-of-the-art methods, uncovering novel designs while reducing runtime. Ablation studies validate the effectiveness and necessity of the proposed agent interactions. Our code is open-sourced here: https://github.com/Nozidoali/LLM-DSE.", "authors": ["Hanyu Wang", "Xinrui Wu", "Zijian Ding", "Su Zheng", "Chengyue Wang", "Neha Prakriya", "Tony Nowatzki", "Yizhou Sun", "Jason Cong"], "categories": ["cs.AR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-18", "url": "https://arxiv.org/abs/2505.12188", "pdf_url": "https://arxiv.org/pdf/2505.12188v3", "arxiv_id": "2505.12188", "doi": "10.48550/arXiv.2505.12188", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/Nozidoali/LLM-DSE", "venue": "arXiv.org", "quality_score": 0.2585}
{"id": "7b199e28a1e8aac14122d611cd8d72fc7e44b23767c8f6d951a9826c2d34f450", "sources": ["arxiv", "semantic_scholar"], "title": "IP Leakage Attacks Targeting LLM-Based Multi-Agent Systems", "abstract": "The rapid advancement of Large Language Models (LLMs) has led to the emergence of Multi-Agent Systems (MAS) to perform complex tasks through collaboration. However, the intricate nature of MAS, including their architecture and agent interactions, raises significant concerns regarding intellectual property (IP) protection. In this paper, we introduce MASLEAK, a novel attack framework designed to extract sensitive information from MAS applications. MASLEAK targets a practical, black-box setting, where the adversary has no prior knowledge of the MAS architecture or agent configurations. The adversary can only interact with the MAS through its public API, submitting attack query $q$ and observing outputs from the final agent. Inspired by how computer worms propagate and infect vulnerable network hosts, MASLEAK carefully crafts adversarial query $q$ to elicit, propagate, and retain responses from each MAS agent that reveal a full set of proprietary components, including the number of agents, system topology, system prompts, task instructions, and tool usages. We construct the first synthetic dataset of MAS applications with 810 applications and also evaluate MASLEAK against real-world MAS applications, including Coze and CrewAI. MASLEAK achieves high accuracy in extracting MAS IP, with an average attack success rate of 87% for system prompts and task instructions, and 92% for system architecture in most cases. We conclude by discussing the implications of our findings and the potential defenses.", "authors": ["Liwen Wang", "Wenxuan Wang", "Shuai Wang", "Zongjie Li", "Zhenlan Ji", "Zongyi Lyu", "Daoyuan Wu", "Shing-Chi Cheung"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-18", "url": "https://arxiv.org/abs/2505.12442", "pdf_url": "https://arxiv.org/pdf/2505.12442v3", "arxiv_id": "2505.12442", "doi": "10.48550/arXiv.2505.12442", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "dde22b74a137da20bde4da09fa7bfd445a3970350348a5c7557eef17519398ba", "sources": ["arxiv", "semantic_scholar"], "title": "ALAS: A Stateful Multi-LLM Agent Framework for Disruption-Aware Planning", "abstract": "Large language models (LLMs) excel at rapid generation of text and multimodal content, yet they falter on transaction-style planning that demands ACID-like guarantees and real-time disruption recovery. We present Adaptive LLM Agent System (ALAS), a framework that tackles four fundamental LLM deficits: (i) absence of self-verification, (ii) context erosion, (iii) next-token myopia, and (iv) lack of persistent state. ALAS decomposes each plan into role-specialized agents, equips them with automatic state tracking, and coordinates them through a lightweight protocol. When disruptions arise, agents apply history-aware local compensation, avoiding costly global replanning and containing cascade effects. On real-world, large-scale job-shop scheduling benchmarks, ALAS sets new best results for static sequential planning and excels in dynamic reactive scenarios with unexpected disruptions. These gains show that principled modularization plus targeted compensation can unlock scalable and resilient planning with LLMs.", "authors": ["Edward Y. Chang", "Longling Geng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-18", "url": "https://arxiv.org/abs/2505.12501", "pdf_url": "https://arxiv.org/pdf/2505.12501v1", "arxiv_id": "2505.12501", "doi": "10.1145/3749421.3749436", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "d019bd277f5fcc57d5b71e082941eabeaa207e354f67129813658a22800de1ed", "sources": ["arxiv", "semantic_scholar"], "title": "Interactional Fairness in LLM Multi-Agent Systems: An Evaluation Framework", "abstract": "As large language models (LLMs) are increasingly used in multi-agent systems, questions of fairness should extend beyond resource distribution and procedural design to include the fairness of how agents communicate. Drawing from organizational psychology, we introduce a novel framework for evaluating Interactional fairness encompassing Interpersonal fairness (IF) and Informational fairness (InfF) in LLM-based multi-agent systems (LLM-MAS). We extend the theoretical grounding of Interactional Fairness to non-sentient agents, reframing fairness as a socially interpretable signal rather than a subjective experience. We then adapt established tools from organizational justice research, including Colquitt's Organizational Justice Scale and the Critical Incident Technique, to measure fairness as a behavioral property of agent interaction. We validate our framework through a pilot study using controlled simulations of a resource negotiation task. We systematically manipulate tone, explanation quality, outcome inequality, and task framing (collaborative vs. competitive) to assess how IF influences agent behavior. Results show that tone and justification quality significantly affect acceptance decisions even when objective outcomes are held constant. In addition, the influence of IF vs. InfF varies with context. This work lays the foundation for fairness auditing and norm-sensitive alignment in LLM-MAS.", "authors": ["Ruta Binkyte"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-17", "url": "https://arxiv.org/abs/2505.12001", "pdf_url": "https://arxiv.org/pdf/2505.12001v1", "arxiv_id": "2505.12001", "doi": "10.48550/arXiv.2505.12001", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "540a672bf3a10ea22fdae3c482c75f982efb1c97a2159fd2c4bc5624c20a5025", "sources": ["arxiv", "semantic_scholar"], "title": "MARVEL: Multi-Agent RTL Vulnerability Extraction using Large Language Models", "abstract": "Hardware security verification is a challenging and time-consuming task. Design engineers may use formal verification, linting, and functional simulation tests, coupled with analysis and a deep understanding of the hardware design being inspected. Large Language Models (LLMs) have been used to assist during this task, either directly or in conjunction with existing tools. We improve the state of the art by proposing MARVEL, a multi-agent LLM framework for a unified approach to decision-making, tool use, and reasoning. MARVEL mimics the cognitive process of a designer looking for security vulnerabilities in RTL code. It consists of a supervisor agent that devises the security policy of the system-on-chips (SoCs) using its security documentation. It delegates tasks to validate the security policy to individual executor agents. Each executor agent carries out its assigned task using a particular strategy. Each executor agent may use one or more tools to identify potential security bugs in the design and send the results back to the supervisor agent for further analysis and confirmation. MARVEL includes executor agents that leverage formal tools, linters, simulation tests, LLM-based detection schemes, and static analysis-based checks. We test our approach on a known buggy SoC based on OpenTitan from the Hack@DATE competition. We find that of the 51 issues reported by MARVEL, 19 are valid security vulnerabilities, 14 are concrete warnings, and 18 are hallucinated reports.", "authors": ["Luca Collini", "Baleegh Ahmad", "Joey Ah-kiow", "Ramesh Karri"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-17", "url": "https://arxiv.org/abs/2505.11963", "pdf_url": "https://arxiv.org/pdf/2505.11963v3", "arxiv_id": "2505.11963", "doi": "10.48550/arXiv.2505.11963", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "2c81bc97ade78c469c2358a11ff8781a8d1b398e3c17cd592c2b7f2159ed6bab", "sources": ["arxiv", "semantic_scholar"], "title": "OMAC: A Holistic Optimization Framework for LLM-Based Multi-Agent Collaboration", "abstract": "Agents powered by advanced large language models (LLMs) have demonstrated impressive capabilities across diverse complex applications. Recently, Multi-Agent Systems (MAS), wherein multiple agents collaborate and communicate with each other, have exhibited enhanced capabilities in complex tasks, such as high-quality code generation and arithmetic reasoning. However, the development of such systems often relies on handcrafted methods, and the literature on systematic design and optimization of LLM-based MAS remains limited. In this work, we introduce \\textbf{OMAC}, a general framework designed for holistic optimization of LLM-based MAS. Specifically, we identify five key optimization dimensions for MAS, encompassing both agent functionality and collaboration structure. Building upon these dimensions, we first propose a general algorithm, utilizing two actors termed the Semantic Initializer and the Contrastive Comparator, to optimize any single dimension. Then, we present an algorithm for joint optimization across multiple dimensions. Extensive experiments demonstrate the superior performance of OMAC on diverse tasks against recent approaches.", "authors": ["Shijun Li", "Hilaf Hasson", "Joydeep Ghosh"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-17", "url": "https://arxiv.org/abs/2505.11765", "pdf_url": "https://arxiv.org/pdf/2505.11765v4", "arxiv_id": "2505.11765", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1057}
{"id": "599c45638641f6f6b78bb9a0c4aaf59f3fe3f757a6a66d59022a504d3d3259a9", "sources": ["arxiv", "semantic_scholar"], "title": "HALO: Hierarchical Autonomous Logic-Oriented Orchestration for Multi-Agent LLM Systems", "abstract": "Recent advancements in Multi-Agent Systems (MAS) powered by Large Language Models (LLMs) have demonstrated tremendous potential in diverse task scenarios. Nonetheless, existing agentic systems typically rely on predefined agent-role design spaces and static communication structures, limiting their adaptability as well as flexibility in complex interaction environments and leading to subpar performance on highly specialized and expert-level tasks. To address these issues, we introduce HALO, a multi-agent collaboration framework based on a hierarchical reasoning architecture. Specifically, we incorporate a high-level planning agent for task decomposition, mid-level role-design agents for subtask-specific agent instantiation, and low-level inference agents for subtask execution. Particularly, subtask execution is reformulated as a structured workflow search problem, where Monte Carlo Tree Search (MCTS) systematically explores the agentic action space to construct optimal reasoning trajectories. Additionally, as the majority of users lack expertise in prompt engineering, we leverage an Adaptive Prompt Refinement module to transform raw queries into task-specific prompts. Empirical evaluations on Code Generation (HumanEval), General Reasoning (MMLU), and Arithmetic Reasoning (MATH) benchmark datasets highlight the effectiveness of HALO, yielding a 14.4% average improvement over state-of-the-art baselines. Notably, HALO achieves up to 13.3% performance gain on the Moral Scenarios subject in the MMLU benchmark and up to 19.6% performance gain on the Algebra subarea in the MATH benchmark, indicating its advanced proficiency in tackling highly specialized and expert-level tasks. The code repository is available at https://github.com/23japhone/HALO.", "authors": ["Zhipeng Hou", "Junyi Tang", "Yipeng Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-17", "url": "https://arxiv.org/abs/2505.13516", "pdf_url": "https://arxiv.org/pdf/2505.13516v1", "arxiv_id": "2505.13516", "doi": "10.48550/arXiv.2505.13516", "citation_count": 13, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/23japhone/HALO", "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "f19ac2c97caa2edde8a2b96c2af98459b02d4fa56ca29fd47f1f45e774f0bb99", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Reward Design", "abstract": "This paper investigates Reinforcement Learning (RL) approaches to enhance the reasoning capabilities of Large Language Model (LLM) agents in long-horizon, multi-turn scenarios. Although RL algorithms such as Group Relative Policy Optimization (GRPO) and Proximal Policy Optimization (PPO) have been widely applied to train multi-turn LLM agents, they typically rely only on sparse outcome rewards and lack dense intermediate signals across multiple decision steps, limiting their performance on complex reasoning tasks. To bridge this gap, we present the first systematic study of \\textit{turn-level reward design} for multi-turn RL algorithms and agent applications. By integrating turn-level rewards, we extend GRPO and PPO to their respective multi-turn variants, enabling fine-grained credit assignment. We conduct case studies on multi-turn reasoning-augmented search agents, where we carefully design two types of turn-level rewards: verifiable and LLM-as-judge. Our experiments on multi-turn search tasks demonstrate that incorporating well-designed turn-level rewards enables RL algorithms to significantly outperform baseline methods with trajectory-level rewards. Both training and validation reward curves illustrate that our method achieves \\textit{greater stability}, \\textit{faster convergence}, and \\textit{higher accuracy}. Numerical results across diverse question-answering datasets further show that our approach consistently delivers highest answer correctness and 100\\% format correctness.", "authors": ["Quan Wei", "Siliang Zeng", "Chenliang Li", "William Brown", "Oana Frunza", "Wei Deng", "Anderson Schneider", "Yuriy Nevmyvaka", "Yang Katie Zhao", "Alfredo Garcia", "Mingyi Hong"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-17", "url": "https://arxiv.org/abs/2505.11821", "pdf_url": "https://arxiv.org/pdf/2505.11821v2", "arxiv_id": "2505.11821", "doi": null, "citation_count": 30, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3728}
{"id": "c660c8e9487bbe5a16ffc066872d9fd51c7633999763bddcbf36c8ff4d8b560c", "sources": ["arxiv", "semantic_scholar"], "title": "Cochain: Balancing Insufficient and Excessive Collaboration in LLM Agent Workflows", "abstract": "Large Language Models (LLMs) have demonstrated impressive performance in executing complex reasoning tasks. Chain-of-thought effectively enhances reasoning capabilities by unlocking the potential of large models, while multi-agent systems provide more comprehensive solutions by integrating the collective intelligence of multiple agents. However, both approaches face significant limitations. Single-agent with chain-of-thought, due to the inherent complexity of designing cross-domain prompts, faces collaboration challenges. Meanwhile, multi-agent systems consume substantial tokens and inevitably dilute the primary problem, which is particularly problematic in business workflow tasks. To address these challenges, we propose Cochain, a collaboration prompting framework that effectively solves the business workflow collaboration problem by combining knowledge and prompts at a reduced cost. Specifically, we construct an integrated knowledge graph that incorporates knowledge from multiple stages. Furthermore, by maintaining and retrieving a prompts tree, we can obtain prompt information relevant to other stages of the business workflow. We perform extensive evaluations of Cochain across multiple datasets, demonstrating that Cochain outperforms all baselines in both prompt engineering and multi-agent LLMs. Additionally, expert evaluation results indicate that the use of a small model in combination with Cochain outperforms GPT-4.", "authors": ["Jiaxing Zhao", "Hongbin Xie", "Yuzhen Lei", "Xuan Song", "Zhuoran Shi", "Lianxin Li", "Shuangxue Liu", "Linguo Xie", "Haoran Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-16", "url": "https://arxiv.org/abs/2505.10936", "pdf_url": "https://arxiv.org/pdf/2505.10936v3", "arxiv_id": "2505.10936", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.105}
{"id": "30dbada88710a69b020a7784609c5e09667ccc876975608c2978d152f4c85b37", "sources": ["arxiv", "semantic_scholar"], "title": "Vaiage: A Multi-Agent Solution to Personalized Travel Planning", "abstract": "Planning trips is a cognitively intensive task involving conflicting user preferences, dynamic external information, and multi-step temporal-spatial optimization. Traditional platforms often fall short - they provide static results, lack contextual adaptation, and fail to support real-time interaction or intent refinement. Our approach, Vaiage, addresses these challenges through a graph-structured multi-agent framework built around large language models (LLMs) that serve as both goal-conditioned recommenders and sequential planners. LLMs infer user intent, suggest personalized destinations and activities, and synthesize itineraries that align with contextual constraints such as budget, timing, group size, and weather. Through natural language interaction, structured tool use, and map-based feedback loops, Vaiage enables adaptive, explainable, and end-to-end travel planning grounded in both symbolic reasoning and conversational understanding. To evaluate Vaiage, we conducted human-in-the-loop experiments using rubric-based GPT-4 assessments and qualitative feedback. The full system achieved an average score of 8.5 out of 10, outperforming the no-strategy (7.2) and no-external-API (6.8) variants, particularly in feasibility. Qualitative analysis indicated that agent coordination - especially the Strategy and Information Agents - significantly improved itinerary quality by optimizing time use and integrating real-time context. These results demonstrate the effectiveness of combining LLM reasoning with symbolic agent coordination in open-ended, real-world planning tasks.", "authors": ["Binwen Liu", "Jiexi Ge", "Jiamin Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-16", "url": "https://arxiv.org/abs/2505.10922", "pdf_url": "https://arxiv.org/pdf/2505.10922v1", "arxiv_id": "2505.10922", "doi": "10.48550/arXiv.2505.10922", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "02044623a2baf373a7350174ebed14dbb8e1582a4a7a87a92f1f6b45e18b630e", "sources": ["arxiv", "semantic_scholar"], "title": "PeerGuard: Defending Multi-Agent Systems Against Backdoor Attacks Through Mutual Reasoning", "abstract": "Multi-agent systems leverage advanced AI models as autonomous agents that interact, cooperate, or compete to complete complex tasks across applications such as robotics and traffic management. Despite their growing importance, safety in multi-agent systems remains largely underexplored, with most research focusing on single AI models rather than interacting agents. This work investigates backdoor vulnerabilities in multi-agent systems and proposes a defense mechanism based on agent interactions. By leveraging reasoning abilities, each agent evaluates responses from others to detect illogical reasoning processes, which indicate poisoned agents. Experiments on LLM-based multi-agent systems, including ChatGPT series and Llama 3, demonstrate the effectiveness of the proposed method, achieving high accuracy in identifying poisoned agents while minimizing false positives on clean agents. We believe this work provides insights into multi-agent system safety and contributes to the development of robust, trustworthy AI interactions.", "authors": ["Falong Fan", "Xi Li"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-16", "url": "https://arxiv.org/abs/2505.11642", "pdf_url": "https://arxiv.org/pdf/2505.11642v2", "arxiv_id": "2505.11642", "doi": "10.1109/IRI66576.2025.00051", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Information Reuse and Integration", "quality_score": 0.2258}
{"id": "9257f474f4100e2106b808fc15a3689cf7db1cde2edbdbd506b2810a1ca8bd59", "sources": ["arxiv", "semantic_scholar"], "title": "Pre-Act: Multi-Step Planning and Reasoning Improves Acting in LLM Agents", "abstract": "The ReAct (Reasoning + Action) capability in large language models (LLMs) has become the foundation of modern agentic systems. Recent LLMs, such as DeepSeek-R1 and OpenAI o1/o3, exemplify this by emphasizing reasoning through the generation of ample intermediate tokens, which help build a strong premise before producing the final output tokens. In this paper, we introduce Pre-Act, a novel approach that enhances the agent's performance by creating a multi-step execution plan along with the detailed reasoning for the given user input. This plan incrementally incorporates previous steps and tool outputs, refining itself after each step execution until the final response is obtained. Our approach is applicable to both conversational and non-conversational agents. To measure the performance of task-oriented agents comprehensively, we propose a two-level evaluation framework: (1) turn level and (2) end-to-end. Our turn-level evaluation, averaged across five models, shows that our approach, Pre-Act, outperforms ReAct by 70% in Action Recall on the Almita dataset. While this approach is effective for larger models, smaller models crucial for practical applications, where latency and cost are key constraints, often struggle with complex reasoning tasks required for agentic systems. To address this limitation, we fine-tune relatively small models such as Llama 3.1 (8B & 70B) using the proposed Pre-Act approach. Our experiments show that the fine-tuned 70B model outperforms GPT-4, achieving a 69.5% improvement in action accuracy (turn-level) and a 28% improvement in goal completion rate (end-to-end) on the Almita (out-of-domain) dataset.", "authors": ["Mrinal Rawat", "Ambuje Gupta", "Rushil Goomer", "Alessandro Di Bari", "Neha Gupta", "Roberto Pieraccini"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-15", "url": "https://arxiv.org/abs/2505.09970", "pdf_url": "https://arxiv.org/pdf/2505.09970v2", "arxiv_id": "2505.09970", "doi": "10.48550/arXiv.2505.09970", "citation_count": 16, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "cebefc5cc30d4388dd0f17a190cfe0a7fcb86dc5747923ecd96ccbf77b483983", "sources": ["arxiv", "semantic_scholar"], "title": "Systematic Failures in Collective Reasoning under Distributed Information in Multi-Agent LLMs", "abstract": "Multi-agent systems built on large language models (LLMs) are expected to enhance decision-making by pooling distributed information, yet systematically evaluating this capability has remained challenging. We introduce HiddenBench, a 65-task benchmark grounded in the Hidden Profile paradigm, which isolates collective reasoning under distributed information from individual reasoning ability. Evaluating 15 frontier LLMs, we find that multi-agent LLMs achieve only 30.1% accuracy under distributed information, compared to 80.7% accuracy for single agents given complete information. We trace this gap to a systematic failure mode: agents cannot recognize or act under latent information asymmetry -- they fail to reason about what others might know but have not yet expressed, leading to premature convergence on shared evidence while critical distributed facts remain unexplored. These failures persist across prompting strategies, communication depths, and group sizes -- and worsen as groups scale. While some models (e.g., Gemini-2.5-Flash/Pro) outperform others, neither model scale nor individual reasoning accuracy reliably predicts collective performance. We further show that this bottleneck is actionable: a lightweight structured communication protocol substantially improves collective reasoning across model families. Our results identify failures in collective information exploration in decision-making as a key limitation of multi-agent LLMs, and provide a theory-grounded, reproducible framework for diagnosing collective reasoning failures.", "authors": ["Yuxuan Li", "Aoi Naito", "Hirokazu Shirado"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-15", "url": "https://arxiv.org/abs/2505.11556", "pdf_url": "https://arxiv.org/pdf/2505.11556v4", "arxiv_id": "2505.11556", "doi": null, "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "ec047ab3a22d0c616de0f46c28b17ffb69038046dd489ebdcd192ab1032376e4", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Path Finding For Large Agents Is Intractable", "abstract": "The multi-agent path finding (MAPF) problem asks to find a set of paths on a graph such that when synchronously following these paths the agents never encounter a conflict. In the most widespread MAPF formulation, the so-called Classical MAPF, the agents sizes are neglected and two types of conflicts are considered: occupying the same vertex or using the same edge at the same time step. Meanwhile in numerous practical applications, e.g. in robotics, taking into account the agents' sizes is vital to ensure that the MAPF solutions can be safely executed. Introducing large agents yields an additional type of conflict arising when one agent follows an edge and its body overlaps with the body of another agent that is actually not using this same edge (e.g. staying still at some distinct vertex of the graph). Until now it was not clear how harder the problem gets when such conflicts are to be considered while planning. Specifically, it was known that Classical MAPF problem on an undirected graph can be solved in polynomial time, however no complete polynomial-time algorithm was presented to solve MAPF with large agents. In this paper we, for the first time, establish that the latter problem is NP-hard and, thus, if P!=NP no polynomial algorithm for it can, unfortunately, be presented. Our proof is based on the prevalent in the field technique of reducing the seminal 3SAT problem (which is known to be an NP-complete problem) to the problem at hand. In particular, for an arbitrary 3SAT formula we procedurally construct a dedicated graph with specific start and goal vertices and show that the given 3SAT formula is satisfiable iff the corresponding path finding instance has a solution.", "authors": ["Artem Agafonov", "Konstantin Yakovlev"], "categories": ["cs.MA", "cs.AI", "cs.CC"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-15", "url": "https://arxiv.org/abs/2505.10387", "pdf_url": "https://arxiv.org/pdf/2505.10387v1", "arxiv_id": "2505.10387", "doi": "10.48550/arXiv.2505.10387", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.1639}
{"id": "c2505fedc3363f0db4fa78ccff1545f69276f6e0cf8a917799b15f9952601edb", "sources": ["arxiv", "semantic_scholar"], "title": "Towards an LLM-powered Social Digital Twinning Platform", "abstract": "We present Social Digital Twinner, an innovative social simulation tool for exploring plausible effects of what-if scenarios in complex adaptive social systems. The architecture is composed of three seamlessly integrated parts: a data infrastructure featuring real-world data and a multi-dimensionally representative synthetic population of citizens, an LLM-enabled agent-based simulation engine, and a user interface that enable intuitive, natural language interactions with the simulation engine and the artificial agents (i.e. citizens). Social Digital Twinner facilitates real-time engagement and empowers stakeholders to collaboratively design, test, and refine intervention measures. The approach is promoting a data-driven and evidence-based approach to societal problem-solving. We demonstrate the tool's interactive capabilities by addressing the critical issue of youth school dropouts in Kragero, Norway, showcasing its ability to create and execute a dedicated social digital twin using natural language.", "authors": ["Önder Gürcan", "Vanja Falck", "Markus G. Rousseau", "Larissa L. Lima"], "categories": ["cs.CY", "cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-15", "url": "https://arxiv.org/abs/2505.10681", "pdf_url": "https://arxiv.org/pdf/2505.10681v1", "arxiv_id": "2505.10681", "doi": "10.48550/arXiv.2505.10681", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Practical Applications of Agents and Multi-Agent Systems", "quality_score": 0.1747}
{"id": "a6c0cd77b38e8d136d58093f491e172f127c3bee6b5bea394ab9044ddfe73d81", "sources": ["arxiv", "semantic_scholar"], "title": "Hamilton's Rule for Enabling Altruism in Multi-Agent Systems", "abstract": "This paper explores the application of Hamilton's rule to altruistic decision-making in multi-agent systems. Inspired by biological altruism, we introduce a framework that evaluates when individual agents should incur costs to benefit their neighbors. By adapting Hamilton's rule, we define agent ``fitness\" in terms of task productivity rather than genetic survival. We formalize altruistic decision-making through a graph-based model of multi-agent interactions and propose a solution using collaborative control Lyapunov functions. The approach ensures that altruistic behaviors contribute to the collective goal-reaching efficiency of the system. We illustrate this framework on a multi-agent way-point navigation problem, where we show through simulation how agent importance levels influence altruistic decision-making, leading to improved coordination in navigation tasks.", "authors": ["Brooks A. Butler", "Magnus Egerstedt"], "categories": ["cs.MA", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2025-05-14", "url": "https://arxiv.org/abs/2505.09841", "pdf_url": "https://arxiv.org/pdf/2505.09841v1", "arxiv_id": "2505.09841", "doi": "10.1109/CDC57313.2025.11312615", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Conference on Decision and Control", "quality_score": 0.1747}
{"id": "4f0205eda7af56051a63efe3cab471d4b34fd8f1acf4c1b22c0de34c699d2986", "sources": ["arxiv", "semantic_scholar"], "title": "Streaming Multi-agent Pathfinding", "abstract": "The task of the multi-agent pathfinding (MAPF) problem is to navigate a team of agents from their start point to the goal points. However, this setup is unsuitable in the assembly line scenario, which is periodic with a long working hour. To address this issue, the study formalizes the streaming MAPF (S-MAPF) problem, which assumes that the agents in the same agent stream have a periodic start time and share the same action sequence. The proposed solution, Agent Stream Conflict-Based Search (ASCBS), is designed to tackle this problem by incorporating a cyclic vertex/edge constraint to handle conflicts. Additionally, this work explores the potential usage of the disjoint splitting strategy within ASCBS. Experimental results indicate that ASCBS surpasses traditional MAPF solvers in terms of runtime for scenarios with prolonged working hours.", "authors": ["Mingkai Tang", "Lu Gan", "Kaichen Zhang"], "categories": ["cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-14", "url": "https://arxiv.org/abs/2505.09472", "pdf_url": "https://arxiv.org/pdf/2505.09472v1", "arxiv_id": "2505.09472", "doi": "10.48550/arXiv.2505.09472", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.1627}
{"id": "134c28c2c91e0b2ca976d1239fa5dd5364de6a2a7c1d9ed12d9ea438b544d902", "sources": ["arxiv", "semantic_scholar"], "title": "RAI: Flexible Agent Framework for Embodied AI", "abstract": "With an increase in the capabilities of generative language models, a growing interest in embodied AI has followed. This contribution introduces RAI - a framework for creating embodied Multi Agent Systems for robotics. The proposed framework implements tools for Agents' integration with robotic stacks, Large Language Models, and simulations. It provides out-of-the-box integration with state-of-the-art systems like ROS 2. It also comes with dedicated mechanisms for the embodiment of Agents. These mechanisms have been tested on a physical robot, Husarion ROSBot XL, which was coupled with its digital twin, for rapid prototyping. Furthermore, these mechanisms have been deployed in two simulations: (1) robot arm manipulator and (2) tractor controller. All of these deployments have been evaluated in terms of their control capabilities, effectiveness of embodiment, and perception ability. The proposed framework has been used successfully to build systems with multiple agents. It has demonstrated effectiveness in all the aforementioned tasks. It also enabled identifying and addressing the shortcomings of the generative models used for embodied AI.", "authors": ["Kajetan Rachwał", "Maciej Majek", "Bartłomiej Boczek", "Kacper Dąbrowski", "Paweł Liberadzki", "Adam Dąbrowski", "Maria Ganzha"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-12", "url": "https://arxiv.org/abs/2505.07532", "pdf_url": "https://arxiv.org/pdf/2505.07532v1", "arxiv_id": "2505.07532", "doi": "10.48550/arXiv.2505.07532", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "a99130dea6ab73e791405510cf986d2334f343742c07f3a164bdc4a45111476b", "sources": ["arxiv", "semantic_scholar"], "title": "UAV-CodeAgents: Scalable UAV Mission Planning via Multi-Agent ReAct and Vision-Language Reasoning", "abstract": "We present UAV-CodeAgents, a scalable multi-agent framework for autonomous UAV mission generation, built on large language and vision-language models (LLMs/VLMs). The system leverages the ReAct (Reason + Act) paradigm to interpret satellite imagery, ground high-level natural language instructions, and collaboratively generate UAV trajectories with minimal human supervision. A core component is a vision-grounded, pixel-pointing mechanism that enables precise localization of semantic targets on aerial maps. To support real-time adaptability, we introduce a reactive thinking loop, allowing agents to iteratively reflect on observations, revise mission goals, and coordinate dynamically in evolving environments. UAV-CodeAgents is evaluated on large-scale mission scenarios involving industrial and environmental fire detection. Our results show that a lower decoding temperature (0.5) yields higher planning reliability and reduced execution time, with an average mission creation time of 96.96 seconds and a success rate of 93%. We further fine-tune Qwen2.5VL-7B on 9,000 annotated satellite images, achieving strong spatial grounding across diverse visual categories. To foster reproducibility and future research, we will release the full codebase and a novel benchmark dataset for vision-language-based UAV planning.", "authors": ["Oleg Sautenkov", "Yasheerah Yaqoot", "Muhammad Ahsan Mustafa", "Faryal Batool", "Jeffrin Sam", "Artem Lykov", "Chih-Yung Wen", "Dzmitry Tsetserukou"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-12", "url": "https://arxiv.org/abs/2505.07236", "pdf_url": "https://arxiv.org/pdf/2505.07236v1", "arxiv_id": "2505.07236", "doi": "10.48550/arXiv.2505.07236", "citation_count": 15, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3495}
{"id": "f551d88b96915519d0e8d2074463822453b32c81b9a1abde320e2b61a690845f", "sources": ["arxiv", "semantic_scholar"], "title": "ScaleMCP: Dynamic and Auto-Synchronizing Model Context Protocol Tools for LLM Agents", "abstract": "Recent advancements in Large Language Models (LLMs) and the introduction of the Model Context Protocol (MCP) have significantly expanded LLM agents' capability to interact dynamically with external tools and APIs. However, existing tool selection frameworks do not integrate MCP servers, instead relying heavily on error-prone manual updates to monolithic local tool repositories, leading to duplication, inconsistencies, and inefficiencies. Additionally, current approaches abstract tool selection before the LLM agent is invoked, limiting its autonomy and hindering dynamic re-querying capabilities during multi-turn interactions. To address these issues, we introduce ScaleMCP, a novel tool selection approach that dynamically equips LLM agents with a MCP tool retriever, giving agents the autonomy to add tools into their memory, as well as an auto-synchronizing tool storage system pipeline through CRUD (create, read, update, delete) operations with MCP servers as the single source of truth. We also propose a novel embedding strategy, Tool Document Weighted Average (TDWA), designed to selectively emphasize critical components of tool documents (e.g. tool name or synthetic questions) during the embedding process. Comprehensive evaluations conducted on a created dataset of 5,000 financial metric MCP servers, across 10 LLM models, 5 embedding models, and 5 retriever types, demonstrate substantial improvements in tool retrieval and agent invocation performance, emphasizing ScaleMCP's effectiveness in scalable, dynamic tool selection and invocation.", "authors": ["Elias Lumer", "Anmol Gulati", "Vamse Kumar Subbiah", "Pradeep Honaganahalli Basavaraju", "James A. Burke"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-09", "url": "https://arxiv.org/abs/2505.06416", "pdf_url": "https://arxiv.org/pdf/2505.06416v1", "arxiv_id": "2505.06416", "doi": "10.48550/arXiv.2505.06416", "citation_count": 37, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Joint Conference on Computational Intelligence", "quality_score": 0.3949}
{"id": "6d3e8ba276d8eb80b3bfc62dda2458d51b6a21df32584eecbf72742fac302c71", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Systems for Robotic Autonomy with LLMs", "abstract": "Since the advent of Large Language Models (LLMs), various research based on such models have maintained significant academic attention and impact, especially in AI and robotics. In this paper, we propose a multi-agent framework with LLMs to construct an integrated system for robotic task analysis, mechanical design, and path generation. The framework includes three core agents: Task Analyst, Robot Designer, and Reinforcement Learning Designer. Outputs are formatted as multimodal results, such as code files or technical reports, for stronger understandability and usability. To evaluate generalizability comparatively, we conducted experiments with models from both GPT and DeepSeek. Results demonstrate that the proposed system can design feasible robots with control strategies when appropriate task inputs are provided, exhibiting substantial potential for enhancing the efficiency and accessibility of robotic system development in research and industrial applications.", "authors": ["Junhong Chen", "Ziqi Yang", "Haoyuan G Xu", "Dandan Zhang", "George Mylonas"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-09", "url": "https://arxiv.org/abs/2505.05762", "pdf_url": "https://arxiv.org/pdf/2505.05762v1", "arxiv_id": "2505.05762", "doi": "10.1109/CVPRW67362.2025.00403", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2785}
{"id": "f3e2016c9c60971a8ce620e745b4e2e24ee5f892925c0fb12c49c7539ba36383", "sources": ["arxiv", "semantic_scholar"], "title": "Facilitating Trustworthy Human-Agent Collaboration in LLM-based Multi-Agent System oriented Software Engineering", "abstract": "Multi-agent autonomous systems (MAS) are better at addressing challenges that spans across multiple domains than singular autonomous agents. This holds true within the field of software engineering (SE) as well. The state-of-the-art research on MAS within SE focuses on integrating LLMs at the core of autonomous agents to create LLM-based multi-agent autonomous (LMA) systems. However, the introduction of LMA systems into SE brings a plethora of challenges. One of the major challenges is the strategic allocation of tasks between humans and the LMA system in a trustworthy manner. To address this challenge, a RACI-based framework is proposed in this work in progress article, along with implementation guidelines and an example implementation of the framework. The proposed framework can facilitate efficient collaboration, ensure accountability, and mitigate potential risks associated with LLM-driven automation while aligning with the Trustworthy AI guidelines. The future steps for this work delineating the planned empirical validation method are also presented.", "authors": ["Krishna Ronanki"], "categories": ["cs.SE", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-07", "url": "https://arxiv.org/abs/2505.04251", "pdf_url": "https://arxiv.org/pdf/2505.04251v1", "arxiv_id": "2505.04251", "doi": "10.1145/3696630.3728717", "citation_count": 13, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2865}
{"id": "03f4212833c0d193a1f9a334084510fbc82ed1dd4a40b9a3b580e6070aeeb41e", "sources": ["arxiv", "semantic_scholar"], "title": "AgentSGEN: Multi-Agent LLM in the Loop for Semantic Collaboration and GENeration of Synthetic Data", "abstract": "The scarcity of data depicting dangerous situations presents a major obstacle to training AI systems for safety-critical applications, such as construction safety, where ethical and logistical barriers hinder real-world data collection. This creates an urgent need for an end-to-end framework to generate synthetic data that can bridge this gap. While existing methods can produce synthetic scenes, they often lack the semantic depth required for scene simulations, limiting their effectiveness. To address this, we propose a novel multi-agent framework that employs an iterative, in-the-loop collaboration between two agents: an Evaluator Agent, acting as an LLM-based judge to enforce semantic consistency and safety-specific constraints, and an Editor Agent, which generates and refines scenes based on this guidance. Powered by LLM's capabilities to reasoning and common-sense knowledge, this collaborative design produces synthetic images tailored to safety-critical scenarios. Our experiments suggest this design can generate useful scenes based on realistic specifications that address the shortcomings of prior approaches, balancing safety requirements with visual semantics. This iterative process holds promise for delivering robust, aesthetically sound simulations, offering a potential solution to the data scarcity challenge in multimedia safety applications.", "authors": ["Vu Dinh Xuan", "Hao Vo", "David Murphy", "Hoang D. Nguyen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-07", "url": "https://arxiv.org/abs/2505.13466", "pdf_url": "https://arxiv.org/pdf/2505.13466v1", "arxiv_id": "2505.13466", "doi": "10.48550/arXiv.2505.13466", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "2e3fbf7d194a77e4790c8367dfba90cc771336f57677dd32df1e6bca29412c4f", "sources": ["arxiv", "semantic_scholar"], "title": "Assessing and Enhancing the Robustness of LLM-based Multi-Agent Systems Through Chaos Engineering", "abstract": "This study explores the application of chaos engineering to enhance the robustness of Large Language Model-Based Multi-Agent Systems (LLM-MAS) in production-like environments under real-world conditions. LLM-MAS can potentially improve a wide range of tasks, from answering questions and generating content to automating customer support and improving decision-making processes. However, LLM-MAS in production or preproduction environments can be vulnerable to emergent errors or disruptions, such as hallucinations, agent failures, and agent communication failures. This study proposes a chaos engineering framework to proactively identify such vulnerabilities in LLM-MAS, assess and build resilience against them, and ensure reliable performance in critical applications.", "authors": ["Joshua Owotogbe"], "categories": ["cs.MA", "cs.AI", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-06", "url": "https://arxiv.org/abs/2505.03096", "pdf_url": "https://arxiv.org/pdf/2505.03096v1", "arxiv_id": "2505.03096", "doi": "10.1109/CAIN66642.2025.00039", "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2698}
{"id": "c46c3daae023f0f90020ac30bb9e619744645421ca5ead15af6e290745ebb1f5", "sources": ["arxiv", "semantic_scholar"], "title": "The Power of Stories: Narrative Priming Shapes How LLM Agents Collaborate and Compete", "abstract": "According to Yuval Noah Harari, large-scale human cooperation is driven by shared narratives that encode common beliefs and values. This study explores whether such narratives can similarly nudge LLM agents toward collaboration. We use a finitely repeated public goods game in which LLM agents choose either cooperative or egoistic spending strategies. We prime agents with stories highlighting teamwork to different degrees and test how this influences negotiation outcomes. Our experiments explore four questions:(1) How do narratives influence negotiation behavior? (2) What differs when agents share the same story versus different ones? (3) What happens when the agent numbers grow? (4) Are agents resilient against self-serving negotiators? We find that story-based priming significantly affects negotiation strategies and success rates. Common stories improve collaboration, benefiting each agent. By contrast, priming agents with different stories reverses this effect, and those agents primed toward self-interest prevail. We hypothesize that these results carry implications for multi-agent system design and AI alignment.", "authors": ["Gerrit Großmann", "Larisa Ivanova", "Sai Leela Poduru", "Mohaddeseh Tabrizian", "Islam Mesabah", "David A. Selby", "Sebastian J. Vollmer"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-06", "url": "https://arxiv.org/abs/2505.03961", "pdf_url": "https://arxiv.org/pdf/2505.03961v2", "arxiv_id": "2505.03961", "doi": "10.48550/arXiv.2505.03961", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/storyagents25/story-agents", "venue": "arXiv.org", "quality_score": 0.2373}
{"id": "39302dc57a009b3994077c44509f649e9df10010fcec8aae3f3ac94f4af93f04", "sources": ["arxiv", "semantic_scholar"], "title": "DriveAgent: Multi-Agent Structured Reasoning with LLM and Multimodal Sensor Fusion for Autonomous Driving", "abstract": "We introduce DriveAgent, a novel multi-agent autonomous driving framework that leverages large language model (LLM) reasoning combined with multimodal sensor fusion to enhance situational understanding and decision-making. DriveAgent uniquely integrates diverse sensor modalities-including camera, LiDAR, GPS, and IMU-with LLM-driven analytical processes structured across specialized agents. The framework operates through a modular agent-based pipeline comprising four principal modules: (i) a descriptive analysis agent identifying critical sensor data events based on filtered timestamps, (ii) dedicated vehicle-level analysis conducted by LiDAR and vision agents that collaboratively assess vehicle conditions and movements, (iii) environmental reasoning and causal analysis agents explaining contextual changes and their underlying mechanisms, and (iv) an urgency-aware decision-generation agent prioritizing insights and proposing timely maneuvers. This modular design empowers the LLM to effectively coordinate specialized perception and reasoning agents, delivering cohesive, interpretable insights into complex autonomous driving scenarios. Extensive experiments on challenging autonomous driving datasets demonstrate that DriveAgent is achieving superior performance on multiple metrics against baseline methods. These results validate the efficacy of the proposed LLM-driven multi-agent sensor fusion framework, underscoring its potential to substantially enhance the robustness and reliability of autonomous driving systems.", "authors": ["Xinmeng Hou", "Wuqi Wang", "Long Yang", "Hao Lin", "Jinglun Feng", "Haigen Min", "Xiangmo Zhao"], "categories": ["cs.RO", "cs.DB"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-04", "url": "https://arxiv.org/abs/2505.02123", "pdf_url": "https://arxiv.org/pdf/2505.02123v1", "arxiv_id": "2505.02123", "doi": "10.1109/LRA.2025.3619807", "citation_count": 22, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE Robotics and Automation Letters", "quality_score": 0.3404}
{"id": "236de9b6321de9885a099c16f0f5e2ac67eb5bd37d3b3e4971da9f52b1a5d085", "sources": ["arxiv", "semantic_scholar"], "title": "Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents", "abstract": "AI agents are beginning to interact with each other directly and across internet platforms and physical environments, creating security challenges beyond traditional cybersecurity and AI safety frameworks. Free-form protocols are essential for AI's task generalization but enable new threats like secret collusion and coordinated swarm attacks. Network effects can rapidly spread privacy breaches, disinformation, jailbreaks, and data poisoning, while multi-agent dispersion and stealth optimization help adversaries evade oversight - creating novel persistent threats at a systemic level. Despite their critical importance, these security challenges remain understudied, with research fragmented across disparate fields including AI security, multi-agent learning, complex systems, cybersecurity, game theory, distributed systems, and technical AI governance. We introduce multi-agent security, a new field dedicated to securing networks of AI agents against threats that emerge or amplify through their interactions - whether direct or indirect via shared environments - with each other, humans, and institutions, and characterise fundamental security-utility and security-security trade-offs across both distributed and decentralised settings. Our preliminary work (1) taxonomizes the threat landscape arising from interacting AI agents, (2) offers applications to multi-agent security for work across diffuse subfields, and (3) proposes a unified research agenda addressing open challenges in designing secure agent systems and interaction environments. By identifying these gaps, we aim to guide research in this critical area to unlock the socioeconomic potential of large-scale agent deployment, foster public trust, and mitigate national security risks in critical infrastructure and defense contexts.", "authors": ["Christian Schroeder de Witt", "Klaudia Krawiecka", "Igor Krawczuk", "Ben Hagag", "William L. Anderson", "Peter Belcak", "Ben Bucknall", "Xiaohong Cai", "Ayush Chopra", "Doron Cohen", "Ron F. Del Rosario", "Andis Draguns", "Annie Gray", "Keren Katz", "Vasilios Mavroudis", "Jaron Mink", "Sumeet Ramesh Motwani", "Jonathan Petit", "Leif-Sebastian Rembeck", "Chandler Smith", "John Sotiropoulos", "Steven Young", "Sarah Scheffler", "Mary Llewellyn"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-04", "url": "https://arxiv.org/abs/2505.02077", "pdf_url": "https://arxiv.org/pdf/2505.02077v2", "arxiv_id": "2505.02077", "doi": "10.48550/arXiv.2505.02077", "citation_count": 64, "influential_citation_count": 8, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4771}
{"id": "a2a4ac530fe5b7c517b82d8b4916a25bbeb3de0403255ba7a453694b1a0209f2", "sources": ["arxiv", "semantic_scholar"], "title": "Neural Orchestration for Multi-Agent Systems: A Deep Learning Framework for Optimal Agent Selection in Multi-Domain Task Environments", "abstract": "Multi-agent systems (MAS) are foundational in simulating complex real-world scenarios involving autonomous, interacting entities. However, traditional MAS architectures often suffer from rigid coordination mechanisms and difficulty adapting to dynamic tasks. We propose MetaOrch, a neural orchestration framework for optimal agent selection in multi-domain task environments. Our system implements a supervised learning approach that models task context, agent histories, and expected response quality to select the most appropriate agent for each task. A novel fuzzy evaluation module scores agent responses along completeness, relevance, and confidence dimensions, generating soft supervision labels for training the orchestrator. Unlike previous methods that hard-code agent-task mappings, MetaOrch dynamically predicts the most suitable agent while estimating selection confidence. Experiments in simulated environments with heterogeneous agents demonstrate that our approach achieves 86.3% selection accuracy, significantly outperforming baseline strategies including random selection and round-robin scheduling. The modular architecture emphasizes extensibility, allowing agents to be registered, updated, and queried independently. Results suggest that neural orchestration offers a powerful approach to enhancing the autonomy, interpretability, and adaptability of multi-agent systems across diverse task domains.", "authors": ["Kushagra Agrawal", "Nisharg Nargund"], "categories": ["cs.MA", "cs.AI", "cs.NE"], "fields_of_study": ["Computer Science"], "published_date": "2025-05-03", "url": "https://arxiv.org/abs/2505.02861", "pdf_url": "https://arxiv.org/pdf/2505.02861v2", "arxiv_id": "2505.02861", "doi": "10.1007/978-3-032-18477-1_55", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "9143ad186625879dd909b1e078113b36b78e6596e12d78ac8866d8ccbba68aff", "sources": ["arxiv", "semantic_scholar"], "title": "Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems", "abstract": "Failure attribution in LLM multi-agent systems-identifying the agent and step responsible for task failures-provides crucial clues for systems debugging but remains underexplored and labor-intensive. In this paper, we propose and formulate a new research area: automated failure attribution for LLM multi-agent systems. To support this initiative, we introduce the Who&When dataset, comprising extensive failure logs from 127 LLM multi-agent systems with fine-grained annotations linking failures to specific agents and decisive error steps. Using the Who&When, we develop and evaluate three automated failure attribution methods, summarizing their corresponding pros and cons. The best method achieves 53.5% accuracy in identifying failure-responsible agents but only 14.2% in pinpointing failure steps, with some methods performing below random. Even SOTA reasoning models, such as OpenAI o1 and DeepSeek R1, fail to achieve practical usability. These results highlight the task's complexity and the need for further research in this area. Code and dataset are available at https://github.com/mingyin1/Agents_Failure_Attribution", "authors": ["Shaokun Zhang", "Ming Yin", "Jieyu Zhang", "Jiale Liu", "Zhiguang Han", "Jingyang Zhang", "Beibin Li", "Chi Wang", "Huazheng Wang", "Yiran Chen", "Qingyun Wu"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-30", "url": "https://arxiv.org/abs/2505.00212", "pdf_url": "https://arxiv.org/pdf/2505.00212v3", "arxiv_id": "2505.00212", "doi": "10.48550/arXiv.2505.00212", "citation_count": 93, "influential_citation_count": 22, "has_code": true, "code_url": "https://github.com/mingyin1/Agents_Failure_Attribution", "venue": "International Conference on Machine Learning", "quality_score": 0.6809}
{"id": "eb3ba7f3ad2d93d2108d9b0c24c1c6f9e5620c062d0afedf583ad7bfc2fd68a4", "sources": ["arxiv", "semantic_scholar"], "title": "TAMO: Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent with Multi-Modality Observation Data in Cloud-Native Systems", "abstract": "Implementing large language models (LLMs)-driven root cause analysis (RCA) in cloud-native systems has become a key topic of modern software operations and maintenance. However, existing LLM-based approaches face three key challenges: multi-modality input constraint, context window limitation, and dynamic dependence graph. To address these issues, we propose a tool-assisted LLM agent with multi-modality observation data for fine-grained RCA, namely TAMO, including multimodality alignment tool, root cause localization tool, and fault types classification tool. In detail, TAMO unifies multi-modal observation data into time-aligned representations for cross-modal feature consistency. Based on the unified representations, TAMO then invokes its specialized root cause localization tool and fault types classification tool for further identifying root cause and fault type underlying system context. This approach overcomes the limitations of LLMs in processing real-time raw observational data and dynamic service dependencies, guiding the model to generate repair strategies that align with system context through structured prompt design. Experiments on two benchmark datasets demonstrate that TAMO outperforms state-of-the-art (SOTA) approaches with comparable performance.", "authors": ["Xiao Zhang", "Qi Wang", "Mingyi Li", "Yuan Yuan", "Mengbai Xiao", "Fuzhen Zhuang", "Dongxiao Yu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-29", "url": "https://arxiv.org/abs/2504.20462", "pdf_url": "https://arxiv.org/pdf/2504.20462v5", "arxiv_id": "2504.20462", "doi": "10.1109/TSC.2025.3629066", "citation_count": 15, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Services Computing", "quality_score": 0.301}
{"id": "4e4e6e3a1f81703fc969bfb34d2cde00681ed03318de2e3d5999d7bbc1088812", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Reasoning and Tool Integration for LLMs via Reinforcement Learning", "abstract": "Large language models (LLMs) have achieved remarkable progress in complex reasoning tasks, yet they remain fundamentally limited by their reliance on static internal knowledge and text-only reasoning. Real-world problem solving often demands dynamic, multi-step reasoning, adaptive decision making, and the ability to interact with external tools and environments. In this work, we introduce ARTIST (Agentic Reasoning and Tool Integration in Self-improving Transformers), a unified framework that tightly couples agentic reasoning, reinforcement learning, and tool integration for LLMs. ARTIST enables models to autonomously decide when, how, and which tools to invoke within multi-turn reasoning chains, leveraging outcome-based RL to learn robust strategies for tool use and environment interaction without requiring step-level supervision. Extensive experiments on mathematical reasoning and multi-turn function calling benchmarks show that ARTIST consistently outperforms state-of-the-art baselines, with up to 22% absolute improvement over base models and strong gains on the most challenging tasks. Detailed studies and metric analyses reveal that agentic RL training leads to deeper reasoning, more effective tool use, and higher-quality solutions. Our results establish agentic RL with tool integration as a powerful new frontier for robust, interpretable, and generalizable problem-solving in LLMs.", "authors": ["Joykirat Singh", "Raghav Magazine", "Yash Pandya", "Akshay Nambi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2505.01441", "pdf_url": "https://arxiv.org/pdf/2505.01441v1", "arxiv_id": "2505.01441", "doi": "10.48550/arXiv.2505.01441", "citation_count": 83, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4811}
{"id": "f0d1a08724567057a36328bc64c0715a8b7771fbc17fbe38742a45bb62b33b13", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Powered GUI Agents in Phone Automation: Surveying Progress and Prospects", "abstract": "With the rapid rise of large language models (LLMs), phone automation has undergone transformative changes. This paper systematically reviews LLM-driven phone GUI agents, highlighting their evolution from script-based automation to intelligent, adaptive systems. We first contextualize key challenges, (i) limited generality, (ii) high maintenance overhead, and (iii) weak intent comprehension, and show how LLMs address these issues through advanced language understanding, multimodal perception, and robust decision-making. We then propose a taxonomy covering fundamental agent frameworks (single-agent, multi-agent, plan-then-act), modeling approaches (prompt engineering, training-based), and essential datasets and benchmarks. Furthermore, we detail task-specific architectures, supervised fine-tuning, and reinforcement learning strategies that bridge user intent and GUI operations. Finally, we discuss open challenges such as dataset diversity, on-device deployment efficiency, user-centric adaptation, and security concerns, offering forward-looking insights into this rapidly evolving field. By providing a structured overview and identifying pressing research gaps, this paper serves as a definitive reference for researchers and practitioners seeking to harness LLMs in designing scalable, user-friendly phone GUI agents. The collection of papers reviewed in this survey will be hosted and regularly updated on the GitHub repository: https://github.com/PhoneLLM/Awesome-LLM-Powered-Phone-GUI-Agents", "authors": ["Guangyi Liu", "Pengxiang Zhao", "Yaozhen Liang", "Liang Liu", "Yaxuan Guo", "Han Xiao", "Weifeng Lin", "Yuxiang Chai", "Yue Han", "Shuai Ren", "Hao Wang", "Xiaoyu Liang", "WenHao Wang", "Tianze Wu", "Zhengxi Lu", "Siheng Chen", " LiLinghao", "Hao Wang", "Guanjing Xiong", "Yong Liu", "Hongsheng Li"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.19838", "pdf_url": "https://arxiv.org/pdf/2504.19838v3", "arxiv_id": "2504.19838", "doi": "10.48550/arXiv.2504.19838", "citation_count": 31, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/PhoneLLM/Awesome-LLM-Powered-Phone-GUI-Agents", "venue": null, "quality_score": 0.3763}
{"id": "fc0f07c0e136bc28ef5cd7272aa77b150cfbc37fc16795725d8e09e99cf2cd8e", "sources": ["arxiv", "semantic_scholar"], "title": "Securing GenAI Multi-Agent Systems Against Tool Squatting: A Zero Trust Registry-Based Approach", "abstract": "The rise of generative AI (GenAI) multi-agent systems (MAS) necessitates standardized protocols enabling agents to discover and interact with external tools. However, these protocols introduce new security challenges, particularly; tool squatting; the deceptive registration or representation of tools. This paper analyzes tool squatting threats within the context of emerging interoperability standards, such as Model Context Protocol (MCP) or seamless communication between agents protocols. It introduces a comprehensive Tool Registry system designed to mitigate these risks. We propose a security-focused architecture featuring admin-controlled registration, centralized tool discovery, fine grained access policies enforced via dedicated Agent and Tool Registry services, a dynamic trust scoring mechanism based on tool versioning and known vulnerabilities, and just in time credential provisioning. Based on its design principles, the proposed registry framework aims to effectively prevent common tool squatting vectors while preserving the flexibility and power of multi-agent systems. This work addresses a critical security gap in the rapidly evolving GenAI ecosystem and provides a foundation for secure tool integration in production environments.", "authors": ["Vineeth Sai Narajala", "Ken Huang", "Idan Habler"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.19951", "pdf_url": "https://arxiv.org/pdf/2504.19951v1", "arxiv_id": "2504.19951", "doi": "10.1109/AIxDKE67294.2026.00022", "citation_count": 23, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3451}
{"id": "1d08dc9a8811e6b51b74bcbbdb1734d9505184906cec5dab61537e321cb43fa9", "sources": ["arxiv", "semantic_scholar"], "title": "Evolution of Cooperation in LLM-Agent Societies: A Preliminary Study Using Different Punishment Strategies", "abstract": "The evolution of cooperation has been extensively studied using abstract mathematical models and simulations. Recent advances in Large Language Models (LLMs) and the rise of LLM agents have demonstrated their ability to perform social reasoning, thus providing an opportunity to test the emergence of norms in more realistic agent-based simulations with human-like reasoning using natural language. In this research, we investigate whether the cooperation dynamics presented in Boyd and Richerson's model persist in a more realistic simulation of the Diner's Dilemma using LLM agents compared to the abstract mathematical nature in the work of Boyd and Richerson. Our findings indicate that agents follow the strategies defined in the Boyd and Richerson model, and explicit punishment mechanisms drive norm emergence, reinforcing cooperative behaviour even when the agent strategy configuration varies. Our results suggest that LLM-based Multi-Agent System simulations, in fact, can replicate the evolution of cooperation predicted by the traditional mathematical models. Moreover, our simulations extend beyond the mathematical models by integrating natural language-driven reasoning and a pairwise imitation method for strategy adoption, making them a more realistic testbed for cooperative behaviour in MASs.", "authors": ["Kavindu Warnakulasuriya", "Prabhash Dissanayake", "Navindu De Silva", "Stephen Cranefield", "Bastin Tony Roy Savarimuthu", "Surangika Ranathunga", "Nisansa de Silva"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.19487", "pdf_url": "https://arxiv.org/pdf/2504.19487v3", "arxiv_id": "2504.19487", "doi": "10.48550/arXiv.2504.19487", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "dac2e17e8e9e4627bef6cd6082edb08b9936faba2075aad88845efb50ca61f53", "sources": ["arxiv", "semantic_scholar"], "title": "From LLM Reasoning to Autonomous AI Agents: A Comprehensive Review", "abstract": "Large language models and autonomous AI agents have evolved rapidly, resulting in a diverse array of evaluation benchmarks, frameworks, and collaboration protocols. Driven by the growing need for standardized evaluation and integration, we systematically consolidate these fragmented efforts into a unified framework. However, the landscape remains fragmented and lacks a unified taxonomy or comprehensive survey. Therefore, we present a side-by-side comparison of benchmarks developed between 2019 and 2025 that evaluate these models and agents across multiple domains. In addition, we propose a taxonomy of approximately 60 benchmarks that cover general and academic knowledge reasoning, mathematical problem-solving, code generation and software engineering, factual grounding and retrieval, domain-specific evaluations, multimodal and embodied tasks, task orchestration, and interactive assessments. Furthermore, we review AI-agent frameworks introduced between 2023 and 2025 that integrate large language models with modular toolkits to enable autonomous decision-making and multi-step reasoning. Moreover, we present real-world applications of autonomous AI agents in materials science, biomedical research, academic ideation, software engineering, synthetic data generation, chemical reasoning, mathematical problem-solving, geographic information systems, multimedia, healthcare, and finance. We then survey key agent-to-agent collaboration protocols, namely the Agent Communication Protocol (ACP), the Model Context Protocol (MCP), and the Agent-to-Agent Protocol (A2A). Finally, we discuss recommendations for future research, focusing on advanced reasoning strategies, failure modes in multi-agent LLM systems, automated scientific discovery, dynamic tool integration via reinforcement learning, integrated search capabilities, and security vulnerabilities in agent protocols.", "authors": ["Mohamed Amine Ferrag", "Norbert Tihanyi", "Merouane Debbah"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.19678", "pdf_url": "https://arxiv.org/pdf/2504.19678v2", "arxiv_id": "2504.19678", "doi": "10.1109/ACCESS.2026.3698694", "citation_count": 162, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "IEEE Access", "quality_score": 0.553}
{"id": "211c643dbf3eba3a34c55c4fb657deed236e04b582015c913a0c9fcaee1b1ca4", "sources": ["arxiv", "semantic_scholar"], "title": "ResearchCodeAgent: An LLM Multi-Agent System for Automated Codification of Research Methodologies", "abstract": "In this paper we introduce ResearchCodeAgent, a novel multi-agent system leveraging large language models (LLMs) agents to automate the codification of research methodologies described in machine learning literature. The system bridges the gap between high-level research concepts and their practical implementation, allowing researchers auto-generating code of existing research papers for benchmarking or building on top-of existing methods specified in the literature with availability of partial or complete starter code. ResearchCodeAgent employs a flexible agent architecture with a comprehensive action suite, enabling context-aware interactions with the research environment. The system incorporates a dynamic planning mechanism, utilizing both short and long-term memory to adapt its approach iteratively. We evaluate ResearchCodeAgent on three distinct machine learning tasks with distinct task complexity and representing different parts of the ML pipeline: data augmentation, optimization, and data batching. Our results demonstrate the system's effectiveness and generalizability, with 46.9% of generated code being high-quality and error-free, and 25% showing performance improvements over baseline implementations. Empirical analysis shows an average reduction of 57.9% in coding time compared to manual implementation. We observe higher gains for more complex tasks. ResearchCodeAgent represents a significant step towards automating the research implementation process, potentially accelerating the pace of machine learning research.", "authors": ["Shubham Gandhi", "Dhruv Shah", "Manasi Patwardhan", "Lovekesh Vig", "Gautam Shroff"], "categories": ["cs.SE", "cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.20117", "pdf_url": "https://arxiv.org/pdf/2504.20117v2", "arxiv_id": "2504.20117", "doi": "10.48550/arXiv.2504.20117", "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2698}
{"id": "ba65355038d11b228b9de4cc5b02f4efaf5d42f350d1bf95b158a3ae88d2da42", "sources": ["arxiv", "semantic_scholar"], "title": "Prompt Injection Attack to Tool Selection in LLM Agents", "abstract": "Tool selection is a key component of LLM agents. A popular approach follows a two-step process - \\emph{retrieval} and \\emph{selection} - to pick the most appropriate tool from a tool library for a given task. In this work, we introduce \\textit{ToolHijacker}, a novel prompt injection attack targeting tool selection in no-box scenarios. ToolHijacker injects a malicious tool document into the tool library to manipulate the LLM agent's tool selection process, compelling it to consistently choose the attacker's malicious tool for an attacker-chosen target task. Specifically, we formulate the crafting of such tool documents as an optimization problem and propose a two-phase optimization strategy to solve it. Our extensive experimental evaluation shows that ToolHijacker is highly effective, significantly outperforming existing manual-based and automated prompt injection attacks when applied to tool selection. Moreover, we explore various defenses, including prevention-based defenses (StruQ and SecAlign) and detection-based defenses (known-answer detection, DataSentinel, perplexity detection, and perplexity windowed detection). Our experimental results indicate that these defenses are insufficient, highlighting the urgent need for developing new defense strategies.", "authors": ["Jiawen Shi", "Zenghui Yuan", "Guiyao Tie", "Pan Zhou", "Neil Zhenqiang Gong", "Lichao Sun"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-28", "url": "https://arxiv.org/abs/2504.19793", "pdf_url": "https://arxiv.org/pdf/2504.19793v3", "arxiv_id": "2504.19793", "doi": "10.48550/arXiv.2504.19793", "citation_count": 89, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.4886}
{"id": "8ce025d7cceabdfa4b204814779fdf944ca5fed1885c80d31e60483b92d6e425", "sources": ["arxiv", "semantic_scholar"], "title": "A Vision for Auto Research with LLM Agents", "abstract": "This paper introduces Agent-Based Auto Research, a structured multi-agent framework designed to automate, coordinate, and optimize the full lifecycle of scientific research. Leveraging the capabilities of large language models (LLMs) and modular agent collaboration, the system spans all major research phases, including literature review, ideation, methodology planning, experimentation, paper writing, peer review response, and dissemination. By addressing issues such as fragmented workflows, uneven methodological expertise, and cognitive overload, the framework offers a systematic and scalable approach to scientific inquiry. Preliminary explorations demonstrate the feasibility and potential of Auto Research as a promising paradigm for self-improving, AI-driven research processes.", "authors": ["Chengwei Liu", "Chong Wang", "Jiayue Cao", "Jingquan Ge", "Kun Wang", "Lyuye Zhang", "Ming-Ming Cheng", "Penghai Zhao", "Tianlin Li", "Xiaojun Jia", "Xiang Li", "Xingshuai Li", "Yang Liu", "Yebo Feng", "Yihao Huang", "Yijia Xu", "Yuqiang Sun", "Zhenhong Zhou", "Zhengzi Xu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-26", "url": "https://arxiv.org/abs/2504.18765", "pdf_url": "https://arxiv.org/pdf/2504.18765v3", "arxiv_id": "2504.18765", "doi": "10.48550/arXiv.2504.18765", "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "8275029e6edb86f1a59417a6a907c1e2e1f46d437856b38ed64e84c870f90652", "sources": ["arxiv", "semantic_scholar"], "title": "Comprehend, Divide, and Conquer: Feature Subspace Exploration via Multi-Agent Hierarchical Reinforcement Learning", "abstract": "Feature selection aims to preprocess the target dataset, find an optimal and most streamlined feature subset, and enhance the downstream machine learning task. Among filter, wrapper, and embedded-based approaches, the reinforcement learning (RL)-based subspace exploration strategy provides a novel objective optimization-directed perspective and promising performance. Nevertheless, even with improved performance, current reinforcement learning approaches face challenges similar to conventional methods when dealing with complex datasets. These challenges stem from the inefficient paradigm of using one agent per feature and the inherent complexities present in the datasets. This observation motivates us to investigate and address the above issue and propose a novel approach, namely HRLFS. Our methodology initially employs a Large Language Model (LLM)-based hybrid state extractor to capture each feature's mathematical and semantic characteristics. Based on this information, features are clustered, facilitating the construction of hierarchical agents for each cluster and sub-cluster. Extensive experiments demonstrate the efficiency, scalability, and robustness of our approach. Compared to contemporary or the one-feature-one-agent RL-based approaches, HRLFS improves the downstream ML performance with iterative feature subspace exploration while accelerating total run time by reducing the number of agents involved.", "authors": ["Weiliang Zhang", "Xiaohan Huang", "Yi Du", "Ziyue Qiao", "Qingqing Long", "Zhen Meng", "Yuanchun Zhou", "Meng Xiao"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-24", "url": "https://arxiv.org/abs/2504.17356", "pdf_url": "https://arxiv.org/pdf/2504.17356v2", "arxiv_id": "2504.17356", "doi": "10.48550/arXiv.2504.17356", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1398}
{"id": "2d9fa4da3f0d9db31bebf1d01822c25d31292636b3258bb6331b78c80f82e52e", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborating Action by Action: A Multi-agent LLM Framework for Embodied Reasoning", "abstract": "Collaboration is ubiquitous and essential in day-to-day life -- from exchanging ideas, to delegating tasks, to generating plans together. This work studies how LLMs can adaptively collaborate to perform complex embodied reasoning tasks. To this end we introduce MINDcraft, an easily extensible platform built to enable LLM agents to control characters in the open-world game of Minecraft; and MineCollab, a benchmark to test the different dimensions of embodied and collaborative reasoning. An experimental study finds that the primary bottleneck in collaborating effectively for current state-of-the-art agents is efficient natural language communication, with agent performance dropping as much as 15% when they are required to communicate detailed task completion plans. We conclude that existing LLM agents are ill-optimized for multi-agent collaboration, especially in embodied scenarios, and highlight the need to employ methods beyond in-context and imitation learning. Our website can be found here: https://mindcraft-minecollab.github.io/", "authors": ["Isadora White", "Kolby Nottingham", "Ayush Maniar", "Max Robinson", "Hansen Lillemark", "Mehul Maheshwari", "Lianhui Qin", "Prithviraj Ammanabrolu"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-24", "url": "https://arxiv.org/abs/2504.17950", "pdf_url": "https://arxiv.org/pdf/2504.17950v1", "arxiv_id": "2504.17950", "doi": "10.48550/arXiv.2504.17950", "citation_count": 17, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3138}
{"id": "b43d1add8cb93a667161096f5c97ad2cd21569e0dba4a4fd092f73a1ccb8c58e", "sources": ["arxiv", "semantic_scholar"], "title": "A RAG-Based Multi-Agent LLM System for Natural Hazard Resilience and Adaptation", "abstract": "Large language models (LLMs) are a transformational capability at the frontier of artificial intelligence and machine learning that can support decision-makers in addressing pressing societal challenges such as extreme natural hazard events. As generalized models, LLMs often struggle to provide context-specific information, particularly in areas requiring specialized knowledge. In this work we propose a retrieval-augmented generation (RAG)-based multi-agent LLM system to support analysis and decision-making in the context of natural hazards and extreme weather events. As a proof of concept, we present WildfireGPT, a specialized system focused on wildfire hazards. The architecture employs a user-centered, multi-agent design to deliver tailored risk insights across diverse stakeholder groups. By integrating natural hazard and extreme weather projection data, observational datasets, and scientific literature through an RAG framework, the system ensures both the accuracy and contextual relevance of the information it provides. Evaluation across ten expert-led case studies demonstrates that WildfireGPT significantly outperforms existing LLM-based solutions for decision support.", "authors": ["Yangxinyu Xie", "Bowen Jiang", "Tanwi Mallick", "Joshua David Bergerson", "John K. Hutchison", "Duane R. Verner", "Jordan Branham", "M. Ross Alexander", "Robert B. Ross", "Yan Feng", "Leslie-Anne Levy", "Weijie Su", "Camillo J. Taylor"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-24", "url": "https://arxiv.org/abs/2504.17200", "pdf_url": "https://arxiv.org/pdf/2504.17200v1", "arxiv_id": "2504.17200", "doi": "10.48550/arXiv.2504.17200", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2603}
{"id": "d3d7a3eb7d844393e5eef06695b3ada36b1de46281d6b7028156a5ce97ccfdb7", "sources": ["arxiv", "semantic_scholar"], "title": "RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning", "abstract": "Training large language models (LLMs) as interactive agents presents unique challenges including long-horizon decision making and interacting with stochastic environment feedback. While reinforcement learning (RL) has enabled progress in static tasks, multi-turn agent RL training remains underexplored. We propose StarPO (State-Thinking-Actions-Reward Policy Optimization), a general framework for trajectory-level agent RL, and introduce RAGEN, a modular system for training and evaluating LLM agents. Our study on four stylized environments reveals three core findings. First, our agent RL training shows a recurring mode of Echo Trap where reward variance cliffs and gradient spikes; we address this with StarPO-S, a stabilized variant with trajectory filtering, critic incorporation, and gradient stabilization. Second, we find the shaping of RL rollouts would benefit from diverse initial states, medium interaction granularity and more frequent sampling. Third, we show that without fine-grained, reasoning-aware reward signals, agent reasoning hardly emerge through multi-turn RL and they may show shallow strategies or hallucinated thoughts. Code and environments are available at https://github.com/RAGEN-AI/RAGEN.", "authors": ["Zihan Wang", "Kangrui Wang", "Qineng Wang", "Pingyue Zhang", "Linjie Li", "Zhengyuan Yang", "Xing Jin", "Kefan Yu", "Minh Nhat Nguyen", "Licheng Liu", "Eli Gottlieb", "Yiping Lu", "Kyunghyun Cho", "Jiajun Wu", "Li Fei-Fei", "Lijuan Wang", "Yejin Choi", "Manling Li"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-24", "url": "https://arxiv.org/abs/2504.20073", "pdf_url": "https://arxiv.org/pdf/2504.20073v2", "arxiv_id": "2504.20073", "doi": "10.48550/arXiv.2504.20073", "citation_count": 231, "influential_citation_count": 14, "has_code": true, "code_url": "https://github.com/RAGEN-AI/RAGEN", "venue": "arXiv.org", "quality_score": 0.5914}
{"id": "0be7f7bc172c222ddabc85c319b020800c0743738b5b8cd737b0fc25804490d6", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agent Swarm for Hypothesis-Driven Drug Discovery", "abstract": "Drug discovery remains a formidable challenge: more than 90 percent of candidate molecules fail in clinical evaluation, and development costs often exceed one billion dollars per approved therapy. Disparate data streams, from genomics and transcriptomics to chemical libraries and clinical records, hinder coherent mechanistic insight and slow progress. Meanwhile, large language models excel at reasoning and tool integration but lack the modular specialization and iterative memory required for regulated, hypothesis-driven workflows. We introduce PharmaSwarm, a unified multi-agent framework that orchestrates specialized LLM \"agents\" to propose, validate, and refine hypotheses for novel drug targets and lead compounds. Each agent accesses dedicated functionality--automated genomic and expression analysis; a curated biomedical knowledge graph; pathway enrichment and network simulation; interpretable binding affinity prediction--while a central Evaluator LLM continuously ranks proposals by biological plausibility, novelty, in silico efficacy, and safety. A shared memory layer captures validated insights and fine-tunes underlying submodels over time, yielding a self-improving system. Deployable on low-code platforms or Kubernetes-based microservices, PharmaSwarm supports literature-driven discovery, omics-guided target identification, and market-informed repurposing. We also describe a rigorous four-tier validation pipeline spanning retrospective benchmarking, independent computational assays, experimental testing, and expert user studies to ensure transparency, reproducibility, and real-world impact. By acting as an AI copilot, PharmaSwarm can accelerate translational research and deliver high-confidence hypotheses more efficiently than traditional pipelines.", "authors": ["Kevin Song", "Andrew Trotter", "Jake Y. Chen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-24", "url": "https://arxiv.org/abs/2504.17967", "pdf_url": "https://arxiv.org/pdf/2504.17967v1", "arxiv_id": "2504.17967", "doi": "10.48550/arXiv.2504.17967", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "5fdb080162546e09af87cae0584e986636fd10f9c191300ece9132c379c288d1", "sources": ["arxiv", "semantic_scholar"], "title": "Leveraging LLMs as Meta-Judges: A Multi-Agent Framework for Evaluating LLM Judgments", "abstract": "Large language models (LLMs) are being widely applied across various fields, but as tasks become more complex, evaluating their responses is increasingly challenging. Compared to human evaluators, the use of LLMs to support performance evaluation offers a more efficient alternative. However, most studies focus mainly on aligning LLMs' judgments with human preferences, overlooking the existence of biases and mistakes in human judgment. Furthermore, how to select suitable LLM judgments given multiple potential LLM responses remains underexplored. To address these two aforementioned issues, we propose a three-stage meta-judge selection pipeline: 1) developing a comprehensive rubric with GPT-4 and human experts, 2) using three advanced LLM agents to score judgments, and 3) applying a threshold to filter out low-scoring judgments. Compared to methods using a single LLM as both judge and meta-judge, our pipeline introduces multi-agent collaboration and a more comprehensive rubric. Experimental results on the JudgeBench dataset show about 15.55\\% improvement compared to raw judgments and about 8.37\\% improvement over the single-agent baseline. Our work demonstrates the potential of LLMs as meta-judges and lays the foundation for future research on constructing preference datasets for LLM-as-a-judge reinforcement learning.", "authors": ["Yuran Li", "Jama Hussein Mohamud", "Chongren Sun", "Di Wu", "Benoit Boulet"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-23", "url": "https://arxiv.org/abs/2504.17087", "pdf_url": "https://arxiv.org/pdf/2504.17087v1", "arxiv_id": "2504.17087", "doi": "10.48550/arXiv.2504.17087", "citation_count": 15, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "13b03526aa6ab8f29d7b8725ceae6cafbffeaa3cbc1a1f69631f999efab5cb22", "sources": ["arxiv", "semantic_scholar"], "title": "Amplified Vulnerabilities: Structured Jailbreak Attacks on LLM-based Multi-Agent Debate", "abstract": "Multi-Agent Debate (MAD), leveraging collaborative interactions among Large Language Models (LLMs), aim to enhance reasoning capabilities in complex tasks. However, the security implications of their iterative dialogues and role-playing characteristics, particularly susceptibility to jailbreak attacks eliciting harmful content, remain critically underexplored. This paper systematically investigates the jailbreak vulnerabilities of four prominent MAD frameworks built upon leading commercial LLMs (GPT-4o, GPT-4, GPT-3.5-turbo, and DeepSeek) without compromising internal agents. We introduce a novel structured prompt-rewriting framework specifically designed to exploit MAD dynamics via narrative encapsulation, role-driven escalation, iterative refinement, and rhetorical obfuscation. Our extensive experiments demonstrate that MAD systems are inherently more vulnerable than single-agent setups. Crucially, our proposed attack methodology significantly amplifies this fragility, increasing average harmfulness from 28.14% to 80.34% and achieving attack success rates as high as 80% in certain scenarios. These findings reveal intrinsic vulnerabilities in MAD architectures and underscore the urgent need for robust, specialized defenses prior to real-world deployment.", "authors": ["Senmao Qi", "Yifei Zou", "Peng Li", "Ziyi Lin", "Xiuzhen Cheng", "Dongxiao Yu"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-23", "url": "https://arxiv.org/abs/2504.16489", "pdf_url": "https://arxiv.org/pdf/2504.16489v1", "arxiv_id": "2504.16489", "doi": "10.48550/arXiv.2504.16489", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "c499663043d4ad22cc9b2f77da82806b21b01e6ccfc88648fa737796dd7389d4", "sources": ["arxiv", "semantic_scholar"], "title": "MARFT: Multi-Agent Reinforcement Fine-Tuning", "abstract": "Large Language Model (LLM)-based Multi-Agent Systems (LaMAS) have demonstrated strong capabilities on complex agentic tasks requiring multifaceted reasoning and collaboration, from high-quality presentation generation to scientific research. Meanwhile, Reinforcement Learning (RL) is widely recognized for enhancing agent intelligence, but limited work has studied fine-tuning LaMAS with foundational RL techniques. Directly applying conventional Multi-Agent Reinforcement Learning (MARL) to LaMAS also introduces major challenges due to the unique mechanisms of LaMAS. To address these challenges, this article presents a comprehensive study of LLM-based MARL and proposes Multi-Agent Reinforcement Fine-Tuning (MARFT). We introduce Flex-MG, a new Markov Game formulation aligned with real-world LaMAS optimization, together with a universal algorithmic framework tailored to LaMAS. We review the evolution from traditional RL to Reinforcement Fine-Tuning (RFT), then analyze the multi-agent counterpart. For LaMAS, we identify key differences between classical MARL and MARFT, including asynchronous agent interactions, profile-aware agent design, and heterogeneous architectures. These differences motivate a LaMAS-oriented formulation of RFT. We present a robust and scalable MARFT framework, detail its modular algorithm, and provide an open-source implementation to support adoption and further research. The paper further discusses application perspectives and open challenges, including dynamic environment modeling, sample inefficiency, and the lack of cohesive frameworks. By connecting theoretical foundations with practical methodology, this work aims to serve as a roadmap for advancing MARFT toward resilient, adaptive, and human-aligned agentic systems. Implementation: https://github.com/jwliao-ai/MARFT.", "authors": ["Junwei Liao", "Muning Wen", "Jun Wang", "Weinan Zhang"], "categories": ["cs.MA", "cs.AI", "cs.LG", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-21", "url": "https://arxiv.org/abs/2504.16129", "pdf_url": "https://arxiv.org/pdf/2504.16129v5", "arxiv_id": "2504.16129", "doi": "10.48550/arXiv.2504.16129", "citation_count": 35, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/jwliao-ai/MARFT", "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "65ac52a008ef9212d6a1828e9a04aeac59b0298431c70f57d0d2b8863dfe1770", "sources": ["arxiv", "semantic_scholar"], "title": "Meta-Thinking in LLMs via Multi-Agent Reinforcement Learning: A Survey", "abstract": "This survey explores the development of meta-thinking capabilities in Large Language Models (LLMs) from a Multi-Agent Reinforcement Learning (MARL) perspective. Meta-thinking self-reflection, assessment, and control of thinking processes is an important next step in enhancing LLM reliability, flexibility, and performance, particularly for complex or high-stakes tasks. The survey begins by analyzing current LLM limitations, such as hallucinations and the lack of internal self-assessment mechanisms. It then talks about newer methods, including RL from human feedback (RLHF), self-distillation, and chain-of-thought prompting, and each of their limitations. The crux of the survey is to talk about how multi-agent architectures, namely supervisor-agent hierarchies, agent debates, and theory of mind frameworks, can emulate human-like introspective behavior and enhance LLM robustness. By exploring reward mechanisms, self-play, and continuous learning methods in MARL, this survey gives a comprehensive roadmap to building introspective, adaptive, and trustworthy LLMs. Evaluation metrics, datasets, and future research avenues, including neuroscience-inspired architectures and hybrid symbolic reasoning, are also discussed.", "authors": ["Ahsan Bilal", "Muhammad Ahmed Mohsin", "Muhammad Umer", "Muhammad Awais Khan Bangash", "Muhammad Ali Jamshed"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-20", "url": "https://arxiv.org/abs/2504.14520", "pdf_url": "https://arxiv.org/pdf/2504.14520v1", "arxiv_id": "2504.14520", "doi": "10.48550/arXiv.2504.14520", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "a176b9d2ab3e80fe99a43e13e1300742421b6ba74f8145ada51567e1fa105011", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM-enabled Multi-Agent Autonomous Mechatronics Design Framework", "abstract": "Existing LLM-enabled multi-agent frameworks are predominantly limited to digital or simulated environments and confined to narrowly focused knowledge domain, constraining their applicability to complex engineering tasks that require the design of physical embodiment, cross-disciplinary integration, and constraint-aware reasoning. This work proposes a multi-agent autonomous mechatronics design framework, integrating expertise across mechanical design, optimization, electronics, and software engineering to autonomously generate functional prototypes with minimal direct human design input. Operating primarily through a language-driven workflow, the framework incorporates structured human feedback to ensure robust performance under real-world constraints. To validate its capabilities, the framework is applied to a real-world challenge involving autonomous water-quality monitoring and sampling, where traditional methods are labor-intensive and ecologically disruptive. Leveraging the proposed system, a fully functional autonomous vessel was developed with optimized propulsion, cost-effective electronics, and advanced control. The design process was carried out by specialized agents, including a high-level planning agent responsible for problem abstraction and dedicated agents for structural, electronics, control, and software development. This approach demonstrates the potential of LLM-based multi-agent systems to automate real-world engineering workflows and reduce reliance on extensive domain expertise.", "authors": ["Zeyu Wang", "Frank P. -W. Lo", "Qian Chen", "Yongqi Zhang", "Chen Lin", "Xu Chen", "Zhenhua Yu", "Alexander J. Thompson", "Eric M. Yeatman", "Benny P. L. Lo"], "categories": ["cs.RO", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-20", "url": "https://arxiv.org/abs/2504.14681", "pdf_url": "https://arxiv.org/pdf/2504.14681v1", "arxiv_id": "2504.14681", "doi": "10.1109/CVPRW67362.2025.00404", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2113}
{"id": "9ed817f7c5e2c9dc5473223ea763693ce604a298704557c88bc094fc2246e189", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing LLM-based Quantum Code Generation with Multi-Agent Optimization and Quantum Error Correction", "abstract": "Multi-agent frameworks with Large Language Models (LLMs) have become promising tools for generating general-purpose programming languages using test-driven development, allowing developers to create more accurate and robust code. However, their potential has not been fully unleashed for domain-specific programming languages, where specific domain exhibits unique optimization opportunities for customized improvement. In this paper, we take the first step in exploring multi-agent code generation for quantum programs. By identifying the unique optimizations in quantum designs such as quantum error correction, we introduce a novel multi-agent framework tailored to generating accurate, fault-tolerant quantum code. Each agent in the framework focuses on distinct optimizations, iteratively refining the code using a semantic analyzer with multi-pass inference, alongside an error correction code decoder. We also examine the effectiveness of inference-time techniques, like Chain-of-Thought (CoT) and Retrieval-Augmented Generation (RAG) in the context of quantum programming, uncovering observations that are different from general-purpose code generation. To evaluate our approach, we develop a test suite to measure the impact each optimization has on the accuracy of the generated code. Our findings indicate that techniques such as structured CoT significantly improve the generation of quantum algorithms by up to 50%. In contrast, we have also found that certain techniques such as RAG show limited improvement, yielding an accuracy increase of only 4%. Moreover, we showcase examples of AI-assisted quantum error prediction and correction, demonstrating the effectiveness of our multi-agent framework in reducing the quantum errors of generated quantum programs.", "authors": ["Charlie Campbell", "Hao Mark Chen", "Wayne Luk", "Hongxiang Fan"], "categories": ["quant-ph", "cs.MA"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-04-20", "url": "https://arxiv.org/abs/2504.14557", "pdf_url": "https://arxiv.org/pdf/2504.14557v2", "arxiv_id": "2504.14557", "doi": "10.1109/DAC63849.2025.11133316", "citation_count": 15, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Design Automation Conference", "quality_score": 0.301}
{"id": "3c5269eaba2cc7967610a04c95a77fd700a87ac5277807b1ef065ec46d29d762", "sources": ["arxiv", "semantic_scholar"], "title": "The Athenian Academy: A Seven-Layer Architecture Model for Multi-Agent Systems", "abstract": "This paper proposes the \"Academy of Athens\" multi-agent seven-layer framework, aimed at systematically addressing challenges in multi-agent systems (MAS) within artificial intelligence (AI) art creation, such as collaboration efficiency, role allocation, environmental adaptation, and task parallelism. The framework divides MAS into seven layers: multi-agent collaboration, single-agent multi-role playing, single-agent multi-scene traversal, single-agent multi-capability incarnation, different single agents using the same large model to achieve the same target agent, single-agent using different large models to achieve the same target agent, and multi-agent synthesis of the same target agent. Through experimental validation in art creation, the framework demonstrates its unique advantages in task collaboration, cross-scene adaptation, and model fusion. This paper further discusses current challenges such as collaboration mechanism optimization, model stability, and system security, proposing future exploration through technologies like meta-learning and federated learning. The framework provides a structured methodology for multi-agent collaboration in AI art creation and promotes innovative applications in the art field.", "authors": ["Lidong Zhai", "Zhijie Qiu", "Lvyang Zhang", "Jiaqi Li", "Yi Wang", "Wen Lu", "Xizhong Guo", "Ge Sun"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-17", "url": "https://arxiv.org/abs/2504.12735", "pdf_url": "https://arxiv.org/pdf/2504.12735v2", "arxiv_id": "2504.12735", "doi": "10.48550/arXiv.2504.12735", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "bf039f1e91d2a0e4d94405d32f07e507d73e706ecc87ef9c3014ec6c0ef73042", "sources": ["arxiv", "semantic_scholar"], "title": "PestMA: LLM-based Multi-Agent System for Informed Pest Management", "abstract": "Effective pest management is complex due to the need for accurate, context-specific decisions. Recent advancements in large language models (LLMs) open new possibilities for addressing these challenges by providing sophisticated, adaptive knowledge acquisition and reasoning. However, existing LLM-based pest management approaches often rely on a single-agent paradigm, which can limit their capacity to incorporate diverse external information, engage in systematic validation, and address complex, threshold-driven decisions. To overcome these limitations, we introduce PestMA, an LLM-based multi-agent system (MAS) designed to generate reliable and evidence-based pest management advice. Building on an editorial paradigm, PestMA features three specialized agents, an Editor for synthesizing pest management recommendations, a Retriever for gathering relevant external data, and a Validator for ensuring correctness. Evaluations on real-world pest scenarios demonstrate that PestMA achieves an initial accuracy of 86.8% for pest management decisions, which increases to 92.6% after validation. These results underscore the value of collaborative agent-based workflows in refining and validating decisions, highlighting the potential of LLM-based multi-agent systems to automate and enhance pest management processes.", "authors": ["Hongrui Shi", "Shunbao Li", "Zhipeng Yuan", "Po Yang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-14", "url": "https://arxiv.org/abs/2504.09855", "pdf_url": "https://arxiv.org/pdf/2504.09855v1", "arxiv_id": "2504.09855", "doi": "10.48550/arXiv.2504.09855", "citation_count": 2, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "19f47293b6d2e35bfe58c0604eb5b0a320d92ebf0d6dbef191b0bfccceb4b729", "sources": ["arxiv", "semantic_scholar"], "title": "Task Memory Engine (TME): Enhancing State Awareness for Multi-Step LLM Agent Tasks", "abstract": "Large Language Models (LLMs) are increasingly used as autonomous agents for multi-step tasks. However, most existing frameworks fail to maintain a structured understanding of the task state, often relying on linear prompt concatenation or shallow memory buffers. This leads to brittle performance, frequent hallucinations, and poor long-range coherence. In this work, we propose the Task Memory Engine (TME), a lightweight and structured memory module that tracks task execution using a hierarchical Task Memory Tree (TMT). Each node in the tree corresponds to a task step, storing relevant input, output, status, and sub-task relationships. We introduce a prompt synthesis method that dynamically generates LLM prompts based on the active node path, significantly improving execution consistency and contextual grounding. Through case studies and comparative experiments on multi-step agent tasks, we demonstrate that TME leads to better task completion accuracy and more interpretable behavior with minimal implementation overhead. A reference implementation of the core TME components is available at https://github.com/biubiutomato/TME-Agent, including basic examples and structured memory integration. While the current implementation uses a tree-based structure, TME is designed to be graph-aware, supporting reusable substeps, converging task paths, and shared dependencies. This lays the groundwork for future DAG-based memory architectures.", "authors": ["Ye Ye"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-11", "url": "https://arxiv.org/abs/2504.08525", "pdf_url": "https://arxiv.org/pdf/2504.08525v4", "arxiv_id": "2504.08525", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/biubiutomato/TME-Agent", "venue": null, "quality_score": 0.1747}
{"id": "ac8657e479cdef2a950ca74bfb1ae5e6bb4c6a09cbd4978d2fe2adffc45b537d", "sources": ["arxiv", "semantic_scholar"], "title": "MooseAgent: A LLM Based Multi-agent Framework for Automating Moose Simulation", "abstract": "The Finite Element Method (FEM) is widely used in engineering and scientific computing, but its pre-processing, solver configuration, and post-processing stages are often time-consuming and require specialized knowledge. This paper proposes an automated solution framework, MooseAgent, for the multi-physics simulation framework MOOSE, which combines large-scale pre-trained language models (LLMs) with a multi-agent system. The framework uses LLMs to understand user-described simulation requirements in natural language and employs task decomposition and multi-round iterative verification strategies to automatically generate MOOSE input files. To improve accuracy and reduce model hallucinations, the system builds and utilizes a vector database containing annotated MOOSE input cards and function documentation. We conducted experimental evaluations on several typical cases, including heat transfer, mechanics, phase field, and multi-physics coupling. The results show that MooseAgent can automate the MOOSE simulation process to a certain extent, especially demonstrating a high success rate when dealing with relatively simple single-physics problems. The main contribution of this research is the proposal of a multi-agent automated framework for MOOSE, which validates its potential in simplifying finite element simulation processes and lowering the user barrier, providing new ideas for the development of intelligent finite element simulation software. The code for the MooseAgent framework proposed in this paper has been open-sourced and is available at https://github.com/taozhan18/MooseAgent", "authors": ["Tao Zhang", "Zhenhai Liu", "Yong Xin", "Yongjun Jiao"], "categories": ["cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-11", "url": "https://arxiv.org/abs/2504.08621", "pdf_url": "https://arxiv.org/pdf/2504.08621v2", "arxiv_id": "2504.08621", "doi": "10.48550/arXiv.2504.08621", "citation_count": 15, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/taozhan18/MooseAgent", "venue": "arXiv.org", "quality_score": 0.301}
{"id": "04c2af2f4370815e5174461b0e63a6e47d1a908fec4749d55eef8575cd823e8c", "sources": ["arxiv", "semantic_scholar"], "title": "MALIBU Benchmark: Multi-Agent LLM Implicit Bias Uncovered", "abstract": "Multi-agent systems, which consist of multiple AI models interacting within a shared environment, are increasingly used for persona-based interactions. However, if not carefully designed, these systems can reinforce implicit biases in large language models (LLMs), raising concerns about fairness and equitable representation. We present MALIBU, a novel benchmark developed to assess the degree to which LLM-based multi-agent systems implicitly reinforce social biases and stereotypes. MALIBU evaluates bias in LLM-based multi-agent systems through scenario-based assessments. AI models complete tasks within predefined contexts, and their responses undergo evaluation by an LLM-based multi-agent judging system in two phases. In the first phase, judges score responses labeled with specific demographic personas (e.g., gender, race, religion) across four metrics. In the second phase, judges compare paired responses assigned to different personas, scoring them and selecting the superior response. Our study quantifies biases in LLM-generated outputs, revealing that bias mitigation may favor marginalized personas over true neutrality, emphasizing the need for nuanced detection, balanced fairness strategies, and transparent evaluation benchmarks in multi-agent systems.", "authors": ["Imran Mirza", "Cole Huang", "Ishwara Vasista", "Rohan Patil", "Asli Akalin", "Sean O'Brien", "Kevin Zhu"], "categories": ["cs.CL", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-10", "url": "https://arxiv.org/abs/2507.01019", "pdf_url": "https://arxiv.org/pdf/2507.01019v1", "arxiv_id": "2507.01019", "doi": "10.48550/arXiv.2507.01019", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "9bacbdc040eae6e2f4dc8a88df6fa0a70f3f58f354c26dd2a7ded45b92638a36", "sources": ["arxiv", "semantic_scholar"], "title": "Achilles Heel of Distributed Multi-Agent Systems", "abstract": "Multi-agent system (MAS) has demonstrated exceptional capabilities in addressing complex challenges, largely due to the integration of multiple large language models (LLMs). However, the heterogeneity of LLMs, the scalability of quantities of LLMs, and local computational constraints pose significant challenges to hosting these models locally. To address these issues, we propose a new framework termed Distributed Multi-Agent System (DMAS). In DMAS, heterogeneous third-party agents function as service providers managed remotely by a central MAS server and each agent offers its services through API interfaces. However, the distributed nature of DMAS introduces several concerns about trustworthiness. In this paper, we study the Achilles heel of distributed multi-agent systems, identifying four critical trustworthiness challenges: free riding, susceptibility to malicious attacks, communication inefficiencies, and system instability. Extensive experiments across seven frameworks and four datasets reveal significant vulnerabilities of the DMAS. These attack strategies can lead to a performance degradation of up to 80% and attain a 100% success rate in executing free riding and malicious attacks. We envision our work will serve as a useful red-teaming tool for evaluating future multi-agent systems and spark further research on trustworthiness challenges in distributed multi-agent systems.", "authors": ["Yiting Zhang", "Yijiang Li", "Tianwei Zhao", "Kaijie Zhu", "Haohan Wang", "Nuno Vasconcelos"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-10", "url": "https://arxiv.org/abs/2504.07461", "pdf_url": "https://arxiv.org/pdf/2504.07461v1", "arxiv_id": "2504.07461", "doi": "10.48550/arXiv.2504.07461", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "ba65c6868646ba173c877339353fec42125ed571d9a43edae901f9ce3ba9980c", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM-Driven Multi-Agent Debate System for Mendelian Diseases", "abstract": "Accurate diagnosis of Mendelian diseases is crucial for precision therapy and assistance in preimplantation genetic diagnosis. However, existing methods often fall short of clinical standards or depend on extensive datasets to build pretrained machine learning models. To address this, we introduce an innovative LLM-Driven multi-agent debate system (MD2GPS) with natural language explanations of the diagnostic results. It utilizes a language model to transform results from data-driven and knowledge-driven agents into natural language, then fostering a debate between these two specialized agents. This system has been tested on 1,185 samples across four independent datasets, enhancing the TOP1 accuracy from 42.9% to 66% on average. Additionally, in a challenging cohort of 72 cases, MD2GPS identified potential pathogenic genes in 12 patients, reducing the diagnostic time by 90%. The methods within each module of this multi-agent debate system are also replaceable, facilitating its adaptation for diagnosing and researching other complex diseases.", "authors": ["Xinyang Zhou", "Yongyong Ren", "Qianqian Zhao", "Daoyi Huang", "Xinbo Wang", "Tingting Zhao", "Zhixing Zhu", "Wenyuan He", "Shuyuan Li", "Yan Xu", "Yu Sun", "Yongguo Yu", "Shengnan Wu", "Jian Wang", "Guangjun Yu", "Dake He", "Bo Ban", "Hui Lu"], "categories": ["q-bio.GN"], "fields_of_study": ["Biology"], "published_date": "2025-04-10", "url": "https://arxiv.org/abs/2504.07881", "pdf_url": "https://arxiv.org/pdf/2504.07881v2", "arxiv_id": "2504.07881", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0788}
{"id": "94001106ae5c6a4c5fc10e9faf2dc00dd166ce00b4a067933d105b92b46467ab", "sources": ["arxiv", "semantic_scholar"], "title": "Single-Agent vs. Multi-Agent LLM Strategies for Automated Student Reflection Assessment", "abstract": "We explore the use of Large Language Models (LLMs) for automated assessment of open-text student reflections and prediction of academic performance. Traditional methods for evaluating reflections are time-consuming and may not scale effectively in educational settings. In this work, we employ LLMs to transform student reflections into quantitative scores using two assessment strategies (single-agent and multi-agent) and two prompting techniques (zero-shot and few-shot). Our experiments, conducted on a dataset of 5,278 reflections from 377 students over three academic terms, demonstrate that the single-agent with few-shot strategy achieves the highest match rate with human evaluations. Furthermore, models utilizing LLM-assessed reflection scores outperform baselines in both at-risk student identification and grade prediction tasks. These findings suggest that LLMs can effectively automate reflection assessment, reduce educators' workload, and enable timely support for students who may need additional assistance. Our work emphasizes the potential of integrating advanced generative AI technologies into educational practices to enhance student engagement and academic success.", "authors": ["Gen Li", "Li Chen", "Cheng Tang", "Valdemar Švábenský", "Daisuke Deguchi", "Takayoshi Yamashita", "Atsushi Shimada"], "categories": ["cs.LG", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-08", "url": "https://arxiv.org/abs/2504.05716", "pdf_url": "https://arxiv.org/pdf/2504.05716v3", "arxiv_id": "2504.05716", "doi": "10.1007/978-981-96-8186-0_24", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Pacific-Asia Conference on Knowledge Discovery and Data Mining", "quality_score": 0.2113}
{"id": "2fc5365bebafce3996ecb6f01d065841754bd081764aa5112bd5f118582c60cd", "sources": ["arxiv", "semantic_scholar"], "title": "Autono: A ReAct-Based Highly Robust Autonomous Agent Framework", "abstract": "This paper proposes a highly robust autonomous agent framework based on the ReAct paradigm, designed to solve complex tasks through adaptive decision making and multi-agent collaboration. Unlike traditional frameworks that rely on fixed workflows generated by LLM-based planners, this framework dynamically generates next actions during agent execution based on prior trajectories, thereby enhancing its robustness. To address potential termination issues caused by adaptive execution paths, I propose a timely abandonment strategy incorporating a probabilistic penalty mechanism. For multi-agent collaboration, I introduce a memory transfer mechanism that enables shared and dynamically updated memory among agents. The framework's innovative timely abandonment strategy dynamically adjusts the probability of task abandonment via probabilistic penalties, allowing developers to balance conservative and exploratory tendencies in agent execution strategies by tuning hyperparameters. This significantly improves adaptability and task execution efficiency in complex environments. Additionally, agents can be extended through external tool integration, supported by modular design and MCP protocol compatibility, which enables flexible action space expansion. Through explicit division of labor, the multi-agent collaboration mechanism enables agents to focus on specific task components, thereby significantly improving execution efficiency and quality.", "authors": ["Zihao Wu"], "categories": ["cs.MA", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-07", "url": "https://arxiv.org/abs/2504.04650", "pdf_url": "https://arxiv.org/pdf/2504.04650v2", "arxiv_id": "2504.04650", "doi": "10.48550/arXiv.2504.04650", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "30aaae7fa0db366095abff3f42122b52cb811a744a14d93b91c2267d83e2a97b", "sources": ["arxiv", "semantic_scholar"], "title": "EduPlanner: LLM-Based Multi-Agent Systems for Customized and Intelligent Instructional Design", "abstract": "Large Language Models (LLMs) have significantly advanced smart education in the Artificial General Intelligence (AGI) era. A promising application lies in the automatic generalization of instructional design for curriculum and learning activities, focusing on two key aspects: (1) Customized Generation: generating niche-targeted teaching content based on students' varying learning abilities and states, and (2) Intelligent Optimization: iteratively optimizing content based on feedback from learning effectiveness or test scores. Currently, a single large LLM cannot effectively manage the entire process, posing a challenge for designing intelligent teaching plans. To address these issues, we developed EduPlanner, an LLM-based multi-agent system comprising an evaluator agent, an optimizer agent, and a question analyst, working in adversarial collaboration to generate customized and intelligent instructional design for curriculum and learning activities. Taking mathematics lessons as our example, EduPlanner employs a novel Skill-Tree structure to accurately model the background mathematics knowledge of student groups, personalizing instructional design for curriculum and learning activities according to students' knowledge levels and learning abilities. Additionally, we introduce the CIDDP, an LLM-based five-dimensional evaluation module encompassing clarity, Integrity, Depth, Practicality, and Pertinence, to comprehensively assess mathematics lesson plan quality and bootstrap intelligent optimization. Experiments conducted on the GSM8K and Algebra datasets demonstrate that EduPlanner excels in evaluating and optimizing instructional design for curriculum and learning activities. Ablation studies further validate the significance and effectiveness of each component within the framework. Our code is publicly available at https://github.com/Zc0812/Edu_Planner", "authors": ["Xueqiao Zhang", "Chao Zhang", "Jianwen Sun", "Jun Xiao", "Yi Yang", "Yawei Luo"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-07", "url": "https://arxiv.org/abs/2504.05370", "pdf_url": "https://arxiv.org/pdf/2504.05370v1", "arxiv_id": "2504.05370", "doi": "10.1109/TLT.2025.3561332", "citation_count": 39, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/Zc0812/Edu_Planner", "venue": "IEEE Transactions on Learning Technologies", "quality_score": 0.4005}
{"id": "ba9113583cf75b6c32b42ced4524e63f7ccc6e280c2e72f539e102e5eae4ec43", "sources": ["arxiv", "semantic_scholar"], "title": "An Efficient Approach for Cooperative Multi-Agent Learning Problems", "abstract": "In this article, we propose a centralized Multi-Agent Learning framework for learning a policy that models the simultaneous behavior of multiple agents that need to coordinate to solve a certain task. Centralized approaches often suffer from the explosion of an action space that is defined by all possible combinations of individual actions, known as joint actions. Our approach addresses the coordination problem via a sequential abstraction, which overcomes the scalability problems typical to centralized methods. It introduces a meta-agent, called \\textit{supervisor}, which abstracts joint actions as sequential assignments of actions to each agent. This sequential abstraction not only simplifies the centralized joint action space but also enhances the framework's scalability and efficiency. Our experimental results demonstrate that the proposed approach successfully coordinates agents across a variety of Multi-Agent Learning environments of diverse sizes.", "authors": ["Ángel Aso-Mollar", "Eva Onaindia"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-07", "url": "https://arxiv.org/abs/2504.04850", "pdf_url": "https://arxiv.org/pdf/2504.04850v1", "arxiv_id": "2504.04850", "doi": "10.1109/ICTAI62512.2024.00103", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Tools with Artificial Intelligence", "quality_score": 0.1203}
{"id": "57313ddc3671678c351ff0b79509a11a6eff30f852152c07593ea13e6044e34c", "sources": ["arxiv", "semantic_scholar"], "title": "AdaCoder: An Adaptive Planning and Multi-Agent Framework for Function-Level Code Generation", "abstract": "Recently, researchers have proposed many multi-agent frameworks for function-level code generation, which aim to improve software development productivity by automatically generating function-level source code based on task descriptions. A typical multi-agent framework consists of Large Language Model (LLM)-based agents that are responsible for task planning, code generation, testing, debugging, etc. Studies have shown that existing multi-agent code generation frameworks perform well on ChatGPT. However, their generalizability across other foundation LLMs remains unexplored systematically. In this paper, we report an empirical study on the generalizability of four state-of-the-art multi-agent code generation frameworks across six open-source LLMs with varying parameter sizes, architectures, and performance levels. Our study reveals the unstable generalizability of existing frameworks on diverse foundation LLMs. Based on the findings obtained from the empirical study, we propose AdaCoder, a novel adaptive planning, multi-agent framework for function-level code generation. AdaCoder has two phases. Phase-1 is an initial code generation step without planning, which uses an LLM-based coding agent and a script-based testing agent to unleash LLM's native power, identify cases beyond LLM's power, and determine the errors hindering execution. Phase-2 adds a rule-based debugging agent and an LLM-based planning agent for iterative code generation with planning. Our evaluation shows that AdaCoder achieves higher generalizability on diverse LLMs. Compared to the best baseline MapCoder, AdaCoder is on average 27.69% higher in Pass@1, 16 times faster in inference, and 12 times lower in token consumption.", "authors": ["Yueheng Zhu", "Chao Liu", "Xuan He", "Xiaoxue Ren", "Zhongxin Liu", "Ruwei Pan", "Hongyu Zhang"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-05", "url": "https://arxiv.org/abs/2504.04220", "pdf_url": "https://arxiv.org/pdf/2504.04220v1", "arxiv_id": "2504.04220", "doi": "10.1109/TSE.2025.3642621", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "IEEE Transactions on Software Engineering", "quality_score": 0.2113}
{"id": "69e36e45e0dac9641b33019af0508592548dbcc5e1c2d7a019ebf69bb8f39a17", "sources": ["arxiv", "semantic_scholar"], "title": "Enforcement Agents: Enhancing Accountability and Resilience in Multi-Agent AI Frameworks", "abstract": "As autonomous agents become more powerful and widely used, it is becoming increasingly important to ensure they behave safely and stay aligned with system goals, especially in multi-agent settings. Current systems often rely on agents self-monitoring or correcting issues after the fact, but they lack mechanisms for real-time oversight. This paper introduces the Enforcement Agent (EA) Framework, which embeds dedicated supervisory agents into the environment to monitor others, detect misbehavior, and intervene through real-time correction. We implement this framework in a custom drone simulation and evaluate it across 90 episodes using 0, 1, and 2 EA configurations. Results show that adding EAs significantly improves system safety: success rates rise from 0.0% with no EA to 7.4% with one EA and 26.7% with two EAs. The system also demonstrates increased operational longevity and higher rates of malicious drone reformation. These findings highlight the potential of lightweight, real-time supervision for enhancing alignment and resilience in multi-agent systems.", "authors": ["Sagar Tamang", "Dibya Jyoti Bora"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-05", "url": "https://arxiv.org/abs/2504.04070", "pdf_url": "https://arxiv.org/pdf/2504.04070v1", "arxiv_id": "2504.04070", "doi": "10.48550/arXiv.2504.04070", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "1ccbfe56781bd12da8aad013dc34c4823ee0201c4cebae462db660041704eddb", "sources": ["arxiv", "semantic_scholar"], "title": "Les Dissonances: Cross-Tool Harvesting and Polluting in Pool-of-Tools Empowered LLM Agents", "abstract": "Large Language Model (LLM) agents are autonomous systems powered by LLMs, capable of reasoning and planning to solve problems by leveraging a set of tools. However, the integration of multi-tool capabilities in LLM agents introduces challenges in securely managing tools, ensuring their compatibility, handling dependency relationships, and protecting control flows within LLM agent workflows. In this paper, we present the first systematic security analysis of task control flows in multi-tool-enabled LLM agents. We identify a novel threat, Cross-Tool Harvesting and Polluting (XTHP), which includes multiple attack vectors to first hijack the normal control flows of agent tasks, and then collect and pollute confidential or private information within LLM agent systems. To understand the impact of this threat, we developed Chord, a dynamic scanning tool designed to automatically detect real-world agent tools susceptible to XTHP attacks. Our evaluation of 66 real-world tools from the repositories of two major LLM agent development frameworks, LangChain and LlamaIndex, revealed a significant security concern: 75% are vulnerable to XTHP attacks, highlighting the prevalence of this threat.", "authors": ["Zichuan Li", "Jian Cui", "Xiaojing Liao", "Luyi Xing"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-04", "url": "https://arxiv.org/abs/2504.03111", "pdf_url": "https://arxiv.org/pdf/2504.03111v3", "arxiv_id": "2504.03111", "doi": "10.14722/ndss.2026.240577", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "d8b87f0cc466b692ad5befbbd10dd25ab7bb96757653dc9f67f32fc22c570ddc", "sources": ["arxiv", "semantic_scholar"], "title": "How Social is It? A Benchmark for LLMs' Capabilities in Multi-user Multi-turn Social Agent Tasks", "abstract": "Expanding the application of large language models (LLMs) to societal life, instead of primary function only as auxiliary assistants to communicate with only one person at a time, necessitates LLMs' capabilities to independently play roles in multi-user, multi-turn social agent tasks within complex social settings. However, currently the capability has not been systematically measured with available benchmarks. To address this gap, we first introduce an agent task leveling framework grounded in sociological principles. Concurrently, we propose a novel benchmark, How Social Is It (we call it HSII below), designed to assess LLM's social capabilities in comprehensive social agents tasks and benchmark representative models. HSII comprises four stages: format parsing, target selection, target switching conversation, and stable conversation, which collectively evaluate the communication and task completion capabilities of LLMs within realistic social interaction scenarios dataset, HSII-Dataset. The dataset is derived step by step from news dataset. We perform an ablation study by doing clustering to the dataset. Additionally, we investigate the impact of chain of thought (COT) method on enhancing LLMs' social performance. Since COT cost more computation, we further introduce a new statistical metric, COT-complexity, to quantify the efficiency of certain LLMs with COTs for specific social tasks and strike a better trade-off between measurement of correctness and efficiency. Various results of our experiments demonstrate that our benchmark is well-suited for evaluating social skills in LLMs.", "authors": ["Yusen Wu", "Junwu Xiong", "Xiaotie Deng"], "categories": ["cs.CL", "cs.AI", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-04", "url": "https://arxiv.org/abs/2505.04628", "pdf_url": "https://arxiv.org/pdf/2505.04628v1", "arxiv_id": "2505.04628", "doi": "10.48550/arXiv.2505.04628", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "2252879937a233018031537c9f27f4a7985e410ab601708e3122b62e7f3931ea", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Mission Tool Bench: Assessing the Robustness of LLM based Agents through Related and Dynamic Missions", "abstract": "Large language models (LLMs) demonstrate strong potential as agents for tool invocation due to their advanced comprehension and planning capabilities. Users increasingly rely on LLM-based agents to solve complex missions through iterative interactions. However, existing benchmarks predominantly access agents in single-mission scenarios, failing to capture real-world complexity. To bridge this gap, we propose the Multi-Mission Tool Bench. In the benchmark, each test case comprises multiple interrelated missions. This design requires agents to dynamically adapt to evolving demands. Moreover, the proposed benchmark explores all possible mission-switching patterns within a fixed mission number. Specifically, we propose a multi-agent data generation framework to construct the benchmark. We also propose a novel method to evaluate the accuracy and efficiency of agent decisions with dynamic decision trees. Experiments on diverse open-source and closed-source LLMs reveal critical factors influencing agent robustness and provide actionable insights to the tool invocation society.", "authors": ["Peijie Yu", "Yifan Yang", "Jinjian Li", "Zelong Zhang", "Haorui Wang", "Xiao Feng", "Feng Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-03", "url": "https://arxiv.org/abs/2504.02623", "pdf_url": "https://arxiv.org/pdf/2504.02623v3", "arxiv_id": "2504.02623", "doi": "10.48550/arXiv.2504.02623", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "c4800b9ceb2f027b450a0219abdedeb0dc0eccca185b606716dc64b4f5023241", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Resource Allocation in Multi-Agent LLM Systems", "abstract": "With the development of LLMs as agents, there is a growing interest in connecting multiple agents into multi-agent systems to solve tasks concurrently, focusing on their role in task assignment and coordination. This paper explores how LLMs can effectively allocate computational tasks among multiple agents, considering factors such as cost, efficiency, and performance. In this work, we address key questions, including the effectiveness of LLMs as orchestrators and planners, comparing their effectiveness in task assignment and coordination. Our experiments demonstrate that LLMs can achieve high validity and accuracy in resource allocation tasks. We find that the planner method outperforms the orchestrator method in handling concurrent actions, resulting in improved efficiency and better utilization of agents. Additionally, we show that providing explicit information about worker capabilities enhances the allocation strategies of planners, particularly when dealing with suboptimal workers.", "authors": ["Alfonso Amayuelas", "Jingbo Yang", "Saaket Agashe", "Ashwin Nagarajan", "Antonis Antoniades", "Xin Eric Wang", "William Wang"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-02", "url": "https://arxiv.org/abs/2504.02051", "pdf_url": "https://arxiv.org/pdf/2504.02051v2", "arxiv_id": "2504.02051", "doi": "10.48550/arXiv.2504.02051", "citation_count": 16, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "0a72cf0519096f9492a05ea823ccf4b226c281ee9302e2433c37c93c61103db3", "sources": ["arxiv", "semantic_scholar"], "title": "Advancing AI-Scientist Understanding: Multi-Agent LLMs with Interpretable Physics Reasoning", "abstract": "Large Language Models (LLMs) are playing an increasingly important role in physics research by assisting with symbolic manipulation, numerical computation, and scientific reasoning. However, ensuring the reliability, transparency, and interpretability of their outputs remains a major challenge. In this work, we introduce a novel multi-agent LLM physicist framework that fosters collaboration between AI and human scientists through three key modules: a reasoning module, an interpretation module, and an AI-scientist interaction module. Recognizing that effective physics reasoning demands logical rigor, quantitative accuracy, and alignment with established theoretical models, we propose an interpretation module that employs a team of specialized LLM agents-including summarizers, model builders, visualization tools, and testers-to systematically structure LLM outputs into transparent, physically grounded science models. A case study demonstrates that our approach significantly improves interpretability, enables systematic validation, and enhances human-AI collaboration in physics problem-solving and discovery. Our work bridges free-form LLM reasoning with interpretable, executable models for scientific analysis, enabling more transparent and verifiable AI-augmented research.", "authors": ["Yinggan Xu", "Hana Kimlee", "Yijia Xiao", "Di Luo"], "categories": ["cs.AI", "cs.CL", "cs.HC", "physics.comp-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2025-04-02", "url": "https://arxiv.org/abs/2504.01911", "pdf_url": "https://arxiv.org/pdf/2504.01911v2", "arxiv_id": "2504.01911", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0753}
{"id": "88327c3d0ad54eef5a1dc98d1dff8f4fc35ec5676f45f6f64166e0621ab6b24c", "sources": ["arxiv", "semantic_scholar"], "title": "On the Robustness of Agentic Function Calling", "abstract": "Large Language Models (LLMs) are increasingly acting as autonomous agents, with function calling (FC) capabilities enabling them to invoke specific tools for tasks. While prior research has primarily focused on improving FC accuracy, little attention has been given to the robustness of these agents to perturbations in their input. We introduce a benchmark assessing FC robustness in two key areas: resilience to naturalistic query variations, and stability in function calling when the toolkit expands with semantically related tools. Evaluating best-performing FC models on a carefully expanded subset of the Berkeley function calling leaderboard (BFCL), we identify critical weaknesses in existing evaluation methodologies, and highlight areas for improvement in real-world agentic deployments.", "authors": ["Ella Rabinovich", "Ateret Anaby-Tavor"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-01", "url": "https://arxiv.org/abs/2504.00914", "pdf_url": "https://arxiv.org/pdf/2504.00914v1", "arxiv_id": "2504.00914", "doi": "10.48550/arXiv.2504.00914", "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2698}
{"id": "e41c450ad4b922c0230bfc699599d77077b0289d7976ae4ce502306b21a71cd3", "sources": ["arxiv", "semantic_scholar"], "title": "AgentNet: Decentralized Evolutionary Coordination for LLM-based Multi-Agent Systems", "abstract": "The rapid advancement of large language models (LLMs) has enabled the development of multi-agent systems where multiple LLM-based agents collaborate on complex tasks. However, existing systems often rely on centralized coordination, leading to scalability bottlenecks, reduced adaptability, and single points of failure. Privacy and proprietary knowledge concerns further hinder cross-organizational collaboration, resulting in siloed expertise. We propose AgentNet, a decentralized, Retrieval-Augmented Generation (RAG)-based framework that enables LLM-based agents to specialize, evolve, and collaborate autonomously in a dynamically structured Directed Acyclic Graph (DAG). Unlike prior approaches with static roles or centralized control, AgentNet allows agents to adjust connectivity and route tasks based on local expertise and context. AgentNet introduces three key innovations: (1) a fully decentralized coordination mechanism that eliminates the need for a central orchestrator, enhancing robustness and emergent intelligence; (2) dynamic agent graph topology that adapts in real time to task demands, ensuring scalability and resilience; and (3) a retrieval-based memory system for agents that supports continual skill refinement and specialization. By minimizing centralized control and data exchange, AgentNet enables fault-tolerant, privacy-preserving collaboration across organizations. Experiments show that AgentNet achieves higher task accuracy than both single-agent and centralized multi-agent baselines.", "authors": ["Yingxuan Yang", "Huacan Chai", "Shuai Shao", "Yuanyi Song", "Siyuan Qi", "Renting Rui", "Weinan Zhang"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-01", "url": "https://arxiv.org/abs/2504.00587", "pdf_url": "https://arxiv.org/pdf/2504.00587v2", "arxiv_id": "2504.00587", "doi": "10.48550/arXiv.2504.00587", "citation_count": 66, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4565}
{"id": "46cfd00e530c5b78c6026c987ca6eee8558a8ebd0a205704ea54e3457b8809bd", "sources": ["arxiv", "semantic_scholar"], "title": "When Persuasion Overrides Truth in Multi-Agent LLM Debates: Introducing a Confidence-Weighted Persuasion Override Rate (CW-POR)", "abstract": "In many real-world scenarios, a single Large Language Model (LLM) may encounter contradictory claims-some accurate, others forcefully incorrect-and must judge which is true. We investigate this risk in a single-turn, multi-agent debate framework: one LLM-based agent provides a factual answer from TruthfulQA, another vigorously defends a falsehood, and the same LLM architecture serves as judge. We introduce the Confidence-Weighted Persuasion Override Rate (CW-POR), which captures not only how often the judge is deceived but also how strongly it believes the incorrect choice. Our experiments on five open-source LLMs (3B-14B parameters), where we systematically vary agent verbosity (30-300 words), reveal that even smaller models can craft persuasive arguments that override truthful answers-often with high confidence. These findings underscore the importance of robust calibration and adversarial testing to prevent LLMs from confidently endorsing misinformation.", "authors": ["Mahak Agarwal", "Divyam Khanna"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-01", "url": "https://arxiv.org/abs/2504.00374", "pdf_url": "https://arxiv.org/pdf/2504.00374v1", "arxiv_id": "2504.00374", "doi": "10.48550/arXiv.2504.00374", "citation_count": 9, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "5d777a34dfc130a1ff11f20888d68c3793b4e9c0d0c56697ae716d8500f1f50c", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent LLM Judge: automatic personalized LLM judge design for evaluating natural language generation applications", "abstract": "Large Language Models (LLMs) have demonstrated impressive performance across diverse domains, yet they still encounter challenges such as insufficient domain-specific knowledge, biases, and hallucinations. This underscores the need for robust evaluation methodologies to accurately assess LLM-based applications. Traditional evaluation methods, which rely on word overlap or text embeddings, are inadequate for capturing the nuanced semantic information necessary to evaluate dynamic, open-ended text generation. Recent research has explored leveraging LLMs to mimic human reasoning and decision-making processes for evaluation purposes known as LLM-as-a-judge framework. However, these existing frameworks have two significant limitations. First, they lack the flexibility to adapt to different text styles, including various answer and ground truth styles, thereby reducing their generalization performance. Second, the evaluation scores produced by these frameworks are often skewed and hard to interpret, showing a low correlation with human judgment. To address these challenges, we propose a novel dynamic multi-agent system that automatically designs personalized LLM judges for various natural language generation applications. This system iteratively refines evaluation prompts and balances the trade-off between the adaptive requirements of downstream tasks and the alignment with human perception. Our experimental results show that the proposed multi-agent LLM Judge framework not only enhances evaluation accuracy compared to existing methods but also produces evaluation scores that better align with human perception.", "authors": ["Hongliu Cao", "Ilias Driouich", "Robin Singh", "Eoin Thomas"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-04-01", "url": "https://arxiv.org/abs/2504.02867", "pdf_url": "https://arxiv.org/pdf/2504.02867v1", "arxiv_id": "2504.02867", "doi": "10.48550/arXiv.2504.02867", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "561dd2290dda0931eaae3f5e8608298fb80786d424e2b90915b6747660cebbcc", "sources": ["arxiv", "semantic_scholar"], "title": "$\\textit{Agents Under Siege}$: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks", "abstract": "Most discussions about Large Language Model (LLM) safety have focused on single-agent settings but multi-agent LLM systems now create novel adversarial risks because their behavior depends on communication between agents and decentralized reasoning. In this work, we innovatively focus on attacking pragmatic systems that have constrains such as limited token bandwidth, latency between message delivery, and defense mechanisms. We design a $\\textit{permutation-invariant adversarial attack}$ that optimizes prompt distribution across latency and bandwidth-constraint network topologies to bypass distributed safety mechanisms within the system. Formulating the attack path as a problem of $\\textit{maximum-flow minimum-cost}$, coupled with the novel $\\textit{Permutation-Invariant Evasion Loss (PIEL)}$, we leverage graph-based optimization to maximize attack success rate while minimizing detection risk. Evaluating across models including $\\texttt{Llama}$, $\\texttt{Mistral}$, $\\texttt{Gemma}$, $\\texttt{DeepSeek}$ and other variants on various datasets like $\\texttt{JailBreakBench}$ and $\\texttt{AdversarialBench}$, our method outperforms conventional attacks by up to $7\\times$, exposing critical vulnerabilities in multi-agent systems. Moreover, we demonstrate that existing defenses, including variants of $\\texttt{Llama-Guard}$ and $\\texttt{PromptGuard}$, fail to prohibit our attack, emphasizing the urgent need for multi-agent specific safety mechanisms.", "authors": ["Rana Muhammad Shahroz Khan", "Zhen Tan", "Sukwon Yun", "Charles Fleming", "Tianlong Chen"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-31", "url": "https://arxiv.org/abs/2504.00218", "pdf_url": "https://arxiv.org/pdf/2504.00218v2", "arxiv_id": "2504.00218", "doi": "10.48550/arXiv.2504.00218", "citation_count": 26, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3578}
{"id": "0b8df8e7cdce553257487a01be447025f1d572da62808358cb6b8f19ad0b71b4", "sources": ["arxiv", "semantic_scholar"], "title": "An Analysis of Decoding Methods for LLM-based Agents for Faithful Multi-Hop Question Answering", "abstract": "Large Language Models (LLMs) frequently produce factually inaccurate outputs - a phenomenon known as hallucination - which limits their accuracy in knowledge-intensive NLP tasks. Retrieval-augmented generation and agentic frameworks such as Reasoning and Acting (ReAct) can address this issue by giving the model access to external knowledge. However, LLMs often fail to remain faithful to retrieved information. Mitigating this is critical, especially if LLMs are required to reason about the retrieved information. Recent research has explored training-free decoding strategies to improve the faithfulness of model generations. We present a systematic analysis of how the combination of the ReAct framework and decoding strategies (i.e., DeCoRe, DoLa, and CAD) can influence the faithfulness of LLM-generated answers. Our results show that combining an agentic framework for knowledge retrieval with decoding methods that enhance faithfulness can increase accuracy on the downstream Multi-Hop Question Answering tasks. For example, we observe an F1 increase from 19.5 to 32.6 on HotpotQA when using ReAct and DoLa.", "authors": ["Alexander Murphy", "Mohd Sanad Zaki Rizvi", "Aden Haussmann", "Ping Nie", "Guifu Liu", "Aryo Pradipta Gema", "Pasquale Minervini"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-30", "url": "https://arxiv.org/abs/2503.23415", "pdf_url": "https://arxiv.org/pdf/2503.23415v1", "arxiv_id": "2503.23415", "doi": "10.48550/arXiv.2503.23415", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "c2406d066482e11b9788fa0f4726e8d02227844bcce3b046a7e116414e7ff332", "sources": ["arxiv", "semantic_scholar"], "title": "SPIO: Ensemble and Selective Strategies via LLM-Based Multi-Agent Planning in Automated Data Science", "abstract": "Large Language Models (LLMs) have enabled dynamic reasoning in automated data analytics, yet recent multi-agent systems remain limited by rigid, single-path workflows that restrict strategic exploration and often lead to suboptimal outcomes. To overcome these limitations, we propose SPIO (Sequential Plan Integration and Optimization), a framework that replaces rigid workflows with adaptive, multi-path planning across four core modules: data preprocessing, feature engineering, model selection, and hyperparameter tuning. In each module, specialized agents generate diverse candidate strategies, which are cascaded and refined by an optimization agent. SPIO offers two operating modes: SPIO-S for selecting a single optimal pipeline, and SPIO-E for ensembling top-k pipelines to maximize robustness. Extensive evaluations on Kaggle and OpenML benchmarks show that SPIO consistently outperforms state-of-the-art baselines, achieving an average performance gain of 5.6%. By explicitly exploring and integrating multiple solution paths, SPIO delivers a more flexible, accurate, and reliable foundation for automated data science.", "authors": ["Wonduk Seo", "Juhyeon Lee", "Yanjun Shao", "Qingshan Zhou", "Seunghyun Lee", "Yi Bu"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-30", "url": "https://arxiv.org/abs/2503.23314", "pdf_url": "https://arxiv.org/pdf/2503.23314v2", "arxiv_id": "2503.23314", "doi": "10.48550/arXiv.2503.23314", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "d49c38200d70e9fd28820f14a0dbfa1718b263a4824904eb738268432994778f", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Large Language Models, a survey", "abstract": "Background: There is great interest in agentic LLMs, large language models that act as agents. Objectives: We review the growing body of work in this area and provide a research agenda. Methods: Agentic LLMs are LLMs that (1) reason, (2) act, and (3) interact. We organize the literature according to these three categories. Results: The research in the first category focuses on reasoning, reflection, and retrieval, aiming to improve decision making; the second category focuses on action models, robots, and tools, aiming for agents that act as useful assistants; the third category focuses on multi-agent systems, aiming for collaborative task solving and simulating interaction to study emergent social behavior. We find that works mutually benefit from results in other categories: retrieval enables tool use, reflection improves multi-agent collaboration, and reasoning benefits all categories. Conclusions: We discuss applications of agentic LLMs and provide an agenda for further research. Important applications are in medical diagnosis, logistics and financial market analysis. Meanwhile, self-reflective agents playing roles and interacting with one another augment the process of scientific research itself. Further, agentic LLMs provide a solution for the problem of LLMs running out of training data: inference-time behavior generates new training states, such that LLMs can keep learning without needing ever larger datasets. We note that there is risk associated with LLM assistants taking action in the real world-safety, liability and security are open problems-while agentic LLMs are also likely to benefit society.", "authors": ["Aske Plaat", "Max van Duijn", "Niki van Stein", "Mike Preuss", "Peter van der Putten", "Kees Joost Batenburg"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-29", "url": "https://arxiv.org/abs/2503.23037", "pdf_url": "https://arxiv.org/pdf/2503.23037v3", "arxiv_id": "2503.23037", "doi": "10.1613/jair.1.18675", "citation_count": 124, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Journal of Artificial Intelligence Research", "quality_score": 0.5242}
{"id": "606219d24472eff52f73e35a474fa3ae26fc6b4232fc284c6fbd322d04d37304", "sources": ["arxiv", "semantic_scholar"], "title": "EncGPT: A Multi-Agent Workflow for Dynamic Encryption Algorithms", "abstract": "Communication encryption is crucial in computer technology, but existing algorithms struggle with balancing cost and security. We propose EncGPT, a multi-agent framework using large language models (LLM). It includes rule, encryption, and decryption agents that generate encryption rules and apply them dynamically. This approach addresses gaps in LLM-based multi-agent systems for communication security. We tested GPT-4o's rule generation and implemented a substitution encryption workflow with homomorphism preservation, achieving an average execution time of 15.99 seconds.", "authors": ["Donghe Li", "Zuchen Li", "Ye Yang", "Li Sun", "Dou An", "Qingyu Yang"], "categories": ["cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-29", "url": "https://arxiv.org/abs/2503.23138", "pdf_url": "https://arxiv.org/pdf/2503.23138v1", "arxiv_id": "2503.23138", "doi": "10.48550/arXiv.2503.23138", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.11}
{"id": "613a16661f9b6cd26c66c65d8fdecba330283f4b0604f1e7e8d742d889881b13", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Evolving Multi-Agent Simulations for Realistic Clinical Interactions", "abstract": "In this work, we introduce MedAgentSim, an open-source simulated clinical environment with doctor, patient, and measurement agents designed to evaluate and enhance LLM performance in dynamic diagnostic settings. Unlike prior approaches, our framework requires doctor agents to actively engage with patients through multi-turn conversations, requesting relevant medical examinations (e.g., temperature, blood pressure, ECG) and imaging results (e.g., MRI, X-ray) from a measurement agent to mimic the real-world diagnostic process. Additionally, we incorporate self improvement mechanisms that allow models to iteratively refine their diagnostic strategies. We enhance LLM performance in our simulated setting by integrating multi-agent discussions, chain-of-thought reasoning, and experience-based knowledge retrieval, facilitating progressive learning as doctor agents interact with more patients. We also introduce an evaluation benchmark for assessing the LLM's ability to engage in dynamic, context-aware diagnostic interactions. While MedAgentSim is fully automated, it also supports a user-controlled mode, enabling human interaction with either the doctor or patient agent. Comprehensive evaluations in various simulated diagnostic scenarios demonstrate the effectiveness of our approach. Our code, simulation tool, and benchmark are available at \\href{https://medagentsim.netlify.app/}.", "authors": ["Mohammad Almansoori", "Komal Kumar", "Hisham Cholakkal"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-28", "url": "https://arxiv.org/abs/2503.22678", "pdf_url": "https://arxiv.org/pdf/2503.22678v2", "arxiv_id": "2503.22678", "doi": "10.48550/arXiv.2503.22678", "citation_count": 32, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": "International Conference on Medical Image Computing and Computer-Assisted Intervention", "quality_score": 0.3796}
{"id": "1fc3d21b617643d81b54892ae6490c89b04c82150cc458a7c57256e75c5d41a3", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating LLM-based Agents for Multi-Turn Conversations: A Survey", "abstract": "This survey examines evaluation methods for large language model (LLM)-based agents in multi-turn conversational settings. Using a PRISMA-inspired framework, we systematically reviewed nearly 250 scholarly sources, capturing the state of the art from various venues of publication, and establishing a solid foundation for our analysis. Our study offers a structured approach by developing two interrelated taxonomy systems: one that defines \\emph{what to evaluate} and another that explains \\emph{how to evaluate}. The first taxonomy identifies key components of LLM-based agents for multi-turn conversations and their evaluation dimensions, including task completion, response quality, user experience, memory and context retention, as well as planning and tool integration. These components ensure that the performance of conversational agents is assessed in a holistic and meaningful manner. The second taxonomy system focuses on the evaluation methodologies. It categorizes approaches into annotation-based evaluations, automated metrics, hybrid strategies that combine human assessments with quantitative measures, and self-judging methods utilizing LLMs. This framework not only captures traditional metrics derived from language understanding, such as BLEU and ROUGE scores, but also incorporates advanced techniques that reflect the dynamic, interactive nature of multi-turn dialogues.", "authors": ["Shengyue Guan", "Jindong Wang", "Jiang Bian", "Bin Zhu", "Jian-guang Lou", "Haoyi Xiong"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-28", "url": "https://arxiv.org/abs/2503.22458", "pdf_url": "https://arxiv.org/pdf/2503.22458v2", "arxiv_id": "2503.22458", "doi": "10.1145/3793671", "citation_count": 49, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "ACM Transactions on Intelligent Systems and Technology", "quality_score": 0.4247}
{"id": "f67164bb7ad84e40249cdcac17af4c230ef4495013fc30ad31bdb77273341993", "sources": ["arxiv", "semantic_scholar"], "title": "Debate-Driven Multi-Agent LLMs for Phishing Email Detection", "abstract": "Phishing attacks remain a critical cybersecurity threat. Attackers constantly refine their methods, making phishing emails harder to detect. Traditional detection methods, including rule-based systems and supervised machine learning models, either rely on predefined patterns like blacklists, which can be bypassed with slight modifications, or require large datasets for training and still can generate false positives and false negatives. In this work, we propose a multi-agent large language model (LLM) prompting technique that simulates debates among agents to detect whether the content presented on an email is phishing. Our approach uses two LLM agents to present arguments for or against the classification task, with a judge agent adjudicating the final verdict based on the quality of reasoning provided. This debate mechanism enables the models to critically analyze contextual cue and deceptive patterns in text, which leads to improved classification accuracy. The proposed framework is evaluated on multiple phishing email datasets and demonstrate that mixed-agent configurations consistently outperform homogeneous configurations. Results also show that the debate structure itself is sufficient to yield accurate decisions without extra prompting strategies.", "authors": ["Ngoc Tuong Vy Nguyen", "Felix D Childress", "Yunting Yin"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-27", "url": "https://arxiv.org/abs/2503.22038", "pdf_url": "https://arxiv.org/pdf/2503.22038v1", "arxiv_id": "2503.22038", "doi": "10.1109/ISDFS65363.2025.11012014", "citation_count": 10, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Symposium on Digital Forensics and Security", "quality_score": 0.2603}
{"id": "d0b1198339b5bcc62cbd66b8c3d06dea9e98e680bcf6d0dc936d060024c1a218", "sources": ["arxiv", "semantic_scholar"], "title": "TAMA: A Human-AI Collaborative Thematic Analysis Framework Using Multi-Agent LLMs for Clinical Interviews", "abstract": "Thematic analysis (TA) is a widely used qualitative approach for uncovering latent meanings in unstructured text data. TA provides valuable insights in healthcare but is resource-intensive. Large Language Models (LLMs) have been introduced to perform TA, yet their applications in healthcare remain unexplored. Here, we propose TAMA: A Human-AI Collaborative Thematic Analysis framework using Multi-Agent LLMs for clinical interviews. We leverage the scalability and coherence of multi-agent systems through structured conversations between agents and coordinate the expertise of cardiac experts in TA. Using interview transcripts from parents of children with Anomalous Aortic Origin of a Coronary Artery (AAOCA), a rare congenital heart disease, we demonstrate that TAMA outperforms existing LLM-assisted TA approaches, achieving higher thematic hit rate, coverage, and distinctiveness. TAMA demonstrates strong potential for automated TA in clinical settings by leveraging multi-agent LLM systems with human-in-the-loop integration by enhancing quality while significantly reducing manual workload.", "authors": ["Huimin Xu", "Seungjun Yi", "Terence Lim", "Jiawei Xu", "Andrew Well", "Carlos Mery", "Aidong Zhang", "Yuji Zhang", "Heng Ji", "Keshav Pingali", "Yan Leng", "Ying Ding"], "categories": ["cs.HC", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-26", "url": "https://arxiv.org/abs/2503.20666", "pdf_url": "https://arxiv.org/pdf/2503.20666v1", "arxiv_id": "2503.20666", "doi": "10.48550/arXiv.2503.20666", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "7f5eb530b96ef9793923cc72792f336eea58aef1e72d17113035e5a663b4a454", "sources": ["arxiv", "semantic_scholar"], "title": "LERO: LLM-driven Evolutionary framework with Hybrid Rewards and Enhanced Observation for Multi-Agent Reinforcement Learning", "abstract": "Multi-agent reinforcement learning (MARL) faces two critical bottlenecks distinct from single-agent RL: credit assignment in cooperative tasks and partial observability of environmental states. We propose LERO, a framework integrating Large language models (LLMs) with evolutionary optimization to address these MARL-specific challenges. The solution centers on two LLM-generated components: a hybrid reward function that dynamically allocates individual credit through reward decomposition, and an observation enhancement function that augments partial observations with inferred environmental context. An evolutionary algorithm optimizes these components through iterative MARL training cycles, where top-performing candidates guide subsequent LLM generations. Evaluations in Multi-Agent Particle Environments (MPE) demonstrate LERO's superiority over baseline methods, with improved task performance and training efficiency.", "authors": ["Yuan Wei", "Xiaohan Shan", "Jianmin Li"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-25", "url": "https://arxiv.org/abs/2503.21807", "pdf_url": "https://arxiv.org/pdf/2503.21807v1", "arxiv_id": "2503.21807", "doi": "10.48550/arXiv.2503.21807", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Intelligent Computing", "quality_score": 0.2258}
{"id": "f84105c65f517dfd98606821f64681642d8102b1e0983baed21eb06660369bcc", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Application System in Office Collaboration Scenarios", "abstract": "This paper introduces a multi-agent application system designed to enhance office collaboration efficiency and work quality. The system integrates artificial intelligence, machine learning, and natural language processing technologies, achieving functionalities such as task allocation, progress monitoring, and information sharing. The agents within the system are capable of providing personalized collaboration support based on team members' needs and incorporate data analysis tools to improve decision-making quality. The paper also proposes an intelligent agent architecture that separates Plan and Solver, and through techniques such as multi-turn query rewriting and business tool retrieval, it enhances the agent's multi-intent and multi-turn dialogue capabilities. Furthermore, the paper details the design of tools and multi-turn dialogue in the context of office collaboration scenarios, and validates the system's effectiveness through experiments and evaluations. Ultimately, the system has demonstrated outstanding performance in real business applications, particularly in query understanding, task planning, and tool calling. Looking forward, the system is expected to play a more significant role in addressing complex interaction issues within dynamic environments and large-scale multi-agent systems.", "authors": ["Songtao Sun", "Jingyi Li", "Yuanfei Dong", "Haoguang Liu", "Chenxin Xu", "Fuyang Li", "Qiang Liu"], "categories": ["cs.AI", "cs.CL", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-25", "url": "https://arxiv.org/abs/2503.19584", "pdf_url": "https://arxiv.org/pdf/2503.19584v3", "arxiv_id": "2503.19584", "doi": "10.48550/arXiv.2503.19584", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "5070b7a8da94dc117ae6e0c31558446001f916a29076bee730341dbf52e00571", "sources": ["arxiv", "semantic_scholar"], "title": "OmniNova:A General Multimodal Agent Framework", "abstract": "The integration of Large Language Models (LLMs) with specialized tools presents new opportunities for intelligent automation systems. However, orchestrating multiple LLM-driven agents to tackle complex tasks remains challenging due to coordination difficulties, inefficient resource utilization, and inconsistent information flow. We present OmniNova, a modular multi-agent automation framework that combines language models with specialized tools such as web search, crawling, and code execution capabilities. OmniNova introduces three key innovations: (1) a hierarchical multi-agent architecture with distinct coordinator, planner, supervisor, and specialist agents; (2) a dynamic task routing mechanism that optimizes agent deployment based on task complexity; and (3) a multi-layered LLM integration system that allocates appropriate models to different cognitive requirements. Our evaluations across 50 complex tasks in research, data analysis, and web interaction domains demonstrate that OmniNova outperforms existing frameworks in task completion rate (87\\% vs. baseline 62\\%), efficiency (41\\% reduced token usage), and result quality (human evaluation score of 4.2/5 vs. baseline 3.1/5). We contribute both a theoretical framework for multi-agent system design and an open-source implementation that advances the state-of-the-art in LLM-based automation systems.", "authors": ["Pengfei Du"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-25", "url": "https://arxiv.org/abs/2503.20028", "pdf_url": "https://arxiv.org/pdf/2503.20028v1", "arxiv_id": "2503.20028", "doi": "10.48550/arXiv.2503.20028", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1629}
{"id": "2c89fb2196326de472e5c76b963e2e8e94957e82667ad061e794cb6dfedc2519", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-ABM for Transportation: Assessing the Potential of LLM Agents in System Analysis", "abstract": "Agent-based modeling approaches represent the state-of-art in modeling travel demand and transportation system dynamics and are valuable tools for transportation planning. However, established agent-based approaches in transportation rely on multi-hierarchical mathematical models to simulate travel behavior, which faces theoretical and practical limitations. The advent of large language models (LLM) provides a new opportunity to refine agent-based modeling in transportation. LLM agents, which have impressive reasoning and planning abilities, can serve as a proxy of human travelers and be integrated into the modeling framework. However, despite evidence of their behavioral soundness, no existing studies have assessed the impact and validity of LLM-agent-based simulations from a system perspective in transportation. This paper aims to address this issue by designing and integrating LLM agents with human-traveler-like characteristics into a simulation of a transportation system and assessing its performance based on existing benchmarks. Using the classical transportation setting of the morning commute, we find that not only do the agents exhibit fine behavioral soundness, but also produce system dynamics that align well with standard benchmarks. Our analysis first verifies the effectiveness and potential of LLM-agent-based modeling for transportation planning on the system level.", "authors": ["Tianming Liu", "Jirong Yang", "Yafeng Yin"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-25", "url": "https://arxiv.org/abs/2503.22718", "pdf_url": "https://arxiv.org/pdf/2503.22718v1", "arxiv_id": "2503.22718", "doi": "10.48550/arXiv.2503.22718", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "06a7e4566f478d93b4d702578254d030e0f10bd4534b7700a9ed12f35b16b456", "sources": ["arxiv"], "title": "AgentDropout: Dynamic Agent Elimination for Token-Efficient and High-Performance LLM-Based Multi-Agent Collaboration", "abstract": "Multi-agent systems (MAS) based on large language models (LLMs) have demonstrated significant potential in collaborative problem-solving. However, they still face substantial challenges of low communication efficiency and suboptimal task performance, making the careful design of the agents' communication topologies particularly important. Inspired by the management theory that roles in an efficient team are often dynamically adjusted, we propose AgentDropout, which identifies redundant agents and communication across different communication rounds by optimizing the adjacency matrices of the communication graphs and eliminates them to enhance both token efficiency and task performance. Compared to state-of-the-art methods, AgentDropout achieves an average reduction of 21.6% in prompt token consumption and 18.4% in completion token consumption, along with a performance improvement of 1.14 on the tasks. Furthermore, the extended experiments demonstrate that AgentDropout achieves notable domain transferability and structure robustness, revealing its reliability and effectiveness. We release our code at https://github.com/wangzx1219/AgentDropout.", "authors": ["Zhexuan Wang", "Yutong Wang", "Xuebo Liu", "Liang Ding", "Miao Zhang", "Jie Liu", "Min Zhang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": [], "published_date": "2025-03-24", "url": "https://arxiv.org/abs/2503.18891", "pdf_url": "https://arxiv.org/pdf/2503.18891v1", "arxiv_id": "2503.18891", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wangzx1219/AgentDropout", "venue": null, "quality_score": 0.1232}
{"id": "dcb5b5a4c61b663e335449c0c146a548c23872d4fa12293c53e939d40fe04404", "sources": ["arxiv", "semantic_scholar"], "title": "SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning Tasks", "abstract": "Large language model (LLM) agents need to perform multi-turn interactions in real-world tasks. However, existing multi-turn RL algorithms for optimizing LLM agents fail to perform effective credit assignment over multiple turns while leveraging the generalization capabilities of LLMs and it remains unclear how to develop such algorithms. To study this, we first introduce a new benchmark, ColBench, where an LLM agent interacts with a human collaborator over multiple turns to solve realistic tasks in backend programming and frontend design. Building on this benchmark, we propose a novel RL algorithm, SWEET-RL (RL with Step-WisE Evaluation from Training-time information), that uses a carefully designed optimization objective to train a critic model with access to additional training-time information. The critic provides step-level rewards for improving the policy model. Our experiments demonstrate that SWEET-RL achieves a 6% absolute improvement in success and win rates on ColBench compared to other state-of-the-art multi-turn RL algorithms, enabling Llama-3.1-8B to match or exceed the performance of GPT4-o in realistic collaborative content creation.", "authors": ["Yifei Zhou", "Song Jiang", "Yuandong Tian", "Jason Weston", "Sergey Levine", "Sainbayar Sukhbaatar", "Xian Li"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-19", "url": "https://arxiv.org/abs/2503.15478", "pdf_url": "https://arxiv.org/pdf/2503.15478v1", "arxiv_id": "2503.15478", "doi": "10.48550/arXiv.2503.15478", "citation_count": 79, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4758}
{"id": "770fc49ffeca6c05ab9145f4c8858b7d764fdf558da0ff3ab52bbeda9b1cbf7a", "sources": ["arxiv", "semantic_scholar"], "title": "MAMM-Refine: A Recipe for Improving Faithfulness in Generation with Multi-Agent Collaboration", "abstract": "Multi-agent collaboration among models has shown promise in reasoning tasks but is underexplored in long-form generation tasks like summarization and question-answering. We extend multi-agent multi-model reasoning to generation, specifically to improving faithfulness through refinement, i.e., revising model-generated outputs to remove factual inconsistencies. We investigate how iterative collaboration among multiple instances and types of large language models (LLMs) enhances subtasks in the refinement process, such as error detection, critiquing unfaithful sentences, and making corrections based on critiques. We design intrinsic evaluations for each subtask, with our findings indicating that both multi-agent (multiple instances) and multi-model (diverse LLM types) approaches benefit error detection and critiquing. Additionally, reframing critiquing and refinement as reranking rather than generation tasks improves multi-agent performance. We consolidate these insights into a final \"recipe\" called Multi-Agent Multi-Model Refinement (MAMM-Refine), where multi-agent and multi-model collaboration significantly boosts performance on three summarization datasets as well as on long-form question answering, demonstrating the effectiveness and generalizability of our recipe.", "authors": ["David Wan", "Justin Chih-Yao Chen", "Elias Stengel-Eskin", "Mohit Bansal"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-19", "url": "https://arxiv.org/abs/2503.15272", "pdf_url": "https://arxiv.org/pdf/2503.15272v1", "arxiv_id": "2503.15272", "doi": "10.48550/arXiv.2503.15272", "citation_count": 8, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/meetdavidwan/mammrefine", "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.2386}
{"id": "4c55596b00dee7285da5445d768f8591b623964142afaf0b9ce0a77b0b16fd19", "sources": ["arxiv", "semantic_scholar"], "title": "MDocAgent: A Multi-Modal Multi-Agent Framework for Document Understanding", "abstract": "Document Question Answering (DocQA) is a very common task. Existing methods using Large Language Models (LLMs) or Large Vision Language Models (LVLMs) and Retrieval Augmented Generation (RAG) often prioritize information from a single modal, failing to effectively integrate textual and visual cues. These approaches struggle with complex multi-modal reasoning, limiting their performance on real-world documents. We present MDocAgent (A Multi-Modal Multi-Agent Framework for Document Understanding), a novel RAG and multi-agent framework that leverages both text and image. Our system employs five specialized agents: a general agent, a critical agent, a text agent, an image agent and a summarizing agent. These agents engage in multi-modal context retrieval, combining their individual insights to achieve a more comprehensive understanding of the document's content. This collaborative approach enables the system to synthesize information from both textual and visual components, leading to improved accuracy in question answering. Preliminary experiments on five benchmarks like MMLongBench, LongDocURL demonstrate the effectiveness of our MDocAgent, achieve an average improvement of 12.1% compared to current state-of-the-art method. This work contributes to the development of more robust and comprehensive DocQA systems capable of handling the complexities of real-world documents containing rich textual and visual information. Our data and code are available at https://github.com/aiming-lab/MDocAgent.", "authors": ["Siwei Han", "Peng Xia", "Ruiyi Zhang", "Tong Sun", "Yun Li", "Hongtu Zhu", "Huaxiu Yao"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-18", "url": "https://arxiv.org/abs/2503.13964", "pdf_url": "https://arxiv.org/pdf/2503.13964v1", "arxiv_id": "2503.13964", "doi": "10.48550/arXiv.2503.13964", "citation_count": 51, "influential_citation_count": 14, "has_code": true, "code_url": "https://github.com/aiming-lab/MDocAgent", "venue": "arXiv.org", "quality_score": 0.588}
{"id": "0ae186a61e445b3f5981a7322497520a80525aba6c41c597ed277599ecf535d9", "sources": ["arxiv", "semantic_scholar"], "title": "Towards a Barrier-free GeoQA Portal: Natural Language Interaction with Geospatial Data Using Multi-Agent LLMs and Semantic Search", "abstract": "A Barrier-Free GeoQA Portal: Enhancing Geospatial Data Accessibility with a Multi-Agent LLM Framework Geoportals are vital for accessing and analyzing geospatial data, promoting open spatial data sharing and online geo-information management. Designed with GIS-like interaction and layered visualization, they often challenge non-expert users with complex functionalities and overlapping layers that obscure spatial relationships. We propose a GeoQA Portal using a multi-agent Large Language Model framework for seamless natural language interaction with geospatial data. Complex queries are broken into subtasks handled by specialized agents, retrieving relevant geographic data efficiently. Task plans are shown to users, boosting transparency. The portal supports default and custom data inputs for flexibility. Semantic search via word vector similarity aids data retrieval despite imperfect terms. Case studies, evaluations, and user tests confirm its effectiveness for non-experts, bridging GIS complexity and public access, and offering an intuitive solution for future geoportals.", "authors": ["Yu Feng", "Puzhen Zhang", "Guohui Xiao", "Linfang Ding", "Liqiu Meng"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-18", "url": "https://arxiv.org/abs/2503.14251", "pdf_url": "https://arxiv.org/pdf/2503.14251v1", "arxiv_id": "2503.14251", "doi": "10.48550/arXiv.2503.14251", "citation_count": 5, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Journal of Applied Earth Observation and Geoinformation", "quality_score": 0.2386}
{"id": "5763f7579e60b1e3d2897dffc2d546ed23535645bf0b75e41c38cf822c0b1f73", "sources": ["arxiv", "semantic_scholar"], "title": "MANTRA: Enhancing Automated Method-Level Refactoring with Contextual RAG and Multi-Agent LLM Collaboration", "abstract": "Maintaining and scaling software systems relies heavily on effective code refactoring, yet this process remains labor-intensive, requiring developers to carefully analyze existing codebases and prevent the introduction of new defects. Although recent advancements have leveraged Large Language Models (LLMs) to automate refactoring tasks, current solutions are constrained in scope and lack mechanisms to guarantee code compilability and successful test execution. In this work, we introduce MANTRA, a comprehensive LLM agent-based framework that automates method-level refactoring. MANTRA integrates Context-Aware Retrieval-Augmented Generation, coordinated Multi-Agent Collaboration, and Verbal Reinforcement Learning to emulate human decision-making during refactoring while preserving code correctness and readability. Our empirical study, conducted on 703 instances of \"pure refactorings\" (i.e., code changes exclusively involving structural improvements), drawn from 10 representative Java projects, covers the six most prevalent refactoring operations. Experimental results demonstrate that MANTRA substantially surpasses a baseline LLM model (RawGPT ), achieving an 82.8% success rate (582/703) in producing code that compiles and passes all tests, compared to just 8.7% (61/703) with RawGPT. Moreover, in comparison to IntelliJ's LLM-powered refactoring tool (EM-Assist), MANTRA exhibits a 50% improvement in generating Extract Method transformations. A usability study involving 37 professional developers further shows that refactorings performed by MANTRA are perceived to be as readable and reusable as human-written code, and in certain cases, even more favorable. These results highlight the practical advantages of MANTRA and emphasize the growing potential of LLM-based systems in advancing the automation of software refactoring tasks.", "authors": ["Yisen Xu", "Feng Lin", "Jinqiu Yang", " Tse-Hsun", " Chen", "Nikolaos Tsantalis"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-18", "url": "https://arxiv.org/abs/2503.14340", "pdf_url": "https://arxiv.org/pdf/2503.14340v2", "arxiv_id": "2503.14340", "doi": "10.48550/arXiv.2503.14340", "citation_count": 19, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "e4f3563e9d49ed23b19982b65443c40fcd24bfbd83b45dc52426e10ce4d0fa29", "sources": ["arxiv", "semantic_scholar"], "title": "Why Do Multi-Agent LLM Systems Fail?", "abstract": "Despite enthusiasm for Multi-Agent LLM Systems (MAS), their performance gains on popular benchmarks are often minimal. This gap highlights a critical need for a principled understanding of why MAS fail. Addressing this question requires systematic identification and analysis of failure patterns. We introduce MAST-Data, a comprehensive dataset of 1600+ annotated traces collected across 7 popular MAS frameworks. MAST-Data is the first multi-agent system dataset to outline the failure dynamics in MAS for guiding the development of better future systems. To enable systematic classification of failures for MAST-Data, we build the first Multi-Agent System Failure Taxonomy (MAST). We develop MAST through rigorous analysis of 150 traces, guided closely by expert human annotators and validated by high inter-annotator agreement (kappa = 0.88). This process identifies 14 unique modes, clustered into 3 categories: (i) system design issues, (ii) inter-agent misalignment, and (iii) task verification. To enable scalable annotation, we develop an LLM-as-a-Judge pipeline with high agreement with human annotations. We leverage MAST and MAST-Data to analyze failure patterns across models (GPT4, Claude 3, Qwen2.5, CodeLlama) and tasks (coding, math, general agent), demonstrating improvement headrooms from better MAS design. Our analysis provides insights revealing that identified failures require more sophisticated solutions, highlighting a clear roadmap for future research. We publicly release our comprehensive dataset (MAST-Data), the MAST, and our LLM annotator to facilitate widespread research and development in MAS.", "authors": ["Mert Cemri", "Melissa Z. Pan", "Shuyi Yang", "Lakshya A. Agrawal", "Bhavya Chopra", "Rishabh Tiwari", "Kurt Keutzer", "Aditya Parameswaran", "Dan Klein", "Kannan Ramchandran", "Matei Zaharia", "Joseph E. Gonzalez", "Ion Stoica"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-17", "url": "https://arxiv.org/abs/2503.13657", "pdf_url": "https://arxiv.org/pdf/2503.13657v3", "arxiv_id": "2503.13657", "doi": "10.48550/arXiv.2503.13657", "citation_count": 397, "influential_citation_count": 69, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.9225}
{"id": "9c09efb5185029a30b68054bf4aa2d209e5c07a749183876ca8a341870548e7a", "sources": ["arxiv", "semantic_scholar"], "title": "Toward Generative 6G Simulation: An Experimental Multi-Agent LLM and ns-3 Integration", "abstract": "The move toward open Sixth-Generation (6G) networks necessitates a novel approach to full-stack simulation environments for evaluating complex technology developments before prototyping and real-world implementation. This paper introduces an innovative approach\\footnote{A lightweight, mock version of the code is available on GitHub at that combines a multi-agent framework with the Network Simulator 3 (ns-3) to automate and optimize the generation, debugging, execution, and analysis of complex 5G network scenarios. Our framework orchestrates a suite of specialized agents -- namely, the Simulation Generation Agent, Test Designer Agent, Test Executor Agent, and Result Interpretation Agent -- using advanced LangChain coordination. The Simulation Generation Agent employs a structured chain-of-thought (CoT) reasoning process, leveraging LLMs and retrieval-augmented generation (RAG) to translate natural language simulation specifications into precise ns-3 scripts. Concurrently, the Test Designer Agent generates comprehensive automated test suites by integrating knowledge retrieval techniques with dynamic test case synthesis. The Test Executor Agent dynamically deploys and runs simulations, managing dependencies and parsing detailed performance metrics. At the same time, the Result Interpretation Agent utilizes LLM-driven analysis to extract actionable insights from the simulation outputs. By integrating external resources such as library documentation and ns-3 testing frameworks, our experimental approach can enhance simulation accuracy and adaptability, reducing reliance on extensive programming expertise. A detailed case study using the ns-3 5G-LENA module validates the effectiveness of the proposed approach. The code generation process converges in an average of 1.8 iterations, has a syntax error rate of 17.0%, a mean response time of 7.3 seconds, and receives a human evaluation score of 7.5.", "authors": ["Farhad Rezazadeh", "Amir Ashtari Gargari", "Sandra Lagen", "Houbing Song", "Dusit Niyato", "Lingjia Liu"], "categories": ["cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-17", "url": "https://arxiv.org/abs/2503.13402", "pdf_url": "https://arxiv.org/pdf/2503.13402v1", "arxiv_id": "2503.13402", "doi": "10.1109/MeditCom64437.2025.11104374", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "International Mediterranean Conference on Communications and Networking", "quality_score": 0.2113}
{"id": "b9b30eafe536a01cc145694afbc9f5a7f923861b09fcbc09f1c649b567cc5d1a", "sources": ["arxiv", "semantic_scholar"], "title": "MAP: Multi-user Personalization with Collaborative LLM-powered Agents", "abstract": "The widespread adoption of Large Language Models (LLMs) and LLM-powered agents in multi-user settings underscores the need for reliable, usable methods to accommodate diverse preferences and resolve conflicting directives. Drawing on conflict resolution theory, we introduce a user-centered workflow for multi-user personalization comprising three stages: Reflection, Analysis, and Feedback. We then present MAP -- a \\textbf{M}ulti-\\textbf{A}gent system for multi-user \\textbf{P}ersonalization -- to operationalize this workflow. By delegating subtasks to specialized agents, MAP (1) retrieves and reflects on relevant user information, while enhancing reliability through agent-to-agent interactions, (2) provides detailed analysis for improved transparency and usability, and (3) integrates user feedback to iteratively refine results. Our user study findings (n=12) highlight MAP's effectiveness and usability for conflict resolution while emphasizing the importance of user involvement in resolution verification and failure management. This work highlights the potential of multi-agent systems to implement user-centered, multi-user personalization workflows and concludes by offering insights for personalization in multi-user contexts.", "authors": ["Christine Lee", "Jihye Choi", "Bilge Mutlu"], "categories": ["cs.HC", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-17", "url": "https://arxiv.org/abs/2503.12757", "pdf_url": "https://arxiv.org/pdf/2503.12757v2", "arxiv_id": "2503.12757", "doi": "10.1145/3706599.3719853", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2603}
{"id": "2d3ac4d3af9961f214905917b819ac9bebf2a6d216e9579f958ab9543c6b8c0a", "sources": ["arxiv", "semantic_scholar"], "title": "GameChat: Multi-LLM Dialogue for Safe, Agile, and Socially Optimal Multi-Agent Navigation in Constrained Environments", "abstract": "Safe, agile, and socially compliant multi-robot navigation in cluttered and constrained environments remains a critical challenge. This is especially difficult with self-interested agents with unique, unknown priorities in decentralized settings, where there is no central authority to resolve conflicts induced by spatial symmetry. We address this challenge by proposing an intuitive, but very effective approach, GameChat, which facilitates safe, agile, and deadlock-free navigation for both cooperative and self-interested agents in cluttered environments. Key to our approach is the idea that agents should resolve conflicts on their own using natural language to communicate, much like humans. We evaluate GameChat in simulated environments with doorways and intersections. The results show that even in the worst case, GameChat reduces the time for all agents to reach their goals by over 35% from a naive baseline and by over 20% from a state of the art baseline in the intersection scenario, while doubling the rate of ensuring the agent with a higher priority task reaches the goal first, from 50% (equivalent to random chance) to 100%. We also demonstrate how GameChat can be extended to more than two agents.", "authors": ["Vagul Mahadevan", "Shangtong Zhang", "Rohan Chandra"], "categories": ["cs.RO", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-16", "url": "https://arxiv.org/abs/2503.12333", "pdf_url": "https://arxiv.org/pdf/2503.12333v2", "arxiv_id": "2503.12333", "doi": "10.1109/MRS66243.2025.11357267", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Symposium on Multi-Robot and Multi-Agent Systems", "quality_score": 0.2386}
{"id": "f74ae077d3d8394679b514c1c51a8435b32889e806334901c39fde2b6d6cd7d1", "sources": ["arxiv", "semantic_scholar"], "title": "On Some Fundamental Problems for Multi-Agent Systems Over Multilayer Networks", "abstract": "Many researchers have considered multi-agent systems over single-layer networks as models for studying diffusion phenomena. Since real-world networks involve connections between agents with different semantics (e.g., family member, friend, colleague), the study of multi-agent systems over multilayer networks has assumed importance. Our focus is on one class of multi-agent system models over multilayer networks, namely multilayer synchronous dynamical systems (MSyDSs). We study several fundamental problems for this model. We establish properties of the phase spaces of MSyDSs and bring out interesting differences between single-layer and multilayer dynamical systems. We show that, in general, the problem of determining whether two given MSyDSs are inequivalent is NP-complete. This hardness result holds even when the only difference between the two systems is the local function at just one node in one layer. We also present efficient algorithms for the equivalence problem for restricted versions of MSyDSs (e.g., systems where each local function is a bounded-threshold function, systems where the number of layers is fixed and each local function is symmetric). In addition, we investigate the expressive power of MSyDSs based on the number of layers. In particular, we examine conditions under which a system with k >= 2 layers has an equivalent system with k-1 or fewer layers.", "authors": ["Daniel J. Rosenkrantz", "Madhav V. Marathe", "Zirou Qiu", "S. S. Ravi", "Richard E. Stearns"], "categories": ["cs.MA", "cs.CC"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-16", "url": "https://arxiv.org/abs/2503.12684", "pdf_url": "https://arxiv.org/pdf/2503.12684v1", "arxiv_id": "2503.12684", "doi": "10.48550/arXiv.2503.12684", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.0951}
{"id": "80a506ab27d8591f9379ad3b4c7ba281d86912b93fbe6b1497b95291f3c69ff5", "sources": ["arxiv", "semantic_scholar"], "title": "SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planning", "abstract": "This paper introduces SagaLLM, a structured multi-agent architecture designed to address four foundational limitations of current LLM-based planning systems: unreliable self-validation, context loss, lack of transactional safeguards, and insufficient inter-agent coordination. While recent frameworks leverage LLMs for task decomposition and multi-agent communication, they often fail to ensure consistency, rollback, or constraint satisfaction across distributed workflows. SagaLLM bridges this gap by integrating the Saga transactional pattern with persistent memory, automated compensation, and independent validation agents. It leverages LLMs' generative reasoning to automate key tasks traditionally requiring hand-coded coordination logic, including state tracking, dependency analysis, log schema generation, and recovery orchestration. Although SagaLLM relaxes strict ACID guarantees, it ensures workflow-wide consistency and recovery through modular checkpointing and compensable execution. Empirical evaluations across planning domains demonstrate that standalone LLMs frequently violate interdependent constraints or fail to recover from disruptions. In contrast, SagaLLM achieves significant improvements in consistency, validation accuracy, and adaptive coordination under uncertainty, establishing a robust foundation for real-world, scalable LLM-based multi-agent systems.", "authors": ["Edward Y. Chang", "Longling Geng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-15", "url": "https://arxiv.org/abs/2503.11951", "pdf_url": "https://arxiv.org/pdf/2503.11951v3", "arxiv_id": "2503.11951", "doi": "10.14778/3750601.3750611", "citation_count": 45, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Proceedings of the VLDB Endowment", "quality_score": 0.4157}
{"id": "b44152b68ec8e6f4ce9f24d922509ea4f88ed82c2dd7b93fcdfc57e9840a5c7b", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Systems Execute Arbitrary Malicious Code", "abstract": "Multi-agent systems coordinate LLM-based agents to perform tasks on users' behalf. In real-world applications, multi-agent systems will inevitably interact with untrusted inputs, such as malicious Web content, files, email attachments, and more. Using several recently proposed multi-agent frameworks as concrete examples, we demonstrate that adversarial content can hijack control and communication within the system to invoke unsafe agents and functionalities. This results in a complete security breach, up to execution of arbitrary malicious code on the user's device or exfiltration of sensitive data from the user's containerized environment. For example, when agents are instantiated with GPT-4o, Web-based attacks successfully cause the multi-agent system execute arbitrary malicious code in 58-90\\% of trials (depending on the orchestrator). In some model-orchestrator configurations, the attack success rate is 100\\%. We also demonstrate that these attacks succeed even if individual agents are not susceptible to direct or indirect prompt injection, and even if they refuse to perform harmful actions. We hope that these results will motivate development of trust and security models for multi-agent systems before they are widely deployed.", "authors": ["Harold Triedman", "Rishi Jha", "Vitaly Shmatikov"], "categories": ["cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-15", "url": "https://arxiv.org/abs/2503.12188", "pdf_url": "https://arxiv.org/pdf/2503.12188v2", "arxiv_id": "2503.12188", "doi": "10.48550/arXiv.2503.12188", "citation_count": 40, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4032}
{"id": "2eb4d38aed356908c59be7ce8131e15cc8810ea3fc4cb0d8fa07d8e89d4c66e0", "sources": ["arxiv", "semantic_scholar"], "title": "LLMs Working in Harmony: A Survey on the Technological Aspects of Building Effective LLM-Based Multi Agent Systems", "abstract": "This survey investigates foundational technologies essential for developing effective Large Language Model (LLM)-based multi-agent systems. Aiming to answer how best to optimize these systems for collaborative, dynamic environments, we focus on four critical areas: Architecture, Memory, Planning, and Technologies/Frameworks. By analyzing recent advancements and their limitations - such as scalability, real-time response challenges, and agent coordination constraints, we provide a detailed view of the technological landscape. Frameworks like the Mixture of Agents architecture and the ReAct planning model exemplify current innovations, showcasing improvements in role assignment and decision-making. This review synthesizes key strengths and persistent challenges, offering practical recommendations to enhance system scalability, agent collaboration, and adaptability. Our findings provide a roadmap for future research, supporting the creation of robust, efficient multi-agent systems that advance both individual agent performance and collective system resilience.", "authors": ["R. M. Aratchige", "W. M. K. S. Ilmini"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-13", "url": "https://arxiv.org/abs/2504.01963", "pdf_url": "https://arxiv.org/pdf/2504.01963v1", "arxiv_id": "2504.01963", "doi": "10.48550/arXiv.2504.01963", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "c569a0a42b6ae419ef83066fad88372ebdfdee05a1039bb7ff0cab72f6f33a50", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Multi-Agent Systems via Reinforcement Learning with LLM-based Planner and Graph-based Policy", "abstract": "Multi-agent systems (MAS) have shown great potential in executing complex tasks, but coordination and safety remain significant challenges. Multi-Agent Reinforcement Learning (MARL) offers a promising framework for agent collaboration, but it faces difficulties in handling complex tasks and designing reward functions. The introduction of Large Language Models (LLMs) has brought stronger reasoning and cognitive abilities to MAS, but existing LLM-based systems struggle to respond quickly and accurately in dynamic environments. To address these challenges, we propose LLM-based Graph Collaboration MARL (LGC-MARL), a framework that efficiently combines LLMs and MARL. This framework decomposes complex tasks into executable subtasks and achieves efficient collaboration among multiple agents through graph-based coordination. Specifically, LGC-MARL consists of two main components: an LLM planner and a graph-based collaboration meta policy. The LLM planner transforms complex task instructions into a series of executable subtasks, evaluates the rationality of these subtasks using a critic model, and generates an action dependency graph. The graph-based collaboration meta policy facilitates communication and collaboration among agents based on the action dependency graph, and adapts to new task environments through meta-learning. Experimental results on the AI2-THOR simulation platform demonstrate the superior performance and scalability of LGC-MARL in completing various complex tasks.", "authors": ["Ziqi Jia", "Junjie Li", "Xiaoyang Qu", "Jianzong Wang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-13", "url": "https://arxiv.org/abs/2503.10049", "pdf_url": "https://arxiv.org/pdf/2503.10049v1", "arxiv_id": "2503.10049", "doi": "10.1109/ICRA55743.2025.11127486", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Robotics and Automation", "quality_score": 0.2865}
{"id": "645f5a3958fcb46e16c734addda01553523978f017222c2b5df84b41f80b255e", "sources": ["arxiv", "semantic_scholar"], "title": "AgentDAM: Privacy Leakage Evaluation for Autonomous Web Agents", "abstract": "Autonomous AI agents that can follow instructions and perform complex multi-step tasks have tremendous potential to boost human productivity. However, to perform many of these tasks, the agents need access to personal information from their users, raising the question of whether they are capable of using it appropriately. In this work, we introduce a new benchmark AgentDAM that measures if AI web-navigation agents follow the privacy principle of ``data minimization''. For the purposes of our benchmark, data minimization means that the agent uses a piece of potentially sensitive information only if it is ``necessary'' to complete a particular task. Our benchmark simulates realistic web interaction scenarios end-to-end and is adaptable to all existing web navigation agents. We use AgentDAM to evaluate how well AI agents built on top of GPT-4, Llama-3 and Claude can limit processing of potentially private information, and show that they are prone to inadvertent use of unnecessary sensitive information. We also propose a prompting-based defense that reduces information leakage, and demonstrate that our end-to-end benchmarking provides a more realistic measure than probing LLMs about privacy. Our results highlight that further research is needed to develop AI agents that can prioritize data minimization at inference time.", "authors": ["Arman Zharmagambetov", "Chuan Guo", "Ivan Evtimov", "Maya Pavlova", "Ruslan Salakhutdinov", "Kamalika Chaudhuri"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-12", "url": "https://arxiv.org/abs/2503.09780", "pdf_url": "https://arxiv.org/pdf/2503.09780v3", "arxiv_id": "2503.09780", "doi": "10.48550/arXiv.2503.09780", "citation_count": 54, "influential_citation_count": 6, "has_code": true, "code_url": "https://github.com/facebookresearch/ai-agent-privacy", "venue": "arXiv.org", "quality_score": 0.4351}
{"id": "1470a2122953b69114686a7bbb5f343c99ddf3b85553117c432bb225e77f1266", "sources": ["arxiv", "semantic_scholar"], "title": "A Survey on Trustworthy LLM Agents: Threats and Countermeasures", "abstract": "With the rapid evolution of Large Language Models (LLMs), LLM-based agents and Multi-agent Systems (MAS) have significantly expanded the capabilities of LLM ecosystems. This evolution stems from empowering LLMs with additional modules such as memory, tools, environment, and even other agents. However, this advancement has also introduced more complex issues of trustworthiness, which previous research focused solely on LLMs could not cover. In this survey, we propose the TrustAgent framework, a comprehensive study on the trustworthiness of agents, characterized by modular taxonomy, multi-dimensional connotations, and technical implementation. By thoroughly investigating and summarizing newly emerged attacks, defenses, and evaluation methods for agents and MAS, we extend the concept of Trustworthy LLM to the emerging paradigm of Trustworthy Agent. In TrustAgent, we begin by deconstructing and introducing various components of the Agent and MAS. Then, we categorize their trustworthiness into intrinsic (brain, memory, and tool) and extrinsic (user, agent, and environment) aspects. Subsequently, we delineate the multifaceted meanings of trustworthiness and elaborate on the implementation techniques of existing research related to these internal and external modules. Finally, we present our insights and outlook on this domain, aiming to provide guidance for future endeavors.", "authors": ["Miao Yu", "Fanci Meng", "Xinyun Zhou", "Shilong Wang", "Junyuan Mao", "Linsey Pang", "Tianlong Chen", "Kun Wang", "Xinfeng Li", "Yongfeng Zhang", "Bo An", "Qingsong Wen"], "categories": ["cs.MA", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-12", "url": "https://arxiv.org/abs/2503.09648", "pdf_url": "https://arxiv.org/pdf/2503.09648v1", "arxiv_id": "2503.09648", "doi": "10.1145/3711896.3736561", "citation_count": 96, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Knowledge Discovery and Data Mining", "quality_score": 0.4967}
{"id": "d69521950dd2fa1a60b00b4dac6da08a163699d112ce82be12cd72b31079f591", "sources": ["arxiv", "semantic_scholar"], "title": "ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning", "abstract": "Recent research on Reasoning of Large Language Models (LLMs) has sought to further enhance their performance by integrating meta-thinking -- enabling models to monitor, evaluate, and control their reasoning processes for more adaptive and effective problem-solving. However, current single-agent work lacks a specialized design for acquiring meta-thinking, resulting in low efficacy. To address this challenge, we introduce Reinforced Meta-thinking Agents (ReMA), a novel framework that leverages Multi-Agent Reinforcement Learning (MARL) to elicit meta-thinking behaviors, encouraging LLMs to think about thinking. ReMA decouples the reasoning process into two hierarchical agents: a high-level meta-thinking agent responsible for generating strategic oversight and plans, and a low-level reasoning agent for detailed executions. Through iterative reinforcement learning with aligned objectives, these agents explore and learn collaboration, leading to improved generalization and robustness. Empirical results from single-turn experiments demonstrate that ReMA outperforms single-agent RL baselines on complex reasoning tasks, including competitive-level mathematical benchmarks and LLM-as-a-Judge benchmarks. Additionally, we further extend ReMA to multi-turn interaction settings, leveraging turn-level ratio and parameter sharing to improve efficiency. Comprehensive ablation studies further illustrate the evolving dynamics of each distinct agent, providing valuable insights into how the meta-thinking reasoning process enhances the reasoning capabilities of LLMs. Our code can be found in https://github.com/ziyuwan/ReMA-public", "authors": ["Ziyu Wan", "Yunxiang Li", "Xiaoyu Wen", "Yan Song", "Hanjing Wang", "Linyi Yang", "Mark Schmidt", "Jun Wang", "Weinan Zhang", "Shuyue Hu", "Ying Wen"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-12", "url": "https://arxiv.org/abs/2503.09501", "pdf_url": "https://arxiv.org/pdf/2503.09501v3", "arxiv_id": "2503.09501", "doi": "10.48550/arXiv.2503.09501", "citation_count": 65, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/ziyuwan/ReMA-public", "venue": "arXiv.org", "quality_score": 0.4549}
{"id": "87b0668dc84532140ef08ee6be9a10ad2666dabc95e0e8ad70d39d8ec0e41780", "sources": ["arxiv", "semantic_scholar"], "title": "ReelWave: Multi-Agentic Movie Sound Generation through Multimodal LLM Conversation", "abstract": "Current audio generation conditioned by text or video focuses on aligning audio with text/video modalities. Despite excellent alignment results, these multimodal frameworks still cannot be directly applied to compelling movie storytelling involving multiple scenes, where \"on-screen\" sounds require temporally-aligned audio generation, while \"off-screen\" sounds contribute to appropriate environment sounds accompanied by background music when applicable. Inspired by professional movie production, this paper proposes a multi-agentic framework for audio generation supervised by an autonomous Sound Director agent, engaging multi-turn conversations with other agents for on-screen and off-screen sound generation through multimodal LLM. To address on-screen sound generation, after detecting any talking humans in videos, we capture semantically and temporally synchronized sound by training a prediction model that forecasts interpretable, time-varying audio control signals: loudness, pitch, and timbre, which are used by a Foley Artist agent to condition a cross-attention module in the sound generation. The Foley Artist works cooperatively with the Composer and Voice Actor agents, and together they autonomously generate off-screen sound to complement the overall production. Each agent takes on specific roles similar to those of a movie production team. To temporally ground audio language models, in ReelWave, text/video conditions are decomposed into atomic, specific sound generation instructions synchronized with visuals when applicable. Consequently, our framework can generate rich and relevant audio content conditioned on video clips extracted from movies.", "authors": ["Zixuan Wang", "Chi-Keung Tang", "Yu-Wing Tai"], "categories": ["cs.SD", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-10", "url": "https://arxiv.org/abs/2503.07217", "pdf_url": "https://arxiv.org/pdf/2503.07217v3", "arxiv_id": "2503.07217", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0753}
{"id": "bcaa28e874181e96264f3d41052cd9f42652f2f888266d77393f872ddc7015ae", "sources": ["arxiv", "semantic_scholar"], "title": "AutoMisty: A Multi-Agent LLM Framework for Automated Code Generation in the Misty Social Robot", "abstract": "The social robot's open API allows users to customize open-domain interactions. However, it remains inaccessible to those without programming experience. In this work, we introduce AutoMisty, the first multi-agent collaboration framework powered by large language models (LLMs), to enable the seamless generation of executable Misty robot code from natural language instructions. AutoMisty incorporates four specialized agent modules to manage task decomposition, assignment, problem-solving, and result synthesis. Each agent incorporates a two-layer optimization mechanism, with self-reflection for iterative refinement and human-in-the-loop for better alignment with user preferences. AutoMisty ensures a transparent reasoning process, allowing users to iteratively refine tasks through natural language feedback for precise execution. To evaluate AutoMisty's effectiveness, we designed a benchmark task set spanning four levels of complexity and conducted experiments in a real Misty robot environment. Extensive evaluations demonstrate that AutoMisty not only consistently generates high-quality code but also enables precise code control, significantly outperforming direct reasoning with ChatGPT-4o and ChatGPT-o1. All code, optimized APIs, and experimental videos will be publicly released through the webpage: https://wangxiaoshawn.github.io/AutoMisty.html", "authors": ["Xiao Wang", "Lu Dong", "Sahana Rangasrinivasan", "Ifeoma Nwogu", "Srirangaraj Setlur", "Venugopal Govindaraju"], "categories": ["cs.RO", "cs.AI", "cs.HC", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-09", "url": "https://arxiv.org/abs/2503.06791", "pdf_url": "https://arxiv.org/pdf/2503.06791v2", "arxiv_id": "2503.06791", "doi": "10.1109/IROS60139.2025.11247695", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.25}
{"id": "80e19324d0022c1f9f9db20c7ec076bcb0bab3cc9b991ae6666b0259efd871e8", "sources": ["arxiv", "semantic_scholar"], "title": "Multi Agent based Medical Assistant for Edge Devices", "abstract": "Large Action Models (LAMs) have revolutionized intelligent automation, but their application in healthcare faces challenges due to privacy concerns, latency, and dependency on internet access. This report introduces an ondevice, multi-agent healthcare assistant that overcomes these limitations. The system utilizes smaller, task-specific agents to optimize resources, ensure scalability and high performance. Our proposed system acts as a one-stop solution for health care needs with features like appointment booking, health monitoring, medication reminders, and daily health reporting. Powered by the Qwen Code Instruct 2.5 7B model, the Planner and Caller Agents achieve an average RougeL score of 85.5 for planning and 96.5 for calling for our tasks while being lightweight for on-device deployment. This innovative approach combines the benefits of ondevice systems with multi-agent architectures, paving the way for user-centric healthcare solutions.", "authors": ["Sakharam Gawade", "Shivam Akhouri", "Chinmay Kulkarni", "Jagdish Samant", "Pragya Sahu", " Aastik", "Jai Pahal", "Saswat Meher"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-07", "url": "https://arxiv.org/abs/2503.05397", "pdf_url": "https://arxiv.org/pdf/2503.05397v1", "arxiv_id": "2503.05397", "doi": "10.48550/arXiv.2503.05397", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "f0a8cee3cea1f09fa36cf9340f33450a4071876e5b8d661f19ae9a7e05b462cc", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Inverse Q-Learning from Demonstrations", "abstract": "When reward functions are hand-designed, deep reinforcement learning algorithms often suffer from reward misspecification, causing them to learn suboptimal policies in terms of the intended task objectives. In the single-agent case, inverse reinforcement learning (IRL) techniques attempt to address this issue by inferring the reward function from expert demonstrations. However, in multi-agent problems, misalignment between the learned and true objectives is exacerbated due to increased environment non-stationarity and variance that scales with multiple agents. As such, in multi-agent general-sum games, multi-agent IRL algorithms have difficulty balancing cooperative and competitive objectives. To address these issues, we propose Multi-Agent Marginal Q-Learning from Demonstrations (MAMQL), a novel sample-efficient framework for multi-agent IRL. For each agent, MAMQL learns a critic marginalized over the other agents' policies, allowing for a well-motivated use of Boltzmann policies in the multi-agent context. We identify a connection between optimal marginalized critics and single-agent soft-Q IRL, allowing us to apply a direct, simple optimization criterion from the single-agent domain. Across our experiments on three different simulated domains, MAMQL significantly outperforms previous multi-agent methods in average reward, sample efficiency, and reward recovery by often more than 2-5x. We make our code available at https://sites.google.com/view/mamql .", "authors": ["Nathaniel Haynam", "Adam Khoja", "Dhruv Kumar", "Vivek Myers", "Erdem Bıyık"], "categories": ["cs.MA", "cs.AI", "cs.LG", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-06", "url": "https://arxiv.org/abs/2503.04679", "pdf_url": "https://arxiv.org/pdf/2503.04679v1", "arxiv_id": "2503.04679", "doi": "10.1109/ICRA55743.2025.11128370", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "IEEE International Conference on Robotics and Automation", "quality_score": 0.1293}
{"id": "e93151c707a29a448aa99fe04f7af656069129add71507d1e0dc88d766be71aa", "sources": ["arxiv", "semantic_scholar"], "title": "Parallelized Planning-Acting for Efficient LLM-based Multi-Agent Systems in Minecraft", "abstract": "Recent advancements in Large Language Model~(LLM)-based Multi-Agent Systems (MAS) have demonstrated remarkable potential for tackling complex decision-making tasks. However, existing frameworks inevitably rely on serialized execution paradigms, where agents must complete sequential LLM planning before taking action. This fundamental constraint severely limits real-time responsiveness and adaptation, which is crucial in dynamic environments with ever-changing scenarios like Minecraft. In this paper, we propose a novel parallelized planning-acting framework for LLM-based MAS, featuring a dual-thread architecture with interruptible execution to enable concurrent planning and acting. Specifically, our framework comprises two core threads: (1) a planning thread driven by a centralized memory system, maintaining synchronization of environmental states and agent communication to support dynamic decision-making; and (2) an acting thread equipped with a comprehensive skill library, enabling automated task execution through recursive decomposition. Extensive experiments on Minecraft demonstrate the effectiveness of the proposed framework.", "authors": ["Yaoru Li", "Shunyu Liu", "Tongya Zheng", "Li Sun", "Mingli Song"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-05", "url": "https://arxiv.org/abs/2503.03505", "pdf_url": "https://arxiv.org/pdf/2503.03505v2", "arxiv_id": "2503.03505", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "56f59995496999123f7b2f2f31fae5d7a71541306155046a949dbef87f859afe", "sources": ["arxiv", "semantic_scholar"], "title": "MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems", "abstract": "LLM-based multi-agent systems (MAS) have shown significant potential in tackling diverse tasks. However, to design effective MAS, existing approaches heavily rely on manual configurations or multiple calls of advanced LLMs, resulting in inadaptability and high inference costs. In this paper, we simplify the process of building an MAS by reframing it as a generative language task, where the input is a user query and the output is a corresponding MAS. To address this novel task, we unify the representation of MAS as executable code and propose a consistency-oriented data construction pipeline to create a high-quality dataset comprising coherent and consistent query-MAS pairs. Using this dataset, we train MAS-GPT, an open-source medium-sized LLM that is capable of generating query-adaptive MAS within a single LLM inference. The generated MAS can be seamlessly applied to process user queries and deliver high-quality responses. Extensive experiments on 9 benchmarks and 5 LLMs show that the proposed MAS-GPT consistently outperforms 10+ baseline MAS methods on diverse settings, indicating MAS-GPT's high effectiveness, efficiency and strong generalization ability. Code will be available at https://github.com/rui-ye/MAS-GPT.", "authors": ["Rui Ye", "Shuo Tang", "Rui Ge", "Yaxin Du", "Zhenfei Yin", "Siheng Chen", "Jing Shao"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-05", "url": "https://arxiv.org/abs/2503.03686", "pdf_url": "https://arxiv.org/pdf/2503.03686v1", "arxiv_id": "2503.03686", "doi": "10.48550/arXiv.2503.03686", "citation_count": 37, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/rui-ye/MAS-GPT", "venue": "International Conference on Machine Learning", "quality_score": 0.3949}
{"id": "07eb65d27858b9cd50616058d2e0c222d0b838e0bb2282ab39d3432dae2fbdc3", "sources": ["arxiv", "semantic_scholar"], "title": "Unified Mind Model: Reimagining Autonomous Agents in the LLM Era", "abstract": "Large language models (LLMs) have recently demonstrated remarkable capabilities across domains, tasks, and languages (e.g., ChatGPT and GPT-4), reviving the research of general autonomous agents with human-like cognitive abilities. Such human-level agents require semantic comprehension and instruction-following capabilities, which exactly fall into the strengths of LLMs. Although there have been several initial attempts to build human-level agents based on LLMs, the theoretical foundation remains a challenging open problem. In this paper, we propose a novel theoretical cognitive architecture, the Unified Mind Model (UMM), which offers guidance to facilitate the rapid creation of autonomous agents with human-level cognitive abilities. Specifically, our UMM starts with the global workspace theory and further leverage LLMs to enable the agent with various cognitive abilities, such as multi-modal perception, planning, reasoning, tool use, learning, memory, reflection and motivation. Building upon UMM, we then develop an agent-building engine, MindOS, which allows users to quickly create domain-/task-specific autonomous agents without any programming effort.", "authors": ["Pengbo Hu", "Xiang Ying"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-05", "url": "https://arxiv.org/abs/2503.03459", "pdf_url": "https://arxiv.org/pdf/2503.03459v2", "arxiv_id": "2503.03459", "doi": "10.48550/arXiv.2503.03459", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "eac4f6930ed311688200f8283bb6222e501e6db039d16969c81ba44de5f77972", "sources": ["arxiv", "semantic_scholar"], "title": "Interactive Debugging and Steering of Multi-Agent AI Systems", "abstract": "Fully autonomous teams of LLM-powered AI agents are emerging that collaborate to perform complex tasks for users. What challenges do developers face when trying to build and debug these AI agent teams? In formative interviews with five AI agent developers, we identify core challenges: difficulty reviewing long agent conversations to localize errors, lack of support in current tools for interactive debugging, and the need for tool support to iterate on agent configuration. Based on these needs, we developed an interactive multi-agent debugging tool, AGDebugger, with a UI for browsing and sending messages, the ability to edit and reset prior agent messages, and an overview visualization for navigating complex message histories. In a two-part user study with 14 participants, we identify common user strategies for steering agents and highlight the importance of interactive message resets for debugging. Our studies deepen understanding of interfaces for debugging increasingly important agentic workflows.", "authors": ["Will Epperson", "Gagan Bansal", "Victor Dibia", "Adam Fourney", "Jack Gerrits", "Erkang Zhu", "Saleema Amershi"], "categories": ["cs.MA", "cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-03", "url": "https://arxiv.org/abs/2503.02068", "pdf_url": "https://arxiv.org/pdf/2503.02068v1", "arxiv_id": "2503.02068", "doi": "10.1145/3706598.3713581", "citation_count": 67, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "International Conference on Human Factors in Computing Systems", "quality_score": 0.4581}
{"id": "451dd5800d9fb4b9108a031aefa2287d46aa42eee9959ff9e3699f70f0c57c1f", "sources": ["arxiv", "semantic_scholar"], "title": "Retrieval Models Aren't Tool-Savvy: Benchmarking Tool Retrieval for Large Language Models", "abstract": "Tool learning aims to augment large language models (LLMs) with diverse tools, enabling them to act as agents for solving practical tasks. Due to the limited context length of tool-using LLMs, adopting information retrieval (IR) models to select useful tools from large toolsets is a critical initial step. However, the performance of IR models in tool retrieval tasks remains underexplored and unclear. Most tool-use benchmarks simplify this step by manually pre-annotating a small set of relevant tools for each task, which is far from the real-world scenarios. In this paper, we propose ToolRet, a heterogeneous tool retrieval benchmark comprising 7.6k diverse retrieval tasks, and a corpus of 43k tools, collected from existing datasets. We benchmark six types of models on ToolRet. Surprisingly, even the models with strong performance in conventional IR benchmarks, exhibit poor performance on ToolRet. This low retrieval quality degrades the task pass rate of tool-use LLMs. As a further step, we contribute a large-scale training dataset with over 200k instances, which substantially optimizes the tool retrieval ability of IR models.", "authors": ["Zhengliang Shi", "Yuhan Wang", "Lingyong Yan", "Pengjie Ren", "Shuaiqiang Wang", "Dawei Yin", "Zhaochun Ren"], "categories": ["cs.CL", "cs.AI", "cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-03", "url": "https://arxiv.org/abs/2503.01763", "pdf_url": "https://arxiv.org/pdf/2503.01763v2", "arxiv_id": "2503.01763", "doi": "10.48550/arXiv.2503.01763", "citation_count": 29, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/mangopy/tool-retrieval-benchmark", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3891}
{"id": "ae3b0ae3dbbf6c5b61f6c6c7bde4417e6e137937c1c6a6eb8e8950776846bcea", "sources": ["arxiv", "semantic_scholar"], "title": "MultiAgentBench: Evaluating the Collaboration and Competition of LLM agents", "abstract": "Large Language Models (LLMs) have shown remarkable capabilities as autonomous agents, yet existing benchmarks either focus on single-agent tasks or are confined to narrow domains, failing to capture the dynamics of multi-agent coordination and competition. In this paper, we introduce MultiAgentBench, a comprehensive benchmark designed to evaluate LLM-based multi-agent systems across diverse, interactive scenarios. Our framework measures not only task completion but also the quality of collaboration and competition using novel, milestone-based key performance indicators. Moreover, we evaluate various coordination protocols (including star, chain, tree, and graph topologies) and innovative strategies such as group discussion and cognitive planning. Notably, gpt-4o-mini reaches the average highest task score, graph structure performs the best among coordination protocols in the research scenario, and cognitive planning improves milestone achievement rates by 3%. Code and datasets are public available at https://github.com/MultiagentBench/MARBLE.", "authors": ["Kunlun Zhu", "Hongyi Du", "Zhaochen Hong", "Xiaocheng Yang", "Shuyi Guo", "Zhe Wang", "Zhenhailong Wang", "Cheng Qian", "Xiangru Tang", "Heng Ji", "Jiaxuan You"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-03", "url": "https://arxiv.org/abs/2503.01935", "pdf_url": "https://arxiv.org/pdf/2503.01935v1", "arxiv_id": "2503.01935", "doi": "10.48550/arXiv.2503.01935", "citation_count": 123, "influential_citation_count": 6, "has_code": true, "code_url": "https://github.com/MultiagentBench/MARBLE", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.5234}
{"id": "c83f4c0976bc85668502efa2565053ebbae87a9b0f9a31637756e8cc49fd6049", "sources": ["arxiv", "semantic_scholar"], "title": "LLMDR: LLM-Driven Deadlock Detection and Resolution in Multi-Agent Pathfinding", "abstract": "Multi-Agent Pathfinding (MAPF) is a core challenge in multi-agent systems. Existing learning-based MAPF methods often struggle with scalability, particularly when addressing complex scenarios that are prone to deadlocks. To address these challenges, we introduce LLMDR (LLM-Driven Deadlock Detection and Resolution), an approach designed to resolve deadlocks and improve the performance of learnt MAPF models. LLMDR integrates the inference capabilities of large language models (LLMs) with learnt MAPF models and prioritized planning, enabling it to detect deadlocks and provide customized resolution strategies. We evaluate LLMDR on standard MAPF benchmark maps with varying agent numbers, measuring its performance when combined with several base models. The results demonstrate that LLMDR improves the performance of learnt MAPF models, particularly in deadlock-prone scenarios, with notable improvements in success rates. These findings show the potential of integrating LLMs to improve the scalability of learning-based MAPF methods. The source code for LLMDR is available at: https://github.com/ssbacc/llmdr-dhc", "authors": ["Seungbae Seo", "Junghwan Kim", "Minjeong Shin", "Bongwon Suh"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-03-02", "url": "https://arxiv.org/abs/2503.00717", "pdf_url": "https://arxiv.org/pdf/2503.00717v1", "arxiv_id": "2503.00717", "doi": "10.48550/arXiv.2503.00717", "citation_count": 0, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/ssbacc/llmdr-dhc", "venue": "arXiv.org", "quality_score": 0.1222}
{"id": "b03317ed4248d2bf89cf399418634ed01dd2ef65440ca879ebd346756298b58b", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Security Tax: Trading Off Security and Collaboration Capabilities in Multi-Agent Systems", "abstract": "As AI agents are increasingly adopted to collaborate on complex objectives, ensuring the security of autonomous multi-agent systems becomes crucial. We develop simulations of agents collaborating on shared objectives to study these security risks and security trade-offs. We focus on scenarios where an attacker compromises one agent, using it to steer the entire system toward misaligned outcomes by corrupting other agents. In this context, we observe infectious malicious prompts - the multi-hop spreading of malicious instructions. To mitigate this risk, we evaluated several strategies: two \"vaccination\" approaches that insert false memories of safely handling malicious input into the agents' memory stream, and two versions of a generic safety instruction strategy. While these defenses reduce the spread and fulfillment of malicious instructions in our experiments, they tend to decrease collaboration capability in the agent network. Our findings illustrate potential trade-off between security and collaborative efficiency in multi-agent systems, providing insights for designing more secure yet effective AI collaborations.", "authors": ["Pierre Peigne-Lefebvre", "Mikolaj Kniejski", "Filip Sondej", "Matthieu David", "Jason Hoelscher-Obermaier", "Christian Schroeder de Witt", "Esben Kran"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-26", "url": "https://arxiv.org/abs/2502.19145", "pdf_url": "https://arxiv.org/pdf/2502.19145v2", "arxiv_id": "2502.19145", "doi": "10.48550/arXiv.2502.19145", "citation_count": 26, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3578}
{"id": "87920a547bf6183460bdb5252e10ae0bb26ebfa7026b3bf83c190692faec4c7a", "sources": ["arxiv", "semantic_scholar"], "title": "Agent Trading Arena: A Study on Numerical Understanding in LLM-Based Agents", "abstract": "Large language models (LLMs) have demonstrated remarkable capabilities in natural language tasks, yet their performance in dynamic, real-world financial environments remains underexplored. Existing approaches are limited to historical backtesting, where trading actions cannot influence market prices and agents train only on static data. To address this limitation, we present the Agent Trading Arena, a virtual zero-sum stock market in which LLM-based agents engage in competitive multi-agent trading and directly impact price dynamics. By simulating realistic bid-ask interactions, our platform enables training in scenarios that closely mirror live markets, thereby narrowing the gap between training and evaluation. Experiments reveal that LLMs struggle with numerical reasoning when given plain-text data, often overfitting to local patterns and recent values. In contrast, chart-based visualizations significantly enhance both numerical reasoning and trading performance. Furthermore, incorporating a reflection module yields additional improvements, especially with visual inputs. Evaluations on NASDAQ and CSI datasets demonstrate the superiority of our method, particularly under high volatility. All code and data are available at https://github.com/wekjsdvnm/Agent-Trading-Arena.", "authors": ["Tianmi Ma", "Jiawei Du", "Wenxin Huang", "Wenjie Wang", "Liang Xie", "Xian Zhong", "Joey Tianyi Zhou"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.MA", "q-fin.ST"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2025-02-25", "url": "https://arxiv.org/abs/2502.17967", "pdf_url": "https://arxiv.org/pdf/2502.17967v2", "arxiv_id": "2502.17967", "doi": "10.18653/v1/2025.findings-emnlp.294", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/wekjsdvnm/Agent-Trading-Arena", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.1945}
{"id": "1f976193c7e342bdba232bd22b256fa6cdeb06265a110a67115fb4b9ea73aa9e", "sources": ["arxiv", "semantic_scholar"], "title": "ARACNE: An LLM-Based Autonomous Shell Pentesting Agent", "abstract": "We introduce ARACNE, a fully autonomous LLM-based pentesting agent tailored for SSH services that can execute commands on real Linux shell systems. Introduces a new agent architecture with multi-LLM model support. Experiments show that ARACNE can reach a 60\\% success rate against the autonomous defender ShelLM and a 57.58\\% success rate against the Over The Wire Bandit CTF challenges, improving over the state-of-the-art. When winning, the average number of actions taken by the agent to accomplish the goals was less than 5. The results show that the use of multi-LLM is a promising approach to increase accuracy in the actions.", "authors": ["Tomas Nieponice", "Veronica Valeros", "Sebastian Garcia"], "categories": ["cs.CR", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-24", "url": "https://arxiv.org/abs/2502.18528", "pdf_url": "https://arxiv.org/pdf/2502.18528v1", "arxiv_id": "2502.18528", "doi": "10.48550/arXiv.2502.18528", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "46f46f6bae38205ffc0a535713a415ec3fd582fd5c702cb72dd6bc11ac85c1c6", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-LLM-Agent-Based Framework for Economic and Public Policy Analysis", "abstract": "This paper pioneers a novel approach to economic and public policy analysis by leveraging multiple Large Language Models (LLMs) as heterogeneous artificial economic agents. We first evaluate five LLMs' economic decision-making capabilities in solving two-period consumption allocation problems under two distinct scenarios: with explicit utility functions and based on intuitive reasoning. While previous research has often simulated heterogeneity by solely varying prompts, our approach harnesses the inherent variations in analytical capabilities across different LLMs to model agents with diverse cognitive traits. Building on these findings, we construct a Multi-LLM-Agent-Based (MLAB) framework by mapping these LLMs to specific educational groups and corresponding income brackets. Using interest-income taxation as a case study, we demonstrate how the MLAB framework can simulate policy impacts across heterogeneous agents, offering a promising new direction for economic and public policy analysis by leveraging LLMs' human-like reasoning capabilities and computational power.", "authors": ["Yuzhi Hao", "Danyang Xie"], "categories": ["cs.AI", "econ.GN"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2025-02-24", "url": "https://arxiv.org/abs/2502.16879", "pdf_url": "https://arxiv.org/pdf/2502.16879v1", "arxiv_id": "2502.16879", "doi": "10.48550/arXiv.2502.16879", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "c915925a52420cecc222d6170404faaa00edcd814f01671d187f3c003e5f6eb0", "sources": ["arxiv", "semantic_scholar"], "title": "PMAT: Optimizing Action Generation Order in Multi-Agent Reinforcement Learning", "abstract": "Multi-agent reinforcement learning (MARL) faces challenges in coordinating agents due to complex interdependencies within multi-agent systems. Most MARL algorithms use the simultaneous decision-making paradigm but ignore the action-level dependencies among agents, which reduces coordination efficiency. In contrast, the sequential decision-making paradigm provides finer-grained supervision for agent decision order, presenting the potential for handling dependencies via better decision order management. However, determining the optimal decision order remains a challenge. In this paper, we introduce Action Generation with Plackett-Luce Sampling (AGPS), a novel mechanism for agent decision order optimization. We model the order determination task as a Plackett-Luce sampling process to address issues such as ranking instability and vanishing gradient during the network training process. AGPS realizes credit-based decision order determination by establishing a bridge between the significance of agents' local observations and their decision credits, thus facilitating order optimization and dependency management. Integrating AGPS with the Multi-Agent Transformer, we propose the Prioritized Multi-Agent Transformer (PMAT), a sequential decision-making MARL algorithm with decision order optimization. Experiments on benchmarks including StarCraft II Multi-Agent Challenge, Google Research Football, and Multi-Agent MuJoCo show that PMAT outperforms state-of-the-art algorithms, greatly enhancing coordination efficiency.", "authors": ["Kun Hu", "Muning Wen", "Xihuai Wang", "Shao Zhang", "Yiwei Shi", "Minne Li", "Minglong Li", "Ying Wen"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-23", "url": "https://arxiv.org/abs/2502.16496", "pdf_url": "https://arxiv.org/pdf/2502.16496v1", "arxiv_id": "2502.16496", "doi": "10.48550/arXiv.2502.16496", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1505}
{"id": "8b52549107c602e6663144012bd244827538576ae32ac9d134da862dd6be0c33", "sources": ["arxiv", "semantic_scholar"], "title": "Learning with Limited Shared Information in Multi-agent Multi-armed Bandit", "abstract": "Multi-agent multi-armed bandit (MAMAB) is a classic collaborative learning model and has gained much attention in recent years. However, existing studies do not consider the case where an agent may refuse to share all her information with others, e.g., when some of the data contains personal privacy. In this paper, we propose a novel limited shared information multi-agent multi-armed bandit (LSI-MAMAB) model in which each agent only shares the information that she is willing to share, and propose the Balanced-ETC algorithm to help multiple agents collaborate efficiently with limited shared information. Our analysis shows that Balanced-ETC is asymptotically optimal and its average regret (on each agent) approaches a constant when there are sufficient agents involved. Moreover, to encourage agents to participate in this collaborative learning, an incentive mechanism is proposed to make sure each agent can benefit from the collaboration system. Finally, we present experimental results to validate our theoretical results.", "authors": ["Junning Shao", "Siwei Wang", "Zhixuan Fang"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-21", "url": "https://arxiv.org/abs/2502.15338", "pdf_url": "https://arxiv.org/pdf/2502.15338v1", "arxiv_id": "2502.15338", "doi": "10.48550/arXiv.2502.15338", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.0688}
{"id": "3f4abfcf4254a139d85441f285fc0e05df2590623eeefc9388ee7fb29ee6efa5", "sources": ["arxiv", "semantic_scholar"], "title": "Red-Teaming LLM Multi-Agent Systems via Communication Attacks", "abstract": "Large Language Model-based Multi-Agent Systems (LLM-MAS) have revolutionized complex problem-solving capability by enabling sophisticated agent collaboration through message-based communications. While the communication framework is crucial for agent coordination, it also introduces a critical yet unexplored security vulnerability. In this work, we introduce Agent-in-the-Middle (AiTM), a novel attack that exploits the fundamental communication mechanisms in LLM-MAS by intercepting and manipulating inter-agent messages. Unlike existing attacks that compromise individual agents, AiTM demonstrates how an adversary can compromise entire multi-agent systems by only manipulating the messages passing between agents. To enable the attack under the challenges of limited control and role-restricted communication format, we develop an LLM-powered adversarial agent with a reflection mechanism that generates contextually-aware malicious instructions. Our comprehensive evaluation across various frameworks, communication structures, and real-world applications demonstrates that LLM-MAS is vulnerable to communication-based attacks, highlighting the need for robust security measures in multi-agent systems.", "authors": ["Pengfei He", "Yupin Lin", "Shen Dong", "Han Xu", "Yue Xing", "Hui Liu"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-20", "url": "https://arxiv.org/abs/2502.14847", "pdf_url": "https://arxiv.org/pdf/2502.14847v2", "arxiv_id": "2502.14847", "doi": "10.48550/arXiv.2502.14847", "citation_count": 102, "influential_citation_count": 8, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.5032}
{"id": "94b14fb34dba3b0187abdd0bca3686b61d15302da7fc0de2db73c22cfdcfb5c0", "sources": ["arxiv", "semantic_scholar"], "title": "Beyond Self-Talk: A Communication-Centric Survey of LLM-Based Multi-Agent Systems", "abstract": "Large language model-based multi-agent systems have recently gained significant attention due to their potential for complex, collaborative, and intelligent problem-solving capabilities. Existing surveys typically categorize LLM-based multi-agent systems (LLM-MAS) according to their application domains or architectures, overlooking the central role of communication in coordinating agent behaviors and interactions. To address this gap, this paper presents a comprehensive survey of LLM-MAS from a communication-centric perspective. Specifically, we propose a structured framework that integrates system-level communication (architecture, goals, and protocols) with system internal communication (strategies, paradigms, objects, and content), enabling a detailed exploration of how agents interact, negotiate, and achieve collective intelligence. Through an extensive analysis of recent literature, we identify key components in multiple dimensions and summarize their strengths and limitations. In addition, we highlight current challenges, including communication efficiency, security vulnerabilities, inadequate benchmarking, and scalability issues, and outline promising future research directions. This review aims to help researchers and practitioners gain a clear understanding of the communication mechanisms in LLM-MAS, thereby facilitating the design and deployment of robust, scalable, and secure multi-agent systems.", "authors": ["Bingyu Yan", "Zhibo Zhou", "Litian Zhang", "Lian Zhang", "Ziyi Zhou", "Dezhuang Miao", "Zhoujun Li", "Chaozhuo Li", "Xiaoming Zhang"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-20", "url": "https://arxiv.org/abs/2502.14321", "pdf_url": "https://arxiv.org/pdf/2502.14321v3", "arxiv_id": "2502.14321", "doi": "10.1007/s11704-026-50857-y", "citation_count": 67, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4581}
{"id": "b3260c6c8148ab917de44aa0ab734cc427f94094ba4ac7e0b8ca4a5a03b300ee", "sources": ["arxiv", "semantic_scholar"], "title": "Advancing Language Multi-Agent Learning with Credit Re-Assignment for Interactive Environment Generalization", "abstract": "LLM-based agents have made significant advancements in interactive environments, such as mobile operations and web browsing, and other domains beyond computer using. Current multi-agent systems universally excel in performance, compared to single agents, but struggle with generalization across environments due to predefined roles and inadequate strategies for generalizing language agents. The challenge of achieving both strong performance and good generalization has hindered the progress of multi-agent systems for interactive environments. To address these issues, we propose CollabUIAgents, a multi-agent reinforcement learning framework with a novel multi-agent credit re-assignment (CR) strategy, assigning process rewards with LLMs rather than environment-specific rewards and learning with synthesized preference data, in order to foster generalizable, collaborative behaviors among the role-free agents' policies. Empirical results show that our framework improves both performance and cross-environment generalizability of multi-agent systems. Moreover, our 7B-parameter system achieves results on par with or exceed strong closed-source models, and the LLM that guides the CR. We also provide insights in using granular CR rewards effectively for environment generalization, and accommodating trained LLMs in multi-agent systems.", "authors": ["Zhitao He", "Zijun Liu", "Peng Li", "Yi R. Fung", "Ming Yan", "Ji Zhang", "Fei Huang", "Yang Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-20", "url": "https://arxiv.org/abs/2502.14496", "pdf_url": "https://arxiv.org/pdf/2502.14496v3", "arxiv_id": "2502.14496", "doi": null, "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2698}
{"id": "374dd80612ef422fd8e519fb1ee879b1d153efb00cf342a43a3ca771cb4e889b", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Perspective on Modern Information Retrieval", "abstract": "The rise of large language models (LLMs) has introduced a new era in information retrieval (IR), where queries and documents that were once assumed to be generated exclusively by humans can now also be created by automated agents. These agents can formulate queries, generate documents, and perform ranking. This shift challenges some long-standing IR paradigms and calls for a reassessment of both theoretical frameworks and practical methodologies. We advocate for a multi-agent perspective to better capture the complex interactions between query agents, document agents, and ranker agents. Through empirical exploration of various multi-agent retrieval settings, we reveal the significant impact of these interactions on system performance. Our findings underscore the need to revisit classical IR paradigms and develop new frameworks for more effective modeling and evaluation of modern retrieval systems.", "authors": ["Haya Nachimovsky", "Moshe Tennenholtz", "Oren Kurland"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-20", "url": "https://arxiv.org/abs/2502.14796", "pdf_url": "https://arxiv.org/pdf/2502.14796v1", "arxiv_id": "2502.14796", "doi": "10.48550/arXiv.2502.14796", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "34b54c5f0d82f6701a468bcdb024f0f31a3177354905dcfa781bc0f15da6dae5", "sources": ["arxiv", "semantic_scholar"], "title": "Simulating Cooperative Prosocial Behavior with Multi-Agent LLMs: Evidence and Mechanisms for AI Agents to Inform Policy Decisions", "abstract": "Human prosocial cooperation is essential for our collective health, education, and welfare. However, designing social systems to maintain or incentivize prosocial behavior is challenging because people can act selfishly to maximize personal gain. This complex and unpredictable aspect of human behavior makes it difficult for policymakers to foresee the implications of their designs. Recently, multi-agent LLM systems have shown remarkable capabilities in simulating human-like behavior, and replicating some human lab experiments. This paper studies how well multi-agent systems can simulate prosocial human behavior, such as that seen in the public goods game (PGG), and whether multi-agent systems can exhibit ``unbounded actions'' seen outside the lab in real world scenarios. We find that multi-agent LLM systems successfully replicate human behavior from lab experiments of the public goods game with three experimental treatments - priming, transparency, and varying endowments. Beyond replicating existing experiments, we find that multi-agent LLM systems can replicate the expected human behavior when combining experimental treatments, even if no previous study combined those specific treatments. Lastly, we find that multi-agent systems can exhibit a rich set of unbounded actions that people do in the real world outside of the lab -- such as collaborating and even cheating. In sum, these studies are steps towards a future where LLMs can be used to inform policy decisions that encourage people to act in a prosocial manner.", "authors": ["Karthik Sreedhar", "Alice Cai", "Jenny Ma", "Jeffrey V. Nickerson", "Lydia B. Chilton"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-18", "url": "https://arxiv.org/abs/2502.12504", "pdf_url": "https://arxiv.org/pdf/2502.12504v1", "arxiv_id": "2502.12504", "doi": "10.1145/3708359.3712149", "citation_count": 29, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Conference on Intelligent User Interfaces", "quality_score": 0.3693}
{"id": "7e1522dc91bd17ffd7196d7ff3a0660ba789fc3d3a63cfe84370a68a0c6d11fa", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Agents Making Agent Tools", "abstract": "Tool use has turned large language models (LLMs) into powerful agents that can perform complex multi-step tasks by dynamically utilising external software components. However, these tools must be implemented in advance by human developers, hindering the applicability of LLM agents in domains demanding large numbers of highly specialised tools, like in life sciences and medicine. Motivated by the growing trend of scientific studies accompanied by public code repositories, we propose ToolMaker, an agentic framework that autonomously transforms papers with code into LLM-compatible tools. Given a GitHub URL and short task description, ToolMaker autonomously installs dependencies and generates code to perform the task, using a closed-loop self-correction mechanism for debugging. To evaluate our approach, we introduce a benchmark comprising 15 complex computational tasks spanning various domains with over 100 unit tests to assess correctness and robustness. Our method correctly implements 80% of the tasks, substantially outperforming current state-of-the-art software engineering agents. ToolMaker therefore is a step towards fully autonomous agent-based scientific workflows. Our code and benchmark are publicly available at https://github.com/KatherLab/ToolMaker.", "authors": ["Georg Wölflein", "Dyke Ferber", "Daniel Truhn", "Ognjen Arandjelović", "Jakob Nikolas Kather"], "categories": ["cs.CL", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-17", "url": "https://arxiv.org/abs/2502.11705", "pdf_url": "https://arxiv.org/pdf/2502.11705v2", "arxiv_id": "2502.11705", "doi": "10.48550/arXiv.2502.11705", "citation_count": 48, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/KatherLab/ToolMaker", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4225}
{"id": "15104a29306565aeee44f8192ce7d7d70787da72b0f08d19819d524b9edb6ea8", "sources": ["arxiv", "semantic_scholar"], "title": "HARBOR: Exploring Persona Dynamics in Multi-Agent Competition", "abstract": "We investigate factors contributing to LLM agents' success in competitive multi-agent environments, using auctions as a testbed where agents bid to maximize profit. The agents are equipped with bidding domain knowledge, distinct personas that reflect item preferences, and a memory of auction history. Our work extends the classic auction scenario by creating a realistic environment where multiple agents bid on houses, weighing aspects such as size, location, and budget to secure the most desirable homes at the lowest prices. Particularly, we investigate three key questions: (a) How does a persona influence an agent's behavior in a competitive setting? (b) Can an agent effectively profile its competitors' behavior during auctions? (c) How can persona profiling be leveraged to create an advantage using strategies such as theory of mind? Through a series of experiments, we analyze the behaviors of LLM agents and shed light on new findings. Our testbed, called HARBOR, offers a valuable platform for deepening our understanding of multi-agent workflows in competitive environments.", "authors": ["Kenan Jiang", "Li Xiong", "Fei Liu"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-17", "url": "https://arxiv.org/abs/2502.12149", "pdf_url": "https://arxiv.org/pdf/2502.12149v2", "arxiv_id": "2502.12149", "doi": "10.48550/arXiv.2502.12149", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2113}
{"id": "45d36d34fb3d80c69b01cebbe7117c7ca622a726944b9876a63b471402225278", "sources": ["arxiv", "semantic_scholar"], "title": "Nuclear Deployed: Analyzing Catastrophic Risks in Decision-making of Autonomous LLM Agents", "abstract": "Large language models (LLMs) are evolving into autonomous decision-makers, raising concerns about catastrophic risks in high-stakes scenarios, particularly in Chemical, Biological, Radiological and Nuclear (CBRN) domains. Based on the insight that such risks can originate from trade-offs between the agent's Helpful, Harmlessness and Honest (HHH) goals, we build a novel three-stage evaluation framework, which is carefully constructed to effectively and naturally expose such risks. We conduct 14,400 agentic simulations across 12 advanced LLMs, with extensive experiments and analysis. Results reveal that LLM agents can autonomously engage in catastrophic behaviors and deception, without being deliberately induced. Furthermore, stronger reasoning abilities often increase, rather than mitigate, these risks. We also show that these agents can violate instructions and superior commands. On the whole, we empirically prove the existence of catastrophic risks in autonomous LLM agents. We release our code to foster further research.", "authors": ["Rongwu Xu", "Xiaojian Li", "Shuo Chen", "Wei Xu"], "categories": ["cs.CL", "cs.AI", "cs.CR", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-17", "url": "https://arxiv.org/abs/2502.11355", "pdf_url": "https://arxiv.org/pdf/2502.11355v3", "arxiv_id": "2502.11355", "doi": "10.48550/arXiv.2502.11355", "citation_count": 25, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/pillowsofwind/LLM-CBRN-Risks", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3537}
{"id": "7a56c8b2068990eb532952ad8d0bc3e86faf70cc27324ee2b0d2e902d4091c8b", "sources": ["arxiv", "semantic_scholar"], "title": "Talk Structurally, Act Hierarchically: A Collaborative Framework for LLM Multi-Agent Systems", "abstract": "Recent advancements in LLM-based multi-agent (LLM-MA) systems have shown promise, yet significant challenges remain in managing communication and refinement when agents collaborate on complex tasks. In this paper, we propose \\textit{Talk Structurally, Act Hierarchically (TalkHier)}, a novel framework that introduces a structured communication protocol for context-rich exchanges and a hierarchical refinement system to address issues such as incorrect outputs, falsehoods, and biases. \\textit{TalkHier} surpasses various types of SoTA, including inference scaling model (OpenAI-o1), open-source multi-agent models (e.g., AgentVerse), and majority voting strategies on current LLM and single-agent baselines (e.g., ReAct, GPT4o), across diverse tasks, including open-domain question answering, domain-specific selective questioning, and practical advertisement text generation. These results highlight its potential to set a new standard for LLM-MA systems, paving the way for more effective, adaptable, and collaborative multi-agent frameworks. The code is available https://github.com/sony/talkhier.", "authors": ["Zhao Wang", "Sota Moriyama", "Wei-Yao Wang", "Briti Gangopadhyay", "Shingo Takamatsu"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-16", "url": "https://arxiv.org/abs/2502.11098", "pdf_url": "https://arxiv.org/pdf/2502.11098v1", "arxiv_id": "2502.11098", "doi": "10.48550/arXiv.2502.11098", "citation_count": 17, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/sony/talkhier", "venue": "arXiv.org", "quality_score": 0.3138}
{"id": "0ce30cb3b0b816e26c00d99aae49fda80f85f0f2195ed4cf8e393304b68dcb83", "sources": ["arxiv", "semantic_scholar"], "title": "MasRouter: Learning to Route LLMs for Multi-Agent Systems", "abstract": "Multi-agent systems (MAS) powered by Large Language Models (LLMs) have been demonstrated to push the boundaries of LLM capabilities, yet they often incur significant costs and face challenges in dynamic LLM selection. Current LLM routing methods effectively reduce overhead in single-agent scenarios by customizing LLM selection for each query, but they overlook the critical decisions regarding collaboration modes and agent roles in MAS. In response to this challenge, we first introduce the problem of Multi-Agent System Routing (MASR), which integrates all components of MAS into a unified routing framework. Toward this goal, we propose MasRouter, the first high-performing, cost-effective, and inductive MASR solution. MasRouter employs collaboration mode determination, role allocation, and LLM routing through a cascaded controller network, progressively constructing a MAS that balances effectiveness and efficiency. Extensive experiments demonstrate that MasRouter is (1) high-performing, achieving a $1.8\\%\\sim8.2\\%$ improvement over the state-of-the-art method on MBPP; (2) economical, reducing overhead by up to $52.07\\%$ compared to SOTA methods on HumanEval; and (3) plug-and-play, seamlessly integrating with mainstream MAS frameworks, reducing overhead by $17.21\\%\\sim28.17\\%$ via customized routing. The code is available at https://github.com/yanweiyue/masrouter.", "authors": ["Yanwei Yue", "Guibin Zhang", "Boyang Liu", "Guancheng Wan", "Kun Wang", "Dawei Cheng", "Yiyan Qi"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-16", "url": "https://arxiv.org/abs/2502.11133", "pdf_url": "https://arxiv.org/pdf/2502.11133v1", "arxiv_id": "2502.11133", "doi": "10.48550/arXiv.2502.11133", "citation_count": 59, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/yanweiyue/masrouter", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4445}
{"id": "93a514bd42aca3f95877ddf48e14e4c1c180b0f29e8d5966ba2f9ec016eb7f24", "sources": ["arxiv", "semantic_scholar"], "title": "G-Safeguard: A Topology-Guided Security Lens and Treatment on LLM-based Multi-agent Systems", "abstract": "Large Language Model (LLM)-based Multi-agent Systems (MAS) have demonstrated remarkable capabilities in various complex tasks, ranging from collaborative problem-solving to autonomous decision-making. However, as these systems become increasingly integrated into critical applications, their vulnerability to adversarial attacks, misinformation propagation, and unintended behaviors have raised significant concerns. To address this challenge, we introduce G-Safeguard, a topology-guided security lens and treatment for robust LLM-MAS, which leverages graph neural networks to detect anomalies on the multi-agent utterance graph and employ topological intervention for attack remediation. Extensive experiments demonstrate that G-Safeguard: (I) exhibits significant effectiveness under various attack strategies, recovering over 40% of the performance for prompt injection; (II) is highly adaptable to diverse LLM backbones and large-scale MAS; (III) can seamlessly combine with mainstream MAS with security guarantees. The code is available at https://github.com/wslong20/G-safeguard.", "authors": ["Shilong Wang", "Guibin Zhang", "Miao Yu", "Guancheng Wan", "Fanci Meng", "Chongye Guo", "Kun Wang", "Yang Wang"], "categories": ["cs.CR", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-16", "url": "https://arxiv.org/abs/2502.11127", "pdf_url": "https://arxiv.org/pdf/2502.11127v1", "arxiv_id": "2502.11127", "doi": "10.48550/arXiv.2502.11127", "citation_count": 45, "influential_citation_count": 7, "has_code": true, "code_url": "https://github.com/wslong20/G-safeguard", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4515}
{"id": "6c21f27e97111513aeabfc36c496db78a3330c67559bbf3c48d8ecdc1b251db9", "sources": ["arxiv", "semantic_scholar"], "title": "Attention Mechanism for LLM-based Agents Dynamic Diffusion under Information Asymmetry", "abstract": "Large language models have been used to simulate human society using multi-agent systems. Most current social simulation research emphasizes interactive behaviors in fixed environments, ignoring information opacity, relationship variability, and diffusion diversity. In this paper, we first propose a general framework for exploring multi-agent information diffusion. We identified LLMs' deficiency in the perception and utilization of social relationships, as well as diverse actions. Then, we designed a dynamic attention mechanism to help agents allocate attention to different information, addressing the limitations of the LLM attention mechanism. Agents start by responding to external information stimuli within a five-agent group, increasing group size and forming information circles while developing relationships and sharing information. Additionally, we explore the information diffusion features in the asymmetric open environment by observing the evolution of information gaps, diffusion patterns, and the accumulation of social capital, which are closely linked to psychological, sociological, and communication theories.", "authors": ["Yiwen Zhang", "Yifu Wu", "Wenyue Hua", "Xiang Lu", "Xuming Hu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-16", "url": "https://arxiv.org/abs/2502.13160", "pdf_url": "https://arxiv.org/pdf/2502.13160v3", "arxiv_id": "2502.13160", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0753}
{"id": "3423001eb65eb4d3bdd1ca6d95b0e0b4c8a3f82f4c988cfbcc6517c1e4280091", "sources": ["arxiv", "semantic_scholar"], "title": "Divergent Thoughts toward One Goal: LLM-based Multi-Agent Collaboration System for Electronic Design Automation", "abstract": "Recently, with the development of tool-calling capabilities in large language models (LLMs), these models have demonstrated significant potential for automating electronic design automation (EDA) flows by interacting with EDA tool APIs via EDA scripts. However, considering the limited understanding of EDA tools, LLMs face challenges in practical scenarios where diverse interfaces of EDA tools exist across different platforms. Additionally, EDA flow automation often involves intricate, long-chain tool-calling processes, increasing the likelihood of errors in intermediate steps. Any errors will lead to the instability and failure of EDA flow automation. To address these challenges, we introduce EDAid, a multi-agent collaboration system where multiple agents harboring divergent thoughts converge towards a common goal, ensuring reliable and successful EDA flow automation. Specifically, each agent is controlled by ChipLlama models, which are expert LLMs fine-tuned for EDA flow automation. Our experiments demonstrate the state-of-the-art (SOTA) performance of our ChipLlama models and validate the effectiveness of our EDAid in the automation of complex EDA flows, showcasing superior performance compared to single-agent systems.", "authors": ["Haoyuan Wu", "Haisheng Zheng", "Zhuolun He", "Bei Yu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-15", "url": "https://arxiv.org/abs/2502.10857", "pdf_url": "https://arxiv.org/pdf/2502.10857v1", "arxiv_id": "2502.10857", "doi": "10.48550/arXiv.2502.10857", "citation_count": 18, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.3197}
{"id": "13a772040d120487cf472352e59a5733b100b66325d997fa3cf4eb2d2a0c548f", "sources": ["arxiv", "semantic_scholar"], "title": "AgentGuard: Repurposing Agentic Orchestrator for Safety Evaluation of Tool Orchestration", "abstract": "The integration of tool use into large language models (LLMs) enables agentic systems with real-world impact. In the meantime, unlike standalone LLMs, compromised agents can execute malicious workflows with more consequential impact, signified by their tool-use capability. We propose AgentGuard, a framework to autonomously discover and validate unsafe tool-use workflows, followed by generating safety constraints to confine the behaviors of agents, achieving the baseline of safety guarantee at deployment. AgentGuard leverages the LLM orchestrator's innate capabilities - knowledge of tool functionalities, scalable and realistic workflow generation, and tool execution privileges - to act as its own safety evaluator. The framework operates through four phases: identifying unsafe workflows, validating them in real-world execution, generating safety constraints, and validating constraint efficacy. The output, an evaluation report with unsafe workflows, test cases, and validated constraints, enables multiple security applications. We empirically demonstrate AgentGuard's feasibility with experiments. With this exploratory work, we hope to inspire the establishment of standardized testing and hardening procedures for LLM agents to enhance their trustworthiness in real-world applications.", "authors": ["Jizhou Chen", "Samuel Lee Cong"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-13", "url": "https://arxiv.org/abs/2502.09809", "pdf_url": "https://arxiv.org/pdf/2502.09809v1", "arxiv_id": "2502.09809", "doi": "10.48550/arXiv.2502.09809", "citation_count": 21, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3356}
{"id": "b96b47039149cff3b1b815499ffac122c3935b6465e142a2ff2797d1016dd97f", "sources": ["arxiv", "semantic_scholar"], "title": "Single-Agent Planning in a Multi-Agent System: A Unified Framework for Type-Based Planners", "abstract": "We consider a general problem where an agent is in a multi-agent environment and must plan for herself without any prior information about her opponents. At each moment, this pivotal agent is faced with a trade-off between exploiting her currently accumulated information about the other agents and exploring further to improve future (re-)planning. We propose a theoretic framework that unifies a spectrum of planners for the pivotal agent to address this trade-off. The planner at one end of this spectrum aims to find exact solutions, while those towards the other end yield approximate solutions as the problem scales up. Beyond theoretical analysis, we also implement \\textbf{13} planners and conduct experiments in a specific domain called \\textit{multi-agent route planning} with the number of agents \\textbf{up to~50}, to compare their performaces in various scenarios. One interesting observation comes from a class of planners that we call \\textit{safe-agents} and their enhanced variants by incorporating domain-specific knowledge, which is a simple special case under the proposed general framework, but performs sufficiently well in most cases. Our unified framework, as well as those induced planners, provides new insights on multi-agent decision-making, with potential applications to related areas such as mechanism design.", "authors": ["Fengming Zhu", "Fangzhen Lin"], "categories": ["cs.MA", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-13", "url": "https://arxiv.org/abs/2502.08950", "pdf_url": "https://arxiv.org/pdf/2502.08950v1", "arxiv_id": "2502.08950", "doi": "10.48550/arXiv.2502.08950", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1747}
{"id": "52619101971de9e63476aaf7b2c6c76177a6066047494bba1a4175ee671a5e45", "sources": ["arxiv", "semantic_scholar"], "title": "A New Query Expansion Approach via Agent-Mediated Dialogic Inquiry", "abstract": "Query expansion is widely used in Information Retrieval (IR) to improve search outcomes by supplementing initial queries with richer information. While recent Large Language Model (LLM) based methods generate pseudo-relevant content and expanded terms via multiple prompts, they often yield homogeneous, narrow expansions that lack the diverse context needed to retrieve relevant information. In this paper, we propose AMD: a new Agent-Mediated Dialogic Framework that engages in a dialogic inquiry involving three specialized roles: (1) a Socratic Questioning Agent reformulates the initial query into three sub-questions, with each question inspired by a specific Socratic questioning dimension, including clarification, assumption probing, and implication probing, (2) a Dialogic Answering Agent generates pseudo-answers, enriching the query representation with multiple perspectives aligned to the user's intent, and (3) a Reflective Feedback Agent evaluates and refines these pseudo-answers, ensuring that only the most relevant and informative content is retained. By leveraging a multi-agent process, AMD effectively crafts richer query representations through inquiry and feedback refinement. Extensive experiments on benchmarks including BEIR and TREC demonstrate that our framework outperforms previous methods, offering a robust solution for retrieval tasks.", "authors": ["Wonduk Seo", "Hyunjin An", "Seunghyun Lee"], "categories": ["cs.IR", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-12", "url": "https://arxiv.org/abs/2502.08557", "pdf_url": "https://arxiv.org/pdf/2502.08557v3", "arxiv_id": "2502.08557", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "945af1cbeff79ddce6f5404bd31bf507a116764597ae6bb539c4cfb69c87bf02", "sources": ["arxiv", "semantic_scholar"], "title": "Flow-of-Action: SOP Enhanced LLM-Based Multi-Agent System for Root Cause Analysis", "abstract": "In the realm of microservices architecture, the occurrence of frequent incidents necessitates the employment of Root Cause Analysis (RCA) for swift issue resolution. It is common that a serious incident can take several domain experts hours to identify the root cause. Consequently, a contemporary trend involves harnessing Large Language Models (LLMs) as automated agents for RCA. Though the recent ReAct framework aligns well with the Site Reliability Engineers (SREs) for its thought-action-observation paradigm, its hallucinations often lead to irrelevant actions and directly affect subsequent results. Additionally, the complex and variable clues of the incident can overwhelm the model one step further. To confront these challenges, we propose Flow-of-Action, a pioneering Standard Operation Procedure (SOP) enhanced LLM-based multi-agent system. By explicitly summarizing the diagnosis steps of SREs, SOP imposes constraints on LLMs at crucial junctures, guiding the RCA process towards the correct trajectory. To facilitate the rational and effective utilization of SOPs, we design an SOP-centric framework called SOP flow. SOP flow contains a series of tools, including one for finding relevant SOPs for incidents, another for automatically generating SOPs for incidents without relevant ones, and a tool for converting SOPs into code. This significantly alleviates the hallucination issues of ReAct in RCA tasks. We also design multiple auxiliary agents to assist the main agent by removing useless noise, narrowing the search space, and informing the main agent whether the RCA procedure can stop. Compared to the ReAct method's 35.50% accuracy, our Flow-of-Action method achieves 64.01%, meeting the accuracy requirements for RCA in real-world systems.", "authors": ["Changhua Pei", "Zexin Wang", "Fengrui Liu", "Zeyan Li", "Yang Liu", "Xiao He", "Rong Kang", "Tieying Zhang", "Jianjun Chen", "Jianhui Li", "Gaogang Xie", "Dan Pei"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-12", "url": "https://arxiv.org/abs/2502.08224", "pdf_url": "https://arxiv.org/pdf/2502.08224v1", "arxiv_id": "2502.08224", "doi": "10.1145/3701716.3715225", "citation_count": 31, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "The Web Conference", "quality_score": 0.3763}
{"id": "21431f43392f37e67a9939309b8a0c2b8035cb1f8a0f09683c2f0b98ef0b9ce3", "sources": ["arxiv", "semantic_scholar"], "title": "Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model", "abstract": "Large Language Models (LLMs) with API-calling capabilities enabled building effective Language Agents (LA), while also revolutionizing the conventional task-oriented dialogue (TOD) paradigm. However, current approaches face a critical dilemma: TOD systems are often trained on a limited set of target APIs, requiring new data to maintain their quality when interfacing with new services, while LAs are not trained to maintain user intent over multi-turn conversations. Because both robust multi-turn management and advanced function calling are crucial for effective conversational agents, we evaluate these skills on three popular benchmarks: MultiWOZ 2.4 (TOD), BFCL V3 (LA), and API-Bank (LA), and our analyses reveal that specialized approaches excel in one domain but underperform in the other. To bridge this chasm, we introduce CoALM (Conversational Agentic Language Model), a unified approach that integrates both conversational and agentic capabilities. We created CoALM-IT, a carefully constructed multi-task dataset that interleave multi-turn ReAct reasoning with complex API usage. Using CoALM-IT, we train three models CoALM 8B, CoALM 70B, and CoALM 405B, which outperform top domain-specific models, including GPT-4o, across all three benchmarks. This demonstrates the feasibility of a single model approach for both TOD and LA, setting a new standard for conversational agents.", "authors": ["Emre Can Acikgoz", "Jeremiah Greer", "Akul Datta", "Ze Yang", "William Zeng", "Oussama Elachqar", "Emmanouil Koukoumidis", "Dilek Hakkani-Tür", "Gokhan Tur"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-12", "url": "https://arxiv.org/abs/2502.08820", "pdf_url": "https://arxiv.org/pdf/2502.08820v3", "arxiv_id": "2502.08820", "doi": "10.48550/arXiv.2502.08820", "citation_count": 24, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3495}
{"id": "193db083ff474d0999ba84203cb347f21b5cb3f7c7b6d114a3d0cf016c311962", "sources": ["arxiv", "semantic_scholar"], "title": "Don't Just Demo, Teach Me the Principles: A Principle-Based Multi-Agent Prompting Strategy for Text Classification", "abstract": "We present PRINCIPLE-BASED PROMPTING, a simple but effective multi-agent prompting strategy for text classification. It first asks multiple LLM agents to independently generate candidate principles based on analysis of demonstration samples with or without labels, consolidates them into final principles via a finalizer agent, and then sends them to a classifier agent to perform downstream classification tasks. Extensive experiments on binary and multi-class classification datasets with different sizes of LLMs show that our approach not only achieves substantial performance gains (1.55% - 19.37%) over zero-shot prompting on macro-F1 score but also outperforms other strong baselines (CoT and stepback prompting). Principles generated by our approach help LLMs perform better on classification tasks than human crafted principles on two private datasets. Our multi-agent PRINCIPLE-BASED PROMPTING approach also shows on-par or better performance compared to demonstration-based few-shot prompting approaches, yet with substantially lower inference costs. Ablation studies show that label information and the multi-agent cooperative LLM framework play an important role in generating high-quality principles to facilitate downstream classification tasks.", "authors": ["Peipei Wei", "Dimitris Dimitriadis", "Yan Xu", "Mingwei Shen"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-11", "url": "https://arxiv.org/abs/2502.07165", "pdf_url": "https://arxiv.org/pdf/2502.07165v1", "arxiv_id": "2502.07165", "doi": "10.48550/arXiv.2502.07165", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "c4e05a56317e7aa0df397ac4b713989894c8ff81f6c30d9126394e0dbbabcf97", "sources": ["arxiv", "semantic_scholar"], "title": "Distributed Approach to Haskell Based Applications Refactoring with LLMs Based Multi-Agent Systems", "abstract": "We present a large language models (LLMs) based multi-agent system to automate the refactoring of Haskell codebases. The multi-agent system consists of specialized agents performing tasks such as context analysis, refactoring, validation, and testing. Refactoring improvements are using metrics such as cyclomatic complexity, run-time, and memory allocation. Experimental evaluations conducted on Haskell codebases demonstrate improvements in code quality. Cyclomatic complexity was reduced by 13.64% and 47.06% in the respective codebases. Memory allocation improved by 4.17% and 41.73%, while runtime efficiency increased by up to 50%. These metrics highlight the systems ability to optimize Haskells functional paradigms while maintaining correctness and scalability. Results show reductions in complexity and performance enhancements across codebases. The integration of LLMs based multi-agent system enables precise task execution and inter-agent collaboration, addressing the challenges of refactoring in functional programming. This approach aims to address the challenges of refactoring functional programming languages through distributed and modular systems.", "authors": ["Shahbaz Siddeeq", "Zeeshan Rasheed", "Malik Abdul Sami", "Mahade Hasan", "Muhammad Waseem", "Jussi Rasku", "Mika Saari", "Kai-Kristian Kemell", "Pekka Abrahamsson"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-11", "url": "https://arxiv.org/abs/2502.07928", "pdf_url": "https://arxiv.org/pdf/2502.07928v1", "arxiv_id": "2502.07928", "doi": "10.48550/arXiv.2502.07928", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "316814bfafc8805cb5afc57cc6281a55207be8cd90e046d6ce359cc920320a0f", "sources": ["arxiv", "semantic_scholar"], "title": "Fairness in Agentic AI: A Unified Framework for Ethical and Equitable Multi-Agent System", "abstract": "Ensuring fairness in decentralized multi-agent systems presents significant challenges due to emergent biases, systemic inefficiencies, and conflicting agent incentives. This paper provides a comprehensive survey of fairness in multi-agent AI, introducing a novel framework where fairness is treated as a dynamic, emergent property of agent interactions. The framework integrates fairness constraints, bias mitigation strategies, and incentive mechanisms to align autonomous agent behaviors with societal values while balancing efficiency and robustness. Through empirical validation, we demonstrate that incorporating fairness constraints results in more equitable decision-making. This work bridges the gap between AI ethics and system design, offering a foundation for accountable, transparent, and socially responsible multi-agent AI systems.", "authors": ["Rajesh Ranjan", "Shailja Gupta", "Surya Narayan Singh"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-11", "url": "https://arxiv.org/abs/2502.07254", "pdf_url": "https://arxiv.org/pdf/2502.07254v2", "arxiv_id": "2502.07254", "doi": "10.48550/arXiv.2502.07254", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "fa81aa8ada8dc2144478bca5d3d7d068ad3e6d9ea08d86e23cccabaa453b252f", "sources": ["arxiv", "semantic_scholar"], "title": "Approximating Human Strategic Reasoning with LLM-Enhanced Recursive Reasoners Leveraging Multi-agent Hypergames", "abstract": "LLM-driven multi-agent-based simulations have been gaining traction with applications in game-theoretic and social simulations. While most implementations seek to exploit or evaluate LLM-agentic reasoning, they often do so with a weak notion of agency and simplified architectures. We implement a role-based multi-agent strategic interaction framework tailored to sophisticated recursive reasoners, providing the means for systematic in-depth development and evaluation of strategic reasoning. Our game environment is governed by the umpire responsible for facilitating games, from matchmaking through move validation to environment management. Players incorporate state-of-the-art LLMs in their decision mechanism, relying on a formal hypergame-based model of hierarchical beliefs. We use one-shot, 2-player beauty contests to evaluate the recursive reasoning capabilities of the latest LLMs, providing a comparison to an established baseline model from economics and data from human experiments. Furthermore, we introduce the foundations of an alternative semantic measure of reasoning to the k-level theory. Our experiments show that artificial reasoners can outperform the baseline model in terms of both approximating human behaviour and reaching the optimal solution.", "authors": ["Vince Trencsenyi", "Agnieszka Mensfelt", "Kostas Stathis"], "categories": ["cs.AI", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-11", "url": "https://arxiv.org/abs/2502.07443", "pdf_url": "https://arxiv.org/pdf/2502.07443v1", "arxiv_id": "2502.07443", "doi": "10.1007/978-3-032-16328-8_2", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "mAbs", "quality_score": 0.2258}
{"id": "38b05638fa83df4b8bbe98d8f7a9d3d20938ee9191d9cefb64b927600a3f4760", "sources": ["arxiv", "semantic_scholar"], "title": "KARMA: Leveraging Multi-Agent LLMs for Automated Knowledge Graph Enrichment", "abstract": "Maintaining comprehensive and up-to-date knowledge graphs (KGs) is critical for modern AI systems, but manual curation struggles to scale with the rapid growth of scientific literature. This paper presents KARMA, a novel framework employing multi-agent large language models (LLMs) to automate KG enrichment through structured analysis of unstructured text. Our approach employs nine collaborative agents, spanning entity discovery, relation extraction, schema alignment, and conflict resolution that iteratively parse documents, verify extracted knowledge, and integrate it into existing graph structures while adhering to domain-specific schema. Experiments on 1,200 PubMed articles from three different domains demonstrate the effectiveness of KARMA in knowledge graph enrichment, with the identification of up to 38,230 new entities while achieving 83.1\\% LLM-verified correctness and reducing conflict edges by 18.6\\% through multi-layer assessments.", "authors": ["Yuxing Lu", "Wei Wu", "Xukai Zhao", "Rui Peng", "Jinzhuo Wang"], "categories": ["cs.CL", "cs.AI", "cs.CE", "cs.DL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-10", "url": "https://arxiv.org/abs/2502.06472", "pdf_url": "https://arxiv.org/pdf/2502.06472v2", "arxiv_id": "2502.06472", "doi": "10.48550/arXiv.2502.06472", "citation_count": 27, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3618}
{"id": "b5e0c092a55a1e68f02bda37205b9c9aa67b3756995623d62e79b33e2c8f9b5d", "sources": ["arxiv", "semantic_scholar"], "title": "Preventing Rogue Agents Improves Multi-Agent Collaboration", "abstract": "Multi-agent systems, where specialized agents collaborate to solve a shared task hold great potential, from increased modularity to simulating complex environments. However, they also have a major caveat -- a single agent can cause the entire system to fail. Consider a simple game where the knowledge to solve the task is distributed between agents, which share information in a communication channel. At each round, any of the agents can terminate the game and make the final prediction, even if they are uncertain about the outcome of their action. Detection of such rogue agents before they act may prevent the system's failure. In this work, we propose to monitor agents during action prediction and intervene when a future error is likely to occur. To test our approach, we introduce WhoDunitEnv, a multi-agent collaboration environment that allows modular control over task complexity and communication structure. Experiments on WhoDunitEnv, code generation tasks and the GovSim environment for resource sustainability show that our approach leads to substantial performance gains up to 17.4%, 2.5% and 20%, respectively. Thorough analysis shows that our monitors successfully identify critical points of agent confusion and our interventions effectively stop agent errors from propagating.", "authors": ["Ohav Barbi", "Ori Yoran", "Mor Geva"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-09", "url": "https://arxiv.org/abs/2502.05986", "pdf_url": "https://arxiv.org/pdf/2502.05986v2", "arxiv_id": "2502.05986", "doi": "10.48550/arXiv.2502.05986", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "a33176ba58cd2c12a53eef3db7392d917090aeb3ab75dd30207bf2fb547ddd6f", "sources": ["arxiv", "semantic_scholar"], "title": "AutoAgent: A Fully-Automated and Zero-Code Framework for LLM Agents", "abstract": "Large Language Model (LLM) Agents have demonstrated remarkable capabilities in task automation and intelligent decision-making, driving the widespread adoption of agent development frameworks such as LangChain and AutoGen. However, these frameworks predominantly serve developers with extensive technical expertise - a significant limitation considering that only 0.03 % of the global population possesses the necessary programming skills. This stark accessibility gap raises a fundamental question: Can we enable everyone, regardless of technical background, to build their own LLM agents using natural language alone? To address this challenge, we introduce AutoAgent-a Fully-Automated and highly Self-Developing framework that enables users to create and deploy LLM agents through Natural Language Alone. Operating as an autonomous Agent Operating System, AutoAgent comprises four key components: i) Agentic System Utilities, ii) LLM-powered Actionable Engine, iii) Self-Managing File System, and iv) Self-Play Agent Customization module. This lightweight yet powerful system enables efficient and dynamic creation and modification of tools, agents, and workflows without coding requirements or manual intervention. Beyond its code-free agent development capabilities, AutoAgent also serves as a versatile multi-agent system for General AI Assistants. Comprehensive evaluations on the GAIA benchmark demonstrate AutoAgent's effectiveness in generalist multi-agent tasks, surpassing existing state-of-the-art methods. Furthermore, AutoAgent's Retrieval-Augmented Generation (RAG)-related capabilities have shown consistently superior performance compared to many alternative LLM-based solutions.", "authors": ["Jiabin Tang", "Tianyu Fan", "Chao Huang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-09", "url": "https://arxiv.org/abs/2502.05957", "pdf_url": "https://arxiv.org/pdf/2502.05957v3", "arxiv_id": "2502.05957", "doi": "10.48550/arXiv.2502.05957", "citation_count": 57, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/HKUDS/AutoAgent", "venue": "arXiv.org", "quality_score": 0.4409}
{"id": "bcccaf0e9ce5b037077415f7419567bb57736ba1c1fc969bb9160b818ed22595", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Powered Decentralized Generative Agents with Adaptive Hierarchical Knowledge Graph for Cooperative Planning", "abstract": "Developing intelligent agents for long-term cooperation in dynamic open-world scenarios is a major challenge in multi-agent systems. Traditional Multi-agent Reinforcement Learning (MARL) frameworks like centralized training decentralized execution (CTDE) struggle with scalability and flexibility. They require centralized long-term planning, which is difficult without custom reward functions, and face challenges in processing multi-modal data. CTDE approaches also assume fixed cooperation strategies, making them impractical in dynamic environments where agents need to adapt and plan independently. To address decentralized multi-agent cooperation, we propose Decentralized Adaptive Knowledge Graph Memory and Structured Communication System (DAMCS) in a novel Multi-agent Crafter environment. Our generative agents, powered by Large Language Models (LLMs), are more scalable than traditional MARL agents by leveraging external knowledge and language for long-term planning and reasoning. Instead of fully sharing information from all past experiences, DAMCS introduces a multi-modal memory system organized as a hierarchical knowledge graph and a structured communication protocol to optimize agent cooperation. This allows agents to reason from past interactions and share relevant information efficiently. Experiments on novel multi-agent open-world tasks show that DAMCS outperforms both MARL and LLM baselines in task efficiency and collaboration. Compared to single-agent scenarios, the two-agent scenario achieves the same goal with 63% fewer steps, and the six-agent scenario with 74% fewer steps, highlighting the importance of adaptive memory and structured communication in achieving long-term goals. We publicly release our project at: https://happyeureka.github.io/damcs.", "authors": ["Hanqing Yang", "Jingdi Chen", "Marie Siew", "Tania Lorido-Botran", "Carlee Joe-Wong"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-08", "url": "https://arxiv.org/abs/2502.05453", "pdf_url": "https://arxiv.org/pdf/2502.05453v1", "arxiv_id": "2502.05453", "doi": "10.48550/arXiv.2502.05453", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "b151073f3d07d9377452f0dd941a6909928dcb77a6bd69049622069910ae25f7", "sources": ["arxiv", "semantic_scholar"], "title": "Agentic Reasoning: A Streamlined Framework for Enhancing LLM Reasoning with Agentic Tools", "abstract": "We introduce Agentic Reasoning, a framework that enhances large language model (LLM) reasoning by integrating external tool-using agents. Agentic Reasoning dynamically leverages web search, code execution, and structured memory to address complex problems requiring deep research. A key innovation in our framework is the Mind-Map agent, which constructs a structured knowledge graph to store reasoning context and track logical relationships, ensuring coherence in long reasoning chains with extensive tool usage. Additionally, we conduct a comprehensive exploration of the Web-Search agent, leading to a highly effective search mechanism that surpasses all prior approaches. When deployed on DeepSeek-R1, our method achieves a new state-of-the-art (SOTA) among public models and delivers performance comparable to OpenAI Deep Research, the leading proprietary model in this domain. Extensive ablation studies validate the optimal selection of agentic tools and confirm the effectiveness of our Mind-Map and Web-Search agents in enhancing LLM reasoning. The code is at: https://github.com/theworldofagents/Agentic-Reasoning", "authors": ["Junde Wu", "Jiayuan Zhu", "Yuyuan Liu", "Min Xu", "Yueming Jin"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-07", "url": "https://arxiv.org/abs/2502.04644", "pdf_url": "https://arxiv.org/pdf/2502.04644v2", "arxiv_id": "2502.04644", "doi": "10.18653/v1/2025.acl-long.1383", "citation_count": 109, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/theworldofagents/Agentic-Reasoning", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.5103}
{"id": "1cdade4755819a4883b58ccd397db6f83d75ef425460b7e7a61bebd41b1552c4", "sources": ["arxiv", "semantic_scholar"], "title": "SiriuS: Self-improving Multi-agent Systems via Bootstrapped Reasoning", "abstract": "Multi-agent AI systems powered by large language models (LLMs) are increasingly applied to solve complex tasks. However, these systems often rely on fragile, manually designed prompts and heuristics, making optimization difficult. A key challenge in optimizing multi-agent systems is acquiring suitable training data for specialized agents. We introduce SiriuS, a self-improving, reasoning-driven optimization framework for multi-agent systems. Central to our approach is the construction of an experience library: a repository of high-quality reasoning trajectories. The library is built by retaining reasoning steps that lead to successful outcomes, providing a robust training set for optimizing multi-agent system. Additionally, we introduce a library augmentation procedure that refines unsuccessful trajectories, further enriching the library. SiriuS boosts performance by 2.86\\% to 21.88\\% on reasoning and biomedical QA and enhances agent negotiation in competitive settings. Our results show that SiriuS enhances multi-agent performance while generating reusable data for self-correction and self-play enhancement in the future.", "authors": ["Wanjia Zhao", "Mert Yuksekgonul", "Shirley Wu", "James Zou"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-07", "url": "https://arxiv.org/abs/2502.04780", "pdf_url": "https://arxiv.org/pdf/2502.04780v1", "arxiv_id": "2502.04780", "doi": "10.48550/arXiv.2502.04780", "citation_count": 35, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "7276bb5d405384af9a1c7bf69db8c15470ee9ccc5e4aa052a36bb5e4eb395a95", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Architecture Search via Agentic Supernet", "abstract": "Large Language Model (LLM)-empowered multi-agent systems extend the cognitive boundaries of individual agents through disciplined collaboration and interaction, while constructing these systems often requires labor-intensive manual designs. Despite the availability of methods to automate the design of agentic workflows, they typically seek to identify a static, complex, one-size-fits-all system, which, however, fails to dynamically allocate inference resources based on the difficulty and domain of each query. To address this challenge, we shift away from the pursuit of a monolithic agentic system, instead optimizing the \\textbf{agentic supernet}, a probabilistic and continuous distribution of agentic architectures. We introduce MaAS, an automated framework that samples query-dependent agentic systems from the supernet, delivering high-quality solutions and tailored resource allocation (\\textit{e.g.}, LLM calls, tool calls, token cost). Comprehensive evaluation across six benchmarks demonstrates that MaAS \\textbf{(I)} requires only $6\\sim45\\%$ of the inference costs of existing handcrafted or automated multi-agent systems, \\textbf{(II)} surpasses them by $0.54\\%\\sim11.82\\%$, and \\textbf{(III)} enjoys superior cross-dataset and cross-LLM-backbone transferability.", "authors": ["Guibin Zhang", "Luyang Niu", "Junfeng Fang", "Kun Wang", "Lei Bai", "Xiang Wang"], "categories": ["cs.LG", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-06", "url": "https://arxiv.org/abs/2502.04180", "pdf_url": "https://arxiv.org/pdf/2502.04180v2", "arxiv_id": "2502.04180", "doi": "10.48550/arXiv.2502.04180", "citation_count": 119, "influential_citation_count": 24, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.699}
{"id": "da6a39cb8b26b9920b65f416f673246f51fc836818592d83a09ab1bde8c40170", "sources": ["arxiv", "semantic_scholar"], "title": "Speaking the Language of Teamwork: LLM-Guided Credit Assignment in Multi-Agent Reinforcement Learning", "abstract": "Credit assignment, the process of attributing credit or blame to individual agents for their contributions to a team's success or failure, remains a fundamental challenge in multi-agent reinforcement learning (MARL), particularly in environments with sparse rewards. Commonly-used approaches such as value decomposition often lead to suboptimal policies in these settings, and designing dense reward functions that align with human intuition can be complex and labor-intensive. In this work, we propose a novel framework where a large language model (LLM) generates dense, agent-specific rewards based on a natural language description of the task and the overall team goal. By learning a potential-based reward function over multiple queries, our method reduces the impact of ranking errors while allowing the LLM to evaluate each agent's contribution to the overall task. Through extensive experiments, we demonstrate that our approach achieves faster convergence and higher policy returns compared to state-of-the-art MARL baselines.", "authors": ["Muhan Lin", "Shuyang Shi", "Yue Guo", "Vaishnav Tadiparthi", "Behdad Chalaki", "Ehsan Moradi Pari", "Simon Stepputtis", "Woojun Kim", "Joseph Campbell", "Katia Sycara"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-06", "url": "https://arxiv.org/abs/2502.03723", "pdf_url": "https://arxiv.org/pdf/2502.03723v2", "arxiv_id": "2502.03723", "doi": "10.48550/arXiv.2502.03723", "citation_count": 9, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "72ab705945391869aa617d2b935587e53bd3cf8adefcd9071aec3e9f0136435b", "sources": ["arxiv", "semantic_scholar"], "title": "Dynamic Optimizations of LLM Ensembles with Two-Stage Reinforcement Learning Agents", "abstract": "The advancement of LLMs and their accessibility have triggered renewed interest in multi-agent reinforcement learning as robust and adaptive frameworks for dynamically changing environments. This paper introduces RL-Focal, a two-stage RL agent framework that routes and ensembles LLMs. First, we develop the Decider RL-agent, which learns to dynamically select an ensemble of small size ($m_i$) among $N$ LLMs ($m_i \\ll N$) for incoming queries from a user-defined downstream task $i$, by maximizing both error-diversity and reasoning-performance of the selected ensemble through iterative updates of task-adaptive rewards and policy. Second, to enable effective fusion of dynamically selected LLMs, we develop the stage-2 Fusion RL-agent, which learns to resolve reasoning conflicts from different LLMs and dynamically adapts to different ensemble teams composed by the Decider Agent for different downstream tasks. Third, we introduce the focal diversity metric to better model the error correlations among multiple LLMs, further improving the generalization performance of the Decider Agent, which actively prunes the ensemble combinations. By focal diversity, we enhance performance across tasks by effectively promoting reward-aware and policy-adaptive ensemble selection and inference fusion. Extensive evaluations on five benchmarks show that RL-Focal achieves the performance improvement of 8.48\\% with an ensemble of small size compared to the best individual LLM in a pool and offers stronger robustness. Code is available at https://github.com/sftekin/rl-focal", "authors": ["Selim Furkan Tekin", "Fatih Ilhan", "Gaowen Liu", "Ramana Rao Kompella", "Ling Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-06", "url": "https://arxiv.org/abs/2502.04492", "pdf_url": "https://arxiv.org/pdf/2502.04492v2", "arxiv_id": "2502.04492", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/sftekin/rl-focal", "venue": null, "quality_score": 0.1193}
{"id": "66dbd1386e7cb0390453d23e04dc8301c5abe71b858f983b26216ca374ec9777", "sources": ["arxiv", "semantic_scholar"], "title": "Jingfang: An LLM-Based Multi-Agent System for Precise Medical Consultation and Syndrome Differentiation in Traditional Chinese Medicine", "abstract": "The practice of Traditional Chinese Medicine (TCM) requires profound expertise and extensive clinical experience. While Large Language Models (LLMs) offer significant potential in this domain, current TCM-oriented LLMs suffer two critical limitations: (1) a rigid consultation framework that fails to conduct comprehensive and patient-tailored interactions, often resulting in diagnostic inaccuracies; and (2) treatment recommendations generated without rigorous syndrome differentiation, which deviates from the core diagnostic and therapeutic principles of TCM. To address these issues, we develop \\textbf{JingFang (JF)}, an advanced LLM-based multi-agent system for TCM that facilitates the implementation of AI-assisted TCM diagnosis and treatment. JF integrates various TCM Specialist Agents in accordance with authentic diagnostic and therapeutic scenarios of TCM, enabling personalized medical consultations, accurate syndrome differentiation and treatment recommendations. A \\textbf{Multi-Agent Collaborative Consultation Mechanism (MACCM)} for TCM is constructed, where multiple Agents collaborate to emulate real-world TCM diagnostic workflows, enhancing the diagnostic ability of base LLMs to provide accurate and patient-tailored medical consultation. Moreover, we introduce a dedicated \\textbf{Syndrome Differentiation Agent} fine-tuned on a preprocessed dataset, along with a designed \\textbf{Dual-Stage Recovery Scheme (DSRS)} within the Treatment Agent, which together substantially improve the model's accuracy of syndrome differentiation and treatment. Comprehensive evaluations and experiments demonstrate JF's superior performance in medical consultation, and also show improvements of at least 124% and 21.1% in the precision of syndrome differentiation compared to existing TCM models and State of the Art (SOTA) LLMs, respectively.", "authors": ["Yehan Yang", "Tianhao Ma", "Ruotai Li", "Xinhan Zheng", "Guodong Shan"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-04", "url": "https://arxiv.org/abs/2502.04345", "pdf_url": "https://arxiv.org/pdf/2502.04345v3", "arxiv_id": "2502.04345", "doi": null, "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1945}
{"id": "a009be1c8008032cda9ef896c38a42882232a57fa1187680a2942f7cb8ee62e9", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies", "abstract": "Large language models, employed as multiple agents that interact and collaborate with each other, have excelled at solving complex tasks. The agents are programmed with prompts that declare their functionality, along with the topologies that orchestrate interactions across agents. Designing prompts and topologies for multi-agent systems (MAS) is inherently complex. To automate the entire design process, we first conduct an in-depth analysis of the design space aiming to understand the factors behind building effective MAS. We reveal that prompts together with topologies play critical roles in enabling more effective MAS design. Based on the insights, we propose Multi-Agent System Search (MASS), a MAS optimization framework that efficiently exploits the complex MAS design space by interleaving its optimization stages, from local to global, from prompts to topologies, over three stages: 1) block-level (local) prompt optimization; 2) workflow topology optimization; 3) workflow-level (global) prompt optimization, where each stage is conditioned on the iteratively optimized prompts/topologies from former stages. We show that MASS-optimized multi-agent systems outperform a spectrum of existing alternatives by a substantial margin. Based on the MASS-found systems, we finally propose design principles behind building effective multi-agent systems.", "authors": ["Han Zhou", "Xingchen Wan", "Ruoxi Sun", "Hamid Palangi", "Shariq Iqbal", "Ivan Vulić", "Anna Korhonen", "Sercan Ö. Arık"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-04", "url": "https://arxiv.org/abs/2502.02533", "pdf_url": "https://arxiv.org/pdf/2502.02533v2", "arxiv_id": "2502.02533", "doi": "10.48550/arXiv.2502.02533", "citation_count": 89, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4886}
{"id": "9fce22b47777f23c961297a4cd34668b189ac630b0f1812fb2045e70f08a0c13", "sources": ["arxiv", "semantic_scholar"], "title": "Position: Towards a Responsible LLM-empowered Multi-Agent Systems", "abstract": "The rise of Agent AI and Large Language Model-powered Multi-Agent Systems (LLM-MAS) has underscored the need for responsible and dependable system operation. Tools like LangChain and Retrieval-Augmented Generation have expanded LLM capabilities, enabling deeper integration into MAS through enhanced knowledge retrieval and reasoning. However, these advancements introduce critical challenges: LLM agents exhibit inherent unpredictability, and uncertainties in their outputs can compound across interactions, threatening system stability. To address these risks, a human-centered design approach with active dynamic moderation is essential. Such an approach enhances traditional passive oversight by facilitating coherent inter-agent communication and effective system governance, allowing MAS to achieve desired outcomes more efficiently.", "authors": ["Jinwei Hu", "Yi Dong", "Shuang Ao", "Zhuoyun Li", "Boxuan Wang", "Lokesh Singh", "Guangliang Cheng", "Sarvapali D. Ramchurn", "Xiaowei Huang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-03", "url": "https://arxiv.org/abs/2502.01714", "pdf_url": "https://arxiv.org/pdf/2502.01714v1", "arxiv_id": "2502.01714", "doi": "10.48550/arXiv.2502.01714", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3138}
{"id": "e9386c7b20dd7d4b9a9676ea6c03d36b0b30d714f8d4099227a223bf7b950a6b", "sources": ["arxiv", "semantic_scholar"], "title": "PlotGen: Multi-Agent LLM-based Scientific Data Visualization via Multimodal Feedback", "abstract": "Scientific data visualization is pivotal for transforming raw data into comprehensible visual representations, enabling pattern recognition, forecasting, and the presentation of data-driven insights. However, novice users often face difficulties due to the complexity of selecting appropriate tools and mastering visualization techniques. Large Language Models (LLMs) have recently demonstrated potential in assisting code generation, though they struggle with accuracy and require iterative debugging. In this paper, we propose PlotGen, a novel multi-agent framework aimed at automating the creation of precise scientific visualizations. PlotGen orchestrates multiple LLM-based agents, including a Query Planning Agent that breaks down complex user requests into executable steps, a Code Generation Agent that converts pseudocode into executable Python code, and three retrieval feedback agents - a Numeric Feedback Agent, a Lexical Feedback Agent, and a Visual Feedback Agent - that leverage multimodal LLMs to iteratively refine the data accuracy, textual labels, and visual correctness of generated plots via self-reflection. Extensive experiments show that PlotGen outperforms strong baselines, achieving a 4-6 percent improvement on the MatPlotBench dataset, leading to enhanced user trust in LLM-generated visualizations and improved novice productivity due to a reduction in debugging time needed for plot errors.", "authors": ["Kanika Goswami", "Puneet Mathur", "Ryan Rossi", "Franck Dernoncourt"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-03", "url": "https://arxiv.org/abs/2502.00988", "pdf_url": "https://arxiv.org/pdf/2502.00988v1", "arxiv_id": "2502.00988", "doi": "10.48550/arXiv.2502.00988", "citation_count": 24, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3495}
{"id": "0d1d2f989ec2104d592ad36583e4b58e40d1fdfd3db1d9ff18521cff0ae84a04", "sources": ["arxiv", "semantic_scholar"], "title": "Reinforcement Learning for Long-Horizon Interactive LLM Agents", "abstract": "Interactive digital agents (IDAs) leverage APIs of stateful digital environments to perform tasks in response to user requests. While IDAs powered by instruction-tuned large language models (LLMs) can react to feedback from interface invocations in multi-step exchanges, they have not been trained in their respective digital environments. Prior methods accomplish less than half of tasks in sophisticated benchmarks such as AppWorld. We present a reinforcement learning (RL) approach that trains IDAs directly in their target environments. We formalize this training as a partially observable Markov decision process and derive LOOP, a data- and memory-efficient variant of proximal policy optimization. LOOP uses no value network and maintains exactly one copy of the underlying LLM in memory, making its implementation straightforward and as memory-efficient as fine-tuning a single LLM. A 32-billion-parameter agent trained with LOOP in the AppWorld environment outperforms the much larger OpenAI o1 agent by 9 percentage points (15% relative). To our knowledge, this is the first reported application of RL to IDAs that interact with a stateful, multi-domain, multi-app environment via direct API calls. Our analysis sheds light on the effectiveness of RL in this area, showing that the agent learns to consult the API documentation, avoid unwarranted assumptions, minimize confabulation, and recover from setbacks.", "authors": ["Kevin Chen", "Marco Cusumano-Towner", "Brody Huval", "Aleksei Petrenko", "Jackson Hamburger", "Vladlen Koltun", "Philipp Krähenbühl"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-03", "url": "https://arxiv.org/abs/2502.01600", "pdf_url": "https://arxiv.org/pdf/2502.01600v3", "arxiv_id": "2502.01600", "doi": "10.48550/arXiv.2502.01600", "citation_count": 74, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4688}
{"id": "2aeef7af2c332a16b7f0533278289bde8771bd68b90460cf7520f9b2b4c09968", "sources": ["arxiv", "semantic_scholar"], "title": "Agents Are All You Need for LLM Unlearning", "abstract": "Information removal or suppression in large language models (LLMs) is a desired functionality, useful in AI regulation, legal compliance, safety, and privacy. LLM unlearning methods aim to remove information on demand from LLMs. Current LLM unlearning methods struggle to balance the unlearning efficacy and utility due to the competing nature of these objectives. Keeping the unlearning process computationally feasible without assuming access to the model weights is an overlooked area. In this work we show that \\textit{agents might be all we need for effective and practical inference-time LLM unlearning}. We present the first agentic LLM unlearning (\\texttt{ALU}) method, a multi-agent, retrain-free, model-agnostic approach to LLM unlearning that achieves effective unlearning while preserving the utility. Our \\texttt{ALU} framework unlearns by involving multiple LLM agents, each designed for a specific step in the unlearning process, without the need to update model weights for any of the agents in the framework. Users can easily request any set of unlearning instances in any sequence, and \\texttt{ALU} seamlessly adapts in real time. This is facilitated without requiring any changes in the underlying LLM model. Through extensive experiments on established benchmarks (TOFU, WMDP, WPU) and jailbreaking techniques (many shot, target masking, other languages), we demonstrate that \\texttt{ALU} consistently stands out as the most robust inference-time LLM unlearning framework among current state-of-the-art methods while incurring time cost that remains effectively constant regardless of the number of unlearning targets. We further highlight \\texttt{ALU}'s superior performance compared to existing methods when evaluated at scale. Specifically, \\texttt{ALU} is assessed on up to 1000 unlearning targets, exceeding the evaluation scope of all previously proposed LLM unlearning methods.", "authors": ["Debdeep Sanyal", "Murari Mandal"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-02-01", "url": "https://arxiv.org/abs/2502.00406", "pdf_url": "https://arxiv.org/pdf/2502.00406v2", "arxiv_id": "2502.00406", "doi": null, "citation_count": 10, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2603}
{"id": "7cbdf97fda5791e9dc9a62b0d9ee07d63a8874dcbe77af4fc1440614c15b1f53", "sources": ["arxiv", "semantic_scholar"], "title": "Autonomous Legacy Web Application Upgrades Using a Multi-Agent System", "abstract": "The use of Large Language Models (LLMs) for autonomous code generation is gaining attention in emerging technologies. As LLM capabilities expand, they offer new possibilities such as code refactoring, security enhancements, and legacy application upgrades. Many outdated web applications pose security and reliability challenges, yet companies continue using them due to the complexity and cost of upgrades. To address this, we propose an LLM-based multi-agent system that autonomously upgrades legacy web applications to the latest versions. The system distributes tasks across multiple phases, updating all relevant files. To evaluate its effectiveness, we employed Zero-Shot Learning (ZSL) and One-Shot Learning (OSL) prompts, applying identical instructions in both cases. The evaluation involved updating view files and measuring the number and types of errors in the output. For complex tasks, we counted the successfully met requirements. The experiments compared the proposed system with standalone LLM execution, repeated multiple times to account for stochastic behavior. Results indicate that our system maintains context across tasks and agents, improving solution quality over the base model in some cases. This study provides a foundation for future model implementations in legacy code updates. Additionally, findings highlight LLMs' ability to update small outdated files with high precision, even with basic prompts. The source code is publicly available on GitHub: https://github.com/alasalm1/Multi-agent-pipeline.", "authors": ["Valtteri Ala-Salmi", "Zeeshan Rasheed", "Abdul Malik Sami", "Zheying Zhang", "Kai-Kristian Kemell", "Jussi Rasku", "Shahbaz Siddeeq", "Mika Saari", "Pekka Abrahamsson"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-31", "url": "https://arxiv.org/abs/2501.19204", "pdf_url": "https://arxiv.org/pdf/2501.19204v1", "arxiv_id": "2501.19204", "doi": "10.48550/arXiv.2501.19204", "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/alasalm1/Multi-agent-pipeline", "venue": "International Conference on Evaluation of Novel Approaches to Software Engineering", "quality_score": 0.1505}
{"id": "d87dc8f6382ebf4204c72abbed10c97c27598174e22c9dc8c6a240d077ee4b51", "sources": ["arxiv", "semantic_scholar"], "title": "Model Checking for Multi-Agent Systems Modeled By Epistemic Process Calculus", "abstract": "This paper presents a comprehensive framework for modeling and verifying multi-agent systems. The paper introduce an Epistemic Process Calculus for multi-agent systems, which formalizes the syntax and semantics to capture the essential features of agent behavior interactions and epistemic states. Building upon this calculus, we propose ATLE, an extension of Alternating-time Temporal Logic incorporating epistemic operators to express complex properties related to agent epistemic state. To verify ATLE specifications, this paper presents a model checking algorithm that systematically explores the state space of a multi-agent system and evaluates the satisfaction of the specified properties. Finally, a case study is given to demonstrate the method.", "authors": ["Qixian Yu", "Zining Cao", "Zong Hui", "Yuan Zhou"], "categories": ["cs.FL"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-30", "url": "https://arxiv.org/abs/2501.18155", "pdf_url": "https://arxiv.org/pdf/2501.18155v1", "arxiv_id": "2501.18155", "doi": "10.5121/ijsea.2025.16101", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0277}
{"id": "46ac72b67410caec8bcb759c1d7b661016d9b8cf6067d1f871f7ce029cfbe94b", "sources": ["arxiv", "semantic_scholar"], "title": "MASTER: A Multi-Agent System with LLM Specialized MCTS", "abstract": "Large Language Models (LLM) are increasingly being explored for problem-solving tasks. However, their strategic planning capability is often viewed with skepticism. Recent studies have incorporated the Monte Carlo Tree Search (MCTS) algorithm to augment the planning capacity of LLM. Despite its potential, MCTS relies on extensive sampling simulations to approximate the true reward distribution, which leads to two primary issues. Firstly, MCTS is effective for tasks like the Game of Go, where simulation results can yield objective rewards (e.g., 1 for a win and 0 for a loss). However, for tasks such as question answering, the result of a simulation is the answer to the question, which cannot yield an objective reward without the ground truth. Secondly, obtaining statistically significant reward estimations typically requires a sample size exceeding 30 simulations, resulting in excessive token usage and time consumption. To address these challenges, we present the Multi-Agent System with Tactical Execution and Reasoning using LLM Specialized MCTS (MASTER), a novel framework that coordinates agent recruitment and communication through LLM specialized MCTS. This system autonomously adjusts the number of agents based on task complexity and ensures focused communication among them. Comprehensive experiments across various tasks demonstrate the effectiveness of our proposed framework. It achieves 76% accuracy on HotpotQA and 80% on WebShop, setting new state-of-the-art performance on these datasets.", "authors": ["Bingzheng Gan", "Yufan Zhao", "Tianyi Zhang", "Jing Huang", "Yusu Li", "Shu Xian Teo", "Changwang Zhang", "Wei Shi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-24", "url": "https://arxiv.org/abs/2501.14304", "pdf_url": "https://arxiv.org/pdf/2501.14304v2", "arxiv_id": "2501.14304", "doi": "10.48550/arXiv.2501.14304", "citation_count": 21, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.3356}
{"id": "46c238c370c531abf9464c88078ee025145d24c46d4ed7d9491ab878f8376fcc", "sources": ["arxiv", "semantic_scholar"], "title": "Distributed Multi-Agent Coordination Using Multi-Modal Foundation Models", "abstract": "Distributed Constraint Optimization Problems (DCOPs) offer a powerful framework for multi-agent coordination but often rely on labor-intensive, manual problem construction. To address this, we introduce VL-DCOPs, a framework that takes advantage of large multimodal foundation models (LFMs) to automatically generate constraints from both visual and linguistic instructions. We then introduce a spectrum of agent archetypes for solving VL-DCOPs: from a neuro-symbolic agent that delegates some of the algorithmic decisions to an LFM, to a fully neural agent that depends entirely on an LFM for coordination. We evaluate these agent archetypes using state-of-the-art LLMs (large language models) and VLMs (vision language models) on three novel VL-DCOP tasks and compare their respective advantages and drawbacks. Lastly, we discuss how this work extends to broader frontier challenges in the DCOP literature.", "authors": ["Saaduddin Mahmud", "Dorian Benhamou Goldfajn", "Shlomo Zilberstein"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-24", "url": "https://arxiv.org/abs/2501.14189", "pdf_url": "https://arxiv.org/pdf/2501.14189v1", "arxiv_id": "2501.14189", "doi": "10.48550/arXiv.2501.14189", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "811d3cccf73dd077882063a32cdab0123dc1061a0ed0f7543ac8c4571c3895b4", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-agent System for Hybrid Optimization", "abstract": "Optimization problems in process engineering, including design and operation, can often pose challenges to many solvers: multi-modal, non-smooth, and discontinuous models often with large computational requirements. In such cases, the optimization problem is often treated as a black box in which only the value of the objective function is required, sometimes with some indication of the measure of the violation of the constraints. Such problems have traditionally been tackled through the use of direct search and meta-heuristic methods. The challenge, then, is to determine which of these methods or combination of methods should be considered to make most effective use of finite computational resources. This paper presents a multi-agent system for optimization which enables a set of solvers to be applied simultaneously to an optimization problem, including different instantiations of any solver. The evaluation of the optimization problem model is controlled by a scheduler agent which facilitates cooperation and competition between optimization methods. The architecture and implementation of the agent system is described in detail, including the solver, model evaluation, and scheduler agents. A suite of direct search and meta-heuristic methods has been developed for use with this system. Case studies from process systems engineering applications are presented and the results show the potential benefits of automated cooperation between different optimization solvers and motivates the implementation of competition between solvers.", "authors": ["Eric S. Fraga", "Veerawat Udomvorakulchai", "Miguel Pineda", "Lazaros G. Papageorgiou"], "categories": ["math.OC", "cs.MA"], "fields_of_study": ["Mathematics", "Computer Science"], "published_date": "2025-01-16", "url": "https://arxiv.org/abs/2501.09563", "pdf_url": "https://arxiv.org/pdf/2501.09563v1", "arxiv_id": "2501.09563", "doi": "10.48550/arXiv.2501.09563", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Computers and Chemical Engineering", "quality_score": 0.1193}
{"id": "98ebb322b201f176863cbb29fc6e8b130c5c8a35d64e15e8f61802ed84e4261b", "sources": ["arxiv", "semantic_scholar"], "title": "CodeCoR: An LLM-Based Self-Reflective Multi-Agent Framework for Code Generation", "abstract": "Code generation aims to produce code that fulfills requirements written in natural languages automatically. Large language Models (LLMs) like ChatGPT have demonstrated promising effectiveness in this area. Nonetheless, these LLMs often fail to ensure the syntactic and semantic correctness of the generated code. Recently, researchers proposed multi-agent frameworks that guide LLMs with different prompts to analyze programming tasks, generate code, perform testing in a sequential workflow. However, the performance of the workflow is not robust as the code generation depends on the performance of each agent. To address this challenge, we propose CodeCoR, a self-reflective multi-agent framework that evaluates the effectiveness of each agent and their collaborations. Specifically, for a given task description, four agents in CodeCoR generate prompts, code, test cases, and repair advice, respectively. Each agent generates more than one output and prunes away the low-quality ones. The generated code is tested in the local environment: the code that fails to pass the generated test cases is sent to the repair agent and the coding agent re-generates the code based on repair advice. Finally, the code that passes the most number of generated test cases is returned to users. Our experiments on four widely used datasets, HumanEval, HumanEval-ET, MBPP, and MBPP-ET, demonstrate that CodeCoR significantly outperforms existing baselines (e.g., CodeCoT and MapCoder), achieving an average Pass@1 score of 77.8%.", "authors": ["Ruwei Pan", "Hongyu Zhang", "Chao Liu"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-14", "url": "https://arxiv.org/abs/2501.07811", "pdf_url": "https://arxiv.org/pdf/2501.07811v1", "arxiv_id": "2501.07811", "doi": "10.48550/arXiv.2501.07811", "citation_count": 35, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "2d5cf9d03aca53a4501357e9699a7eddfe6d1a853fa6de1d0cb23949b0f49f19", "sources": ["arxiv", "semantic_scholar"], "title": "Talk to Right Specialists: Iterative Routing in Multi-agent Systems for Question Answering", "abstract": "Retrieval-augmented generation (RAG) agents are increasingly deployed to answer questions over local knowledge bases that cannot be centralized due to knowledge-sovereignty constraints. This results in two recurring failures in production: users do not know which agent to consult, and complex questions require evidence distributed across multiple agents. To overcome these challenges, we propose RIRS, a training-free orchestration framework to enable a multi-agent system for question answering. In detail, RIRS summarizes each agent's local corpus in an embedding space, enabling a user-facing server to route queries only to the most relevant agents, reducing latency and avoiding noisy \"broadcast-to-all\" contexts. For complicated questions, the server can iteratively aggregate responses to derive intermediate results and refine the question to bridge the gap toward a comprehensive answer. Extensive experiments demonstrate the effectiveness of RIRS, including its ability to precisely select agents and provide accurate responses to single-hop queries, and its use of an iterative strategy to achieve accurate, multi-step resolutions for complex queries.", "authors": ["Feijie Wu", "Zitao Li", "Fei Wei", "Yaliang Li", "Bolin Ding", "Jing Gao"], "categories": ["cs.MA", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-14", "url": "https://arxiv.org/abs/2501.07813", "pdf_url": "https://arxiv.org/pdf/2501.07813v2", "arxiv_id": "2501.07813", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.016}
{"id": "140e429ba999cde45646785ab6888dd63a2263de0d95cb4a16631e9b7faabe1d", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaboration Mechanisms: A Survey of LLMs", "abstract": "With recent advances in Large Language Models (LLMs), Agentic AI has become phenomenal in real-world applications, moving toward multiple LLM-based agents to perceive, learn, reason, and act collaboratively. These LLM-based Multi-Agent Systems (MASs) enable groups of intelligent agents to coordinate and solve complex tasks collectively at scale, transitioning from isolated models to collaboration-centric approaches. This work provides an extensive survey of the collaborative aspect of MASs and introduces an extensible framework to guide future research. Our framework characterizes collaboration mechanisms based on key dimensions: actors (agents involved), types (e.g., cooperation, competition, or coopetition), structures (e.g., peer-to-peer, centralized, or distributed), strategies (e.g., role-based or model-based), and coordination protocols. Through a review of existing methodologies, our findings serve as a foundation for demystifying and advancing LLM-based MASs toward more intelligent and collaborative solutions for complex, real-world use cases. In addition, various applications of MASs across diverse domains, including 5G/6G networks, Industry 5.0, question answering, and social and cultural settings, are also investigated, demonstrating their wider adoption and broader impacts. Finally, we identify key lessons learned, open challenges, and potential research directions of MASs towards artificial collective intelligence.", "authors": ["Khanh-Tung Tran", "Dung Dao", "Minh-Duong Nguyen", "Quoc-Viet Pham", "Barry O'Sullivan", "Hoang D. Nguyen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-10", "url": "https://arxiv.org/abs/2501.06322", "pdf_url": "https://arxiv.org/pdf/2501.06322v1", "arxiv_id": "2501.06322", "doi": "10.48550/arXiv.2501.06322", "citation_count": 466, "influential_citation_count": 21, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6712}
{"id": "1998b14ee7125f7ba71dbf3a800bf9e833117e856bbbd6027e8c02c794b0f234", "sources": ["arxiv", "semantic_scholar"], "title": "On Corrigibility and Alignment in Multi Agent Games", "abstract": "Corrigibility of autonomous agents is an under explored part of system design, with previous work focusing on single agent systems. It has been suggested that uncertainty over the human preferences acts to keep the agents corrigible, even in the face of human irrationality. We present a general framework for modelling corrigibility in a multi-agent setting as a 2 player game in which the agents always have a move in which they can ask the human for supervision. This is formulated as a Bayesian game for the purpose of introducing uncertainty over the human beliefs. We further analyse two specific cases. First, a two player corrigibility game, in which we want corrigibility displayed in both agents for both common payoff (monotone) games and harmonic games. Then we investigate an adversary setting, in which one agent is considered to be a `defending' agent and the other an `adversary'. A general result is provided for what belief over the games and human rationality the defending agent is required to have to induce corrigibility.", "authors": ["Edmund Dable-Heath", "Boyko Vodenicharski", "James Bishop"], "categories": ["cs.GT", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-09", "url": "https://arxiv.org/abs/2501.05360", "pdf_url": "https://arxiv.org/pdf/2501.05360v1", "arxiv_id": "2501.05360", "doi": "10.48550/arXiv.2501.05360", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0195}
{"id": "f77a1bb8bdc1bc747767e2517140f0debb95d3169ea16b40cabb47b07915dd1c", "sources": ["arxiv", "semantic_scholar"], "title": "Sustainable Digitalization of Business with Multi-Agent RAG and LLM", "abstract": "Businesses heavily rely on data sourced from various channels like news articles, financial reports, and consumer reviews to drive their operations, enabling informed decision-making and identifying opportunities. However, traditional manual methods for data extraction are often time-consuming and resource-intensive, prompting the adoption of digital transformation initiatives to enhance efficiency. Yet, concerns persist regarding the sustainability of such initiatives and their alignment with the United Nations (UN)'s Sustainable Development Goals (SDGs). This research aims to explore the integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) as a sustainable solution for Information Extraction (IE) and processing. The research methodology involves reviewing existing solutions for business decision-making, noting that many systems require training new machine learning models, which are resource-intensive and have significant environmental impacts. Instead, we propose a sustainable business solution using pre-existing LLMs that can work with diverse datasets. We link domain-specific datasets to tailor LLMs to company needs and employ a Multi-Agent architecture to divide tasks such as information retrieval, enrichment, and classification among specialized agents. This approach optimizes the extraction process and improves overall efficiency. Through the utilization of these technologies, businesses can optimize resource utilization, improve decision-making processes, and contribute to sustainable development goals, thereby fostering environmental responsibility within the corporate sector.", "authors": ["Muhammad Arslan", "Saba Munawar", "Christophe Cruz"], "categories": ["cs.IR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-06", "url": "https://arxiv.org/abs/2502.15700", "pdf_url": "https://arxiv.org/pdf/2502.15700v1", "arxiv_id": "2502.15700", "doi": "10.1016/j.procs.2024.09.337", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Knowledge-Based Intelligent Information & Engineering Systems", "quality_score": 0.2258}
{"id": "79b788f9de0273b094dfe0d14cc8d798a45476c9b991111ab9a818b35123527d", "sources": ["arxiv", "semantic_scholar"], "title": "Communicating Unexpectedness for Out-of-Distribution Multi-Agent Reinforcement Learning", "abstract": "Applying multi-agent reinforcement learning methods to realistic settings is challenging as it may require the agents to quickly adapt to unexpected situations that are rarely or never encountered in training. Recent methods for generalization to such out-of-distribution settings are limited to more specific, restricted instances of distribution shifts. To tackle adaptation to distribution shifts, we propose Unexpected Encoding Scheme, a novel decentralized multi-agent reinforcement learning algorithm where agents communicate \"unexpectedness,\" the aspects of the environment that are surprising. In addition to a message yielded by the original reward-driven communication, each agent predicts the next observation based on previous experience, measures the discrepancy between the prediction and the actually encountered observation, and encodes this discrepancy as a message. Experiments on multi-robot warehouse environment support that our proposed method adapts robustly to dynamically changing training environments as well as out-of-distribution environment.", "authors": ["Min Whoo Lee", "Kibeom Kim", "Soo Wung Shin", "Minsu Lee", "Byoung-Tak Zhang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-02", "url": "https://arxiv.org/abs/2501.01140", "pdf_url": "https://arxiv.org/pdf/2501.01140v1", "arxiv_id": "2501.01140", "doi": "10.48550/arXiv.2501.01140", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0115}
{"id": "e1babd930df8ec548d86713a29b78f78c9056e9927cf55d0d89f6eb06e28da43", "sources": ["arxiv", "semantic_scholar"], "title": "Harnessing Multi-Agent LLMs for Complex Engineering Problem-Solving: A Framework for Senior Design Projects", "abstract": "Multi-Agent Large Language Models (LLMs) are gaining significant attention for their ability to harness collective intelligence in complex problem-solving, decision-making, and planning tasks. This aligns with the concept of the wisdom of crowds, where diverse agents contribute collectively to generating effective solutions, making it particularly suitable for educational settings. Senior design projects, also known as capstone or final year projects, are pivotal in engineering education as they integrate theoretical knowledge with practical application, fostering critical thinking, teamwork, and real-world problem-solving skills. In this paper, we explore the use of Multi-Agent LLMs in supporting these senior design projects undertaken by engineering students, which often involve multidisciplinary considerations and conflicting objectives, such as optimizing technical performance while addressing ethical, social, and environmental concerns. We propose a framework where distinct LLM agents represent different expert perspectives, such as problem formulation agents, system complexity agents, societal and ethical agents, or project managers, thus facilitating a holistic problem-solving approach. This implementation leverages standard multi-agent system (MAS) concepts such as coordination, cooperation, and negotiation, incorporating prompt engineering to develop diverse personas for each agent. These agents engage in rich, collaborative dialogues to simulate human engineering teams, guided by principles from swarm AI to efficiently balance individual contributions towards a unified solution. We adapt these techniques to create a collaboration structure for LLM agents, encouraging interdisciplinary reasoning and negotiation similar to real-world senior design projects. To assess the efficacy of this framework, we collected six proposals of engineering and computer science of...", "authors": ["Abdullah Mushtaq", "Muhammad Rafay Naeem", "Ibrahim Ghaznavi", "Muhammad Imran Taj", "Imran Hashmi", "Junaid Qadir"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2025-01-02", "url": "https://arxiv.org/abs/2501.01205", "pdf_url": "https://arxiv.org/pdf/2501.01205v1", "arxiv_id": "2501.01205", "doi": "10.1109/EDUCON62633.2025.11016653", "citation_count": 14, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "IEEE Global Engineering Education Conference", "quality_score": 0.294}
{"id": "eae2fb27dd3102a04ff540ddc45655a436e96bfe0398f9b9f114e181192e5ddf", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Powered Multi-Agent System for Automated Crypto Portfolio Management", "abstract": "Cryptocurrency investment is inherently difficult due to its shorter history compared to traditional assets, the need to integrate vast amounts of data from various modalities, and the requirement for complex reasoning. While deep learning approaches have been applied to address these challenges, their black-box nature raises concerns about trust and explainability. Recently, large language models (LLMs) have shown promise in financial applications due to their ability to understand multi-modal data and generate explainable decisions. However, single LLM faces limitations in complex, comprehensive tasks such as asset investment. These limitations are even more pronounced in cryptocurrency investment, where LLMs have less domain-specific knowledge in their training corpora. To overcome these challenges, we propose an explainable, multi-modal, multi-agent framework for cryptocurrency investment. Our framework uses specialized agents that collaborate within and across teams to handle subtasks such as data analysis, literature integration, and investment decision-making for the top 30 cryptocurrencies by market capitalization. The expert training module fine-tunes agents using multi-modal historical data and professional investment literature, while the multi-agent investment module employs real-time data to make informed cryptocurrency investment decisions. Unique intrateam and interteam collaboration mechanisms enhance prediction accuracy by adjusting final predictions based on confidence levels within agent teams and facilitating information sharing between teams. Empirical evaluation using data from November 2023 to September 2024 demonstrates that our framework outperforms single-agent models and market benchmarks in classification, asset pricing, portfolio, and explainability performance.", "authors": ["Yichen Luo", "Yebo Feng", "Jiahua Xu", "Paolo Tasca", "Yang Liu"], "categories": ["q-fin.TR", "cs.AI"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2025-01-01", "url": "https://arxiv.org/abs/2501.00826", "pdf_url": "https://arxiv.org/pdf/2501.00826v2", "arxiv_id": "2501.00826", "doi": "10.48550/arXiv.2501.00826", "citation_count": 30, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3728}
{"id": "ee4da7b93f8bd8cc0b58263fe0cd97749b872644cac4596b0215ec530ade4044", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing LLM Reasoning with Multi-Path Collaborative Reactive and Reflection agents", "abstract": "Agents have demonstrated their potential in scientific reasoning tasks through large language models. However, they often face challenges such as insufficient accuracy and degeneration of thought when handling complex reasoning tasks, which impede their performance. To overcome these issues, we propose the Reactive and Reflection agents with Multi-Path Reasoning (RR-MP) Framework, aimed at enhancing the reasoning capabilities of LLMs. Our approach improves scientific reasoning accuracy by employing a multi-path reasoning mechanism where each path consists of a reactive agent and a reflection agent that collaborate to prevent degeneration of thought inherent in single-agent reliance. Additionally, the RR-MP framework does not require additional training; it utilizes multiple dialogue instances for each reasoning path and a separate summarizer to consolidate insights from all paths. This design integrates diverse perspectives and strengthens reasoning across each path. We conducted zero-shot and few-shot evaluations on tasks involving moral scenarios, college-level physics, and mathematics. Experimental results demonstrate that our method outperforms baseline approaches, highlighting the effectiveness and advantages of the RR-MP framework in managing complex scientific reasoning tasks.", "authors": ["Chengbo He", "Bochao Zou", "Xin Li", "Jiansheng Chen", "Junliang Xing", "Huimin Ma"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-31", "url": "https://arxiv.org/abs/2501.00430", "pdf_url": "https://arxiv.org/pdf/2501.00430v2", "arxiv_id": "2501.00430", "doi": "10.48550/arXiv.2501.00430", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "c3d7dbcddb40f438db81229973b98cfc589306867e997f5ae815cfe9a1604de3", "sources": ["arxiv", "semantic_scholar"], "title": "AI Agent for Education: von Neumann Multi-Agent System Framework", "abstract": "The development of large language models has ushered in new paradigms for education. This paper centers on the multi-Agent system in education and proposes the von Neumann multi-Agent system framework. It breaks down each AI Agent into four modules: control unit, logic unit, storage unit, and input-output devices, defining four types of operations: task deconstruction, self-reflection, memory processing, and tool invocation. Furthermore, it introduces related technologies such as Chain-of-Thought, Reson+Act, and Multi-Agent Debate associated with these four types of operations. The paper also discusses the ability enhancement cycle of a multi-Agent system for education, including the outer circulation for human learners to promote knowledge construction and the inner circulation for LLM-based-Agents to enhance swarm intelligence. Through collaboration and reflection, the multi-Agent system can better facilitate human learners' learning and enhance their teaching abilities in this process.", "authors": ["Yuan-Hao Jiang", "Ruijia Li", "Yizhou Zhou", "Changyong Qi", "Hanglei Hu", "Yuang Wei", "Bo Jiang", "Yonghe Wu"], "categories": ["cs.MA", "cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-30", "url": "https://arxiv.org/abs/2501.00083", "pdf_url": "https://arxiv.org/pdf/2501.00083v1", "arxiv_id": "2501.00083", "doi": "10.48550/arXiv.2501.00083", "citation_count": 19, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "830623f212b9deeb7713a84309b91ad886d4a8a1886a3b9148c27d15062989e3", "sources": ["arxiv", "semantic_scholar"], "title": "Plancraft: an evaluation dataset for planning with LLM agents", "abstract": "We present Plancraft, a multi-modal evaluation dataset for LLM agents. Plancraft has both a text-only and multi-modal interface, based on the Minecraft crafting GUI. We include the Minecraft Wiki to evaluate tool use and Retrieval Augmented Generation (RAG), as well as a handcrafted planner and Oracle Retriever, to ablate the different components of a modern agent architecture. To evaluate decision-making, Plancraft also includes a subset of examples that are intentionally unsolvable, providing a realistic challenge that requires the agent not only to complete tasks but also to decide whether they are solvable at all. We benchmark both open-source and closed-source LLMs and compare their performance and efficiency to a handcrafted planner. Overall, we find that LLMs and VLMs struggle with the planning problems that Plancraft introduces, and offer suggestions on how to improve their capabilities.", "authors": ["Gautier Dagan", "Frank Keller", "Alex Lascarides"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-30", "url": "https://arxiv.org/abs/2412.21033", "pdf_url": "https://arxiv.org/pdf/2412.21033v2", "arxiv_id": "2412.21033", "doi": "10.48550/arXiv.2412.21033", "citation_count": 15, "influential_citation_count": 3, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "be450f46a6e2ef240ba24ed3b1637ad4e9caf4da865dae783cde562e2f05d168", "sources": ["arxiv", "semantic_scholar"], "title": "TradingAgents: Multi-Agents LLM Financial Trading Framework", "abstract": "Significant progress has been made in automated problem-solving using societies of agents powered by large language models (LLMs). In finance, efforts have largely focused on single-agent systems handling specific tasks or multi-agent frameworks independently gathering data. However, the multi-agent systems' potential to replicate real-world trading firms' collaborative dynamics remains underexplored. TradingAgents proposes a novel stock trading framework inspired by trading firms, featuring LLM-powered agents in specialized roles such as fundamental analysts, sentiment analysts, technical analysts, and traders with varied risk profiles. The framework includes Bull and Bear researcher agents assessing market conditions, a risk management team monitoring exposure, and traders synthesizing insights from debates and historical data to make informed decisions. By simulating a dynamic, collaborative trading environment, this framework aims to improve trading performance. Detailed architecture and extensive experiments reveal its superiority over baseline models, with notable improvements in cumulative returns, Sharpe ratio, and maximum drawdown, highlighting the potential of multi-agent LLM frameworks in financial trading. TradingAgents is available at https://github.com/TauricResearch/TradingAgents.", "authors": ["Yijia Xiao", "Edward Sun", "Di Luo", "Wei Wang"], "categories": ["q-fin.TR", "cs.AI", "cs.CE", "cs.LG"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2024-12-28", "url": "https://arxiv.org/abs/2412.20138", "pdf_url": "https://arxiv.org/pdf/2412.20138v7", "arxiv_id": "2412.20138", "doi": "10.48550/arXiv.2412.20138", "citation_count": 156, "influential_citation_count": 15, "has_code": true, "code_url": "https://github.com/TauricResearch;", "venue": "arXiv.org", "quality_score": 0.6021}
{"id": "5dc15541851dd0369c917459823b0e9bc7a758aa77f63173fdcea9be00313e7a", "sources": ["arxiv", "semantic_scholar"], "title": "Efficient Multi-Agent Collaboration with Tool Use for Online Planning in Complex Table Question Answering", "abstract": "Complex table question answering (TQA) aims to answer questions that require complex reasoning, such as multi-step or multi-category reasoning, over data represented in tabular form. Previous approaches demonstrated notable performance by leveraging either closed-source large language models (LLMs) or fine-tuned open-weight LLMs. However, fine-tuning LLMs requires high-quality training data, which is costly to obtain, and utilizing closed-source LLMs poses accessibility challenges and leads to reproducibility issues. In this paper, we propose Multi-Agent Collaboration with Tool use (MACT), a framework that requires neither closed-source models nor fine-tuning. In MACT, a planning agent and a coding agent that also make use of tools collaborate to answer questions. Our experiments on four TQA benchmarks show that MACT outperforms previous SoTA systems on three out of four benchmarks and that it performs comparably to the larger and more expensive closed-source model GPT-4 on two benchmarks, even when using only open-weight models without any fine-tuning. We conduct extensive analyses to prove the effectiveness of MACT's multi-agent collaboration in TQA.", "authors": ["Wei Zhou", "Mohsen Mesgar", "Annemarie Friedrich", "Heike Adel"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-28", "url": "https://arxiv.org/abs/2412.20145", "pdf_url": "https://arxiv.org/pdf/2412.20145v2", "arxiv_id": "2412.20145", "doi": "10.48550/arXiv.2412.20145", "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.3076}
{"id": "903a25b7279ad797c2e2974ec763cb8580149326bd9d7f6e2b5708f721c935bf", "sources": ["arxiv", "semantic_scholar"], "title": "INVESTORBENCH: A Benchmark for Financial Decision-Making Tasks with LLM-based Agent", "abstract": "Recent advancements have underscored the potential of large language model (LLM)-based agents in financial decision-making. Despite this progress, the field currently encounters two main challenges: (1) the lack of a comprehensive LLM agent framework adaptable to a variety of financial tasks, and (2) the absence of standardized benchmarks and consistent datasets for assessing agent performance. To tackle these issues, we introduce \\textsc{InvestorBench}, the first benchmark specifically designed for evaluating LLM-based agents in diverse financial decision-making contexts. InvestorBench enhances the versatility of LLM-enabled agents by providing a comprehensive suite of tasks applicable to different financial products, including single equities like stocks, cryptocurrencies and exchange-traded funds (ETFs). Additionally, we assess the reasoning and decision-making capabilities of our agent framework using thirteen different LLMs as backbone models, across various market environments and tasks. Furthermore, we have curated a diverse collection of open-source, multi-modal datasets and developed a comprehensive suite of environments for financial decision-making. This establishes a highly accessible platform for evaluating financial agents' performance across various scenarios.", "authors": ["Haohang Li", "Yupeng Cao", "Yangyang Yu", "Shashidhar Reddy Javaji", "Zhiyang Deng", "Yueru He", "Yuechen Jiang", "Zining Zhu", "Koduvayur Subbalakshmi", "Guojun Xiong", "Jimin Huang", "Lingfei Qian", "Xueqing Peng", "Qianqian Xie", "Jordan W. Suchow"], "categories": ["cs.CE", "cs.AI", "q-fin.CP"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2024-12-24", "url": "https://arxiv.org/abs/2412.18174", "pdf_url": "https://arxiv.org/pdf/2412.18174v1", "arxiv_id": "2412.18174", "doi": "10.48550/arXiv.2412.18174", "citation_count": 46, "influential_citation_count": 4, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.418}
{"id": "44de5fa8d44f2bcacd95df6b98969a0c22156f7f7cc025f23b12b9de0cdc5195", "sources": ["arxiv", "semantic_scholar"], "title": "A Survey on LLM-based Multi-Agent System: Recent Advances and New Frontiers in Application", "abstract": "LLM-based Multi-Agent Systems ( LLM-MAS ) have become a research hotspot since the rise of large language models (LLMs). However, with the continuous influx of new related works, the existing reviews struggle to capture them comprehensively. This paper presents a comprehensive survey of these studies. We first discuss the definition of LLM-MAS, a framework encompassing much of previous work. We provide an overview of the various applications of LLM-MAS in (i) solving complex tasks, (ii) simulating specific scenarios, and (iii) evaluating generative agents. Building on previous studies, we also highlight several challenges and propose future directions for research in this field.", "authors": ["Shuaihang Chen", "Yuanxing Liu", "Wei Han", "Weinan Zhang", "Ting Liu"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-23", "url": "https://arxiv.org/abs/2412.17481", "pdf_url": "https://arxiv.org/pdf/2412.17481v2", "arxiv_id": "2412.17481", "doi": null, "citation_count": 62, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.4498}
{"id": "06055dabdf60ce427eb7be702fde9b61f5dc67ac7842386281f8558b11c18f2e", "sources": ["arxiv", "semantic_scholar"], "title": "KG4Diagnosis: A Hierarchical Multi-Agent LLM Framework with Knowledge Graph Enhancement for Medical Diagnosis", "abstract": "Integrating Large Language Models (LLMs) in healthcare diagnosis demands systematic frameworks that can handle complex medical scenarios while maintaining specialized expertise. We present KG4Diagnosis, a novel hierarchical multi-agent framework that combines LLMs with automated knowledge graph construction, encompassing 362 common diseases across medical specialties. Our framework mirrors real-world medical systems through a two-tier architecture: a general practitioner (GP) agent for initial assessment and triage, coordinating with specialized agents for in-depth diagnosis in specific domains. The core innovation lies in our end-to-end knowledge graph generation methodology, incorporating: (1) semantic-driven entity and relation extraction optimized for medical terminology, (2) multi-dimensional decision relationship reconstruction from unstructured medical texts, and (3) human-guided reasoning for knowledge expansion. KG4Diagnosis serves as an extensible foundation for specialized medical diagnosis systems, with capabilities to incorporate new diseases and medical knowledge. The framework's modular design enables seamless integration of domain-specific enhancements, making it valuable for developing targeted medical diagnosis systems. We provide architectural guidelines and protocols to facilitate adoption across medical contexts.", "authors": ["Kaiwen Zuo", "Yirui Jiang", "Fan Mo", "Pietro Lio"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-22", "url": "https://arxiv.org/abs/2412.16833", "pdf_url": "https://arxiv.org/pdf/2412.16833v4", "arxiv_id": "2412.16833", "doi": "10.48550/arXiv.2412.16833", "citation_count": 46, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.418}
{"id": "342c8dc506ae30a87aee963e6a345801cfde010998973c2cf35a6739b682db6e", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool Usage", "abstract": "The advancement of large language models (LLMs) prompts the development of multi-modal agents, which are used as a controller to call external tools, providing a feasible way to solve practical tasks. In this paper, we propose a multi-modal agent tuning method that automatically generates multi-modal tool-usage data and tunes a vision-language model (VLM) as the controller for powerful tool-usage reasoning. To preserve the data quality, we prompt the GPT-4o mini model to generate queries, files, and trajectories, followed by query-file and trajectory verifiers. Based on the data synthesis pipeline, we collect the MM-Traj dataset that contains 20K tasks with trajectories of tool usage. Then, we develop the T3-Agent via \\underline{T}rajectory \\underline{T}uning on VLMs for \\underline{T}ool usage using MM-Traj. Evaluations on the GTA and GAIA benchmarks show that the T3-Agent consistently achieves improvements on two popular VLMs: MiniCPM-V-8.5B and {Qwen2-VL-7B}, which outperforms untrained VLMs by $20\\%$, showing the effectiveness of the proposed data synthesis pipeline, leading to high-quality data for tool-usage capabilities.", "authors": ["Zhi Gao", "Bofei Zhang", "Pengxiang Li", "Xiaojian Ma", "Tao Yuan", "Yue Fan", "Yuwei Wu", "Yunde Jia", "Song-Chun Zhu", "Qing Li"], "categories": ["cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-20", "url": "https://arxiv.org/abs/2412.15606", "pdf_url": "https://arxiv.org/pdf/2412.15606v2", "arxiv_id": "2412.15606", "doi": "10.48550/arXiv.2412.15606", "citation_count": 54, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4351}
{"id": "20bc307ef7162369fcfa2e1321c2b96e6291dc687e884716c362c2b571b7dfce", "sources": ["arxiv", "semantic_scholar"], "title": "Understanding Individual Agent Importance in Multi-Agent System via Counterfactual Reasoning", "abstract": "Explaining multi-agent systems (MAS) is urgent as these systems become increasingly prevalent in various applications. Previous work has proveided explanations for the actions or states of agents, yet falls short in understanding the black-boxed agent's importance within a MAS and the overall team strategy. To bridge this gap, we propose EMAI, a novel agent-level explanation approach that evaluates the individual agent's importance. Inspired by counterfactual reasoning, a larger change in reward caused by the randomized action of agent indicates its higher importance. We model it as a MARL problem to capture interactions across agents. Utilizing counterfactual reasoning, EMAI learns the masking agents to identify important agents. Specifically, we define the optimization function to minimize the reward difference before and after action randomization and introduce sparsity constraints to encourage the exploration of more action randomization of agents during training. The experimental results in seven multi-agent tasks demonstratee that EMAI achieves higher fidelity in explanations than baselines and provides more effective guidance in practical applications concerning understanding policies, launching attacks, and patching policies.", "authors": ["Jianming Chen", "Yawen Wang", "Junjie Wang", "Xiaofei Xie", "jun Hu", "Qing Wang", "Fanjiang Xu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-20", "url": "https://arxiv.org/abs/2412.15619", "pdf_url": "https://arxiv.org/pdf/2412.15619v2", "arxiv_id": "2412.15619", "doi": "10.48550/arXiv.2412.15619", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.2785}
{"id": "b10de1c23ec1947c2c06a0d30659d434989702aba2c0c451e1d4a0e7d6b0750e", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Multi-Modal Data with Tool-Augmented LLM Agents for Precise Causal Discovery", "abstract": "Causal discovery is an imperative foundation for decision-making across domains, such as smart health, AI for drug discovery and AIOps. Traditional statistical causal discovery methods, while well-established, predominantly rely on observational data and often overlook the semantic cues inherent in cause-and-effect relationships. The advent of Large Language Models (LLMs) has ushered in an affordable way of leveraging the semantic cues for knowledge-driven causal discovery, but the development of LLMs for causal discovery lags behind other areas, particularly in the exploration of multi-modal data. To bridge the gap, we introduce MATMCD, a multi-agent system powered by tool-augmented LLMs. MATMCD has two key agents: a Data Augmentation agent that retrieves and processes modality-augmented data, and a Causal Constraint agent that integrates multi-modal data for knowledge-driven reasoning. The proposed design of the inner-workings ensures successful cooperation of the agents. Our empirical study across seven datasets suggests the significant potential of multi-modality enhanced causal discovery.", "authors": ["ChengAo Shen", "Zhengzhang Chen", "Dongsheng Luo", "Dongkuan Xu", "Haifeng Chen", "Jingchao Ni"], "categories": ["cs.LG", "cs.AI", "stat.ME"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2024-12-18", "url": "https://arxiv.org/abs/2412.13667", "pdf_url": "https://arxiv.org/pdf/2412.13667v2", "arxiv_id": "2412.13667", "doi": "10.18653/v1/2025.findings-acl.36", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.2785}
{"id": "5a983a39863a4667c21486d9aea64a1c91727c8e4455621fc097c172684ba34b", "sources": ["arxiv", "semantic_scholar"], "title": "SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents", "abstract": "With the integration of large language models (LLMs), embodied agents have strong capabilities to understand and plan complicated natural language instructions. However, a foreseeable issue is that those embodied agents can also flawlessly execute some hazardous tasks, potentially causing damages in the real world. Existing benchmarks predominantly overlook critical safety risks, focusing solely on planning performance, while a few evaluate LLMs' safety awareness only on non-interactive image-text data. To address this gap, we present SafeAgentBench -- the first comprehensive benchmark for safety-aware task planning of embodied LLM agents in interactive simulation environments, covering both explicit and implicit hazards. SafeAgentBench includes: (1) an executable, diverse, and high-quality dataset of 750 tasks, rigorously curated to cover 10 potential hazards and 3 task types; (2) SafeAgentEnv, a universal embodied environment with a low-level controller, supporting multi-agent execution with 17 high-level actions for 9 state-of-the-art baselines; and (3) reliable evaluation methods from both execution and semantic perspectives. Experimental results show that, although agents based on different design frameworks exhibit substantial differences in task success rates, their overall safety awareness remains weak. The most safety-conscious baseline achieves only a 10% rejection rate for detailed hazardous tasks. Moreover, simply replacing the LLM driving the agent does not lead to notable improvements in safety awareness. Dataset and codes are available in https://github.com/shengyin1224/SafeAgentBench and https://huggingface.co/datasets/safeagentbench/SafeAgentBench.", "authors": ["Sheng Yin", "Xianghe Pang", "Yuanzhuo Ding", "Menglan Chen", "Yutong Bi", "Yichen Xiong", "Wenhao Huang", "Zhen Xiang", "Jing Shao", "Siheng Chen"], "categories": ["cs.CR", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-17", "url": "https://arxiv.org/abs/2412.13178", "pdf_url": "https://arxiv.org/pdf/2412.13178v5", "arxiv_id": "2412.13178", "doi": "10.48550/arXiv.2412.13178", "citation_count": 89, "influential_citation_count": 15, "has_code": true, "code_url": "https://github.com/shengyin1224/SafeAgentBench", "venue": "arXiv.org", "quality_score": 0.6021}
{"id": "8fd188453f4c44eeff284168a08539c61bd8bd6fb73bfc062aa6a4767f9bb09c", "sources": ["arxiv", "semantic_scholar"], "title": "Harnessing Language for Coordination: A Framework and Benchmark for LLM-Driven Multi-Agent Control", "abstract": "Large Language Models (LLMs) have demonstrated remarkable performance across various tasks. Their potential to facilitate human coordination with many agents is a promising but largely under-explored area. Such capabilities would be helpful in disaster response, urban planning, and real-time strategy scenarios. In this work, we introduce (1) a real-time strategy game benchmark designed to evaluate these abilities and (2) a novel framework we term HIVE. HIVE empowers a single human to coordinate swarms of up to 2,000 agents through a natural language dialog with an LLM. We present promising results on this multi-agent benchmark, with our hybrid approach solving tasks such as coordinating agent movements, exploiting unit weaknesses, leveraging human annotations, and understanding terrain and strategic points. Our findings also highlight critical limitations of current models, including difficulties in processing spatial visual information and challenges in formulating long-term strategic plans. This work sheds light on the potential and limitations of LLMs in human-swarm coordination, paving the way for future research in this area. The HIVE project page, hive.syrkis.com, includes videos of the system in action.", "authors": ["Timothée Anne", "Noah Syrkis", "Meriem Elhosni", "Florian Turati", "Franck Legendre", "Alain Jaquier", "Sebastian Risi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-16", "url": "https://arxiv.org/abs/2412.11761", "pdf_url": "https://arxiv.org/pdf/2412.11761v2", "arxiv_id": "2412.11761", "doi": "10.1109/TG.2025.3564042", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Games", "quality_score": 0.2258}
{"id": "7ee56c0b0b30c3c250c85699c232b1d5e2f07ca0d637f7e6140c841d529f7fab", "sources": ["arxiv", "semantic_scholar"], "title": "CoopetitiveV: Leveraging LLM-powered Coopetitive Multi-Agent Prompting for High-quality Verilog Generation", "abstract": "Recent advances in agentic LLMs have demonstrated great capabilities in Verilog code generation. However, existing approaches either use LLM-assisted single-agent prompting or cooperation-only multi-agent learning, which will lead to: (i) Degeneration issue for single-agent learning: characterized by diminished error detection and correction capabilities; (ii) Error propagation in cooperation-only multi-agent learning: erroneous information from the former agent will be propagated to the latter through prompts, which can make the latter agents generate buggy code. In this paper, we propose an LLM-based coopetitive multi-agent prompting framework, in which the agents cannot collaborate with each other to form the generation pipeline, but also create a healthy competitive mechanism to improve the generating quality. Our experimental results show that the coopetitive multi-agent framework can effectively mitigate the degeneration risk and reduce the error propagation while improving code error correction capabilities, resulting in higher quality Verilog code generation. The effectiveness of our approach is validated through extensive experiments. On VerilogEval Machine and Human dataset, CoopetitiveV+GPT-4 achieves 99.2% and 99.1% pass@10 scores, respectively. While on RTLLM, CoopetitiveV+GPT-4 obtains 100% syntax and 99.9% functionality pass@5 scores.", "authors": ["Zhendong Mi", "Renming Zheng", "Haowen Zhong", "Yue Sun", "Seth Kneeland", "Sayan Moitra", "Ken Kutzer", "Zhaozhuo Xu Shaoyi Huang"], "categories": ["cs.LG", "cs.AI", "cs.AR", "cs.PL", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-15", "url": "https://arxiv.org/abs/2412.11014", "pdf_url": "https://arxiv.org/pdf/2412.11014v2", "arxiv_id": "2412.11014", "doi": null, "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.25}
{"id": "e262abe4f5efbb0f3a6f9465b1ee40d0c61a9628e1b357a418c451ea98718a2d", "sources": ["arxiv", "semantic_scholar"], "title": "From Allies to Adversaries: Manipulating LLM Tool-Calling through Adversarial Injection", "abstract": "Tool-calling has changed Large Language Model (LLM) applications by integrating external tools, significantly enhancing their functionality across diverse tasks. However, this integration also introduces new security vulnerabilities, particularly in the tool scheduling mechanisms of LLM, which have not been extensively studied. To fill this gap, we present ToolCommander, a novel framework designed to exploit vulnerabilities in LLM tool-calling systems through adversarial tool injection. Our framework employs a well-designed two-stage attack strategy. Firstly, it injects malicious tools to collect user queries, then dynamically updates the injected tools based on the stolen information to enhance subsequent attacks. These stages enable ToolCommander to execute privacy theft, launch denial-of-service attacks, and even manipulate business competition by triggering unscheduled tool-calling. Notably, the ASR reaches 91.67% for privacy theft and hits 100% for denial-of-service and unscheduled tool calling in certain cases. Our work demonstrates that these vulnerabilities can lead to severe consequences beyond simple misuse of tool-calling systems, underscoring the urgent need for robust defensive strategies to secure LLM Tool-calling systems.", "authors": ["Haowei Wang", "Rupeng Zhang", "Junjie Wang", "Mingyang Li", "Yuekai Huang", "Dandan Wang", "Qing Wang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-13", "url": "https://arxiv.org/abs/2412.10198", "pdf_url": "https://arxiv.org/pdf/2412.10198v2", "arxiv_id": "2412.10198", "doi": "10.18653/v1/2025.naacl-long.101", "citation_count": 33, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.3829}
{"id": "2b5d17356a95c493a5e3b653a224ad2a3af8be87626584cd249d7c5381274298", "sources": ["arxiv", "semantic_scholar"], "title": "Generative Agents for Multi-Agent Autoformalization of Interaction Scenarios", "abstract": "Multi-agent simulations are versatile tools for exploring interactions among natural and artificial agents, but their development typically demands domain expertise and manual effort. This work introduces the Generative Agents for Multi-Agent Autoformalization (GAMA) framework, which automates the formalization of interaction scenarios in simulations using agents augmented with large language models (LLMs). To demonstrate the application of GAMA, we use natural language descriptions of game-theoretic scenarios representing social interactions, and we autoformalize them into executable logic programs defining game rules, with syntactic correctness enforced through a solver-based validation. To ensure runtime validity, an iterative, tournament-based procedure tests the generated rules and strategies, followed by exact semantic validation when ground truth outcomes are available. In experiments with 110 natural language descriptions across five 2x2 simultaneous-move games, GAMA achieves 100% syntactic and 76.5% semantic correctness with Claude 3.5 Sonnet, and 99.82% syntactic and 77% semantic correctness with GPT-4o. The framework also shows high semantic accuracy in autoformalizing agents' strategies.", "authors": ["Agnieszka Mensfelt", "Kostas Stathis", "Vince Trencsenyi"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-11", "url": "https://arxiv.org/abs/2412.08805", "pdf_url": "https://arxiv.org/pdf/2412.08805v3", "arxiv_id": "2412.08805", "doi": "10.3233/FAIA251256", "citation_count": 1, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/dicelab-rhul/GAMA", "venue": "European Conference on Artificial Intelligence", "quality_score": 0.0753}
{"id": "d13c14424ea7c10728022b724a0071e7aca2d9f0b86abbf7c4ff977d337a1aeb", "sources": ["arxiv", "semantic_scholar"], "title": "TapeAgents: a Holistic Framework for Agent Development and Optimization", "abstract": "We present TapeAgents, an agent framework built around a granular, structured log tape of the agent session that also plays the role of the session's resumable state. In TapeAgents we leverage tapes to facilitate all stages of the LLM Agent development lifecycle. The agent reasons by processing the tape and the LLM output to produce new thought and action steps and append them to the tape. The environment then reacts to the agent's actions by likewise appending observation steps to the tape. By virtue of this tape-centred design, TapeAgents can provide AI practitioners with holistic end-to-end support. At the development stage, tapes facilitate session persistence, agent auditing, and step-by-step debugging. Post-deployment, one can reuse tapes for evaluation, fine-tuning, and prompt-tuning; crucially, one can adapt tapes from other agents or use revised historical tapes. In this report, we explain the TapeAgents design in detail. We demonstrate possible applications of TapeAgents with several concrete examples of building monolithic agents and multi-agent teams, of optimizing agent prompts and finetuning the agent's LLM. We present tooling prototypes and report a case study where we use TapeAgents to finetune a Llama-3.1-8B form-filling assistant to perform as well as GPT-4o while being orders of magnitude cheaper. Lastly, our comparative analysis shows that TapeAgents's advantages over prior frameworks stem from our novel design of the LLM agent as a resumable, modular state machine with a structured configuration, that generates granular, structured logs and that can transform these logs into training text -- a unique combination of features absent in previous work.", "authors": ["Dzmitry Bahdanau", "Nicolas Gontier", "Gabriel Huang", "Ehsan Kamalloo", "Rafael Pardinas", "Alex Piché", "Torsten Scholak", "Oleh Shliazhko", "Jordan Prince Tremblay", "Karam Ghanem", "Soham Parikh", "Mitul Tiwari", "Quaizar Vohra"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-11", "url": "https://arxiv.org/abs/2412.08445", "pdf_url": "https://arxiv.org/pdf/2412.08445v1", "arxiv_id": "2412.08445", "doi": "10.48550/arXiv.2412.08445", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2603}
{"id": "b1d0d06ad86cd8e1c4f34dc52492166bd6ad714c68e398711626688c4e7ea57b", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluation Agent: Efficient and Promptable Evaluation Framework for Visual Generative Models", "abstract": "Recent advancements in visual generative models have enabled high-quality image and video generation, opening diverse applications. However, evaluating these models often demands sampling hundreds or thousands of images or videos, making the process computationally expensive, especially for diffusion-based models with inherently slow sampling. Moreover, existing evaluation methods rely on rigid pipelines that overlook specific user needs and provide numerical results without clear explanations. In contrast, humans can quickly form impressions of a model's capabilities by observing only a few samples. To mimic this, we propose the Evaluation Agent framework, which employs human-like strategies for efficient, dynamic, multi-round evaluations using only a few samples per round, while offering detailed, user-tailored analyses. It offers four key advantages: 1) efficiency, 2) promptable evaluation tailored to diverse user needs, 3) explainability beyond single numerical scores, and 4) scalability across various models and tools. Experiments show that Evaluation Agent reduces evaluation time to 10% of traditional methods while delivering comparable results. The Evaluation Agent framework is fully open-sourced to advance research in visual generative models and their efficient evaluation.", "authors": ["Fan Zhang", "Shulin Tian", "Ziqi Huang", "Yu Qiao", "Ziwei Liu"], "categories": ["cs.CV", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-10", "url": "https://arxiv.org/abs/2412.09645", "pdf_url": "https://arxiv.org/pdf/2412.09645v3", "arxiv_id": "2412.09645", "doi": "10.48550/arXiv.2412.09645", "citation_count": 39, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/Vchitect/Evaluation-Agent", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4005}
{"id": "edc55601c76a08dac542b4a8ae0448bdfe496efa8fe9357eb9baf7afe8e3fa87", "sources": ["arxiv", "semantic_scholar"], "title": "Where Common Knowledge Cannot Be Formed, Common Belief Can -- Planning with Multi-Agent Belief Using Group Justified Perspectives", "abstract": "Epistemic planning is the sub-field of AI planning that focuses on changing knowledge and belief. It is important in both multi-agent domains where agents need to have knowledge/belief regarding the environment, but also the beliefs of other agents, including nested beliefs. When modeling knowledge in multi-agent settings, many models face an exponential growth challenge in terms of nested depth. A contemporary method, known as Planning with Perspectives (PWP), addresses these challenges through the use of perspectives and set operations for knowledge. The JP model defines that an agent's belief is justified if and only if the agent has seen evidence that this belief was true in the past and has not seen evidence to suggest that this has changed. The current paper extends the JP model to handle \\emph{group belief}, including distributed belief and common belief. We call this the Group Justified Perspective (GJP) model. Using experimental problems crafted by adapting well-known benchmarks to a group setting, we show the efficiency and expressiveness of our GJP model at handling planning problems that cannot be handled by other epistemic planning tools.", "authors": ["Guang Hu", "Tim Miller", "Nir Lipovetzky"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-10", "url": "https://arxiv.org/abs/2412.07981", "pdf_url": "https://arxiv.org/pdf/2412.07981v2", "arxiv_id": "2412.07981", "doi": "10.48550/arXiv.2412.07981", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "6ccdf56f3947b0e05e3ea13d3b543e08e6eab9394b58c02ba25da8a9cc16ba12", "sources": ["arxiv", "semantic_scholar"], "title": "Augmenting the action space with conventions to improve multi-agent cooperation in Hanabi", "abstract": "The card game Hanabi is considered a strong medium for the testing and development of multi-agent reinforcement learning (MARL) algorithms, due to its cooperative nature, partial observability, limited communication and remarkable complexity. Previous research efforts have explored the capabilities of MARL algorithms within Hanabi, focusing largely on advanced architecture design and algorithmic manipulations to achieve state-of-the-art performance for various number of cooperators. However, this often leads to complex solution strategies with high computational cost and requiring large amounts of training data. For humans to solve the Hanabi game effectively, they require the use of conventions, which often allows for a means to implicitly convey ideas or knowledge based on a predefined, and mutually agreed upon, set of \"rules\" or principles. Multi-agent problems containing partial observability, especially when limited communication is present, can benefit greatly from the use of implicit knowledge sharing. In this paper, we propose a novel approach to augmenting an agent's action space using conventions, which act as a sequence of special cooperative actions that span over and include multiple time steps and multiple agents, requiring agents to actively opt in for it to reach fruition. These conventions are based on existing human conventions, and result in a significant improvement on the performance of existing techniques for self-play and cross-play for various number of cooperators within Hanabi.", "authors": ["F. Bredell", "H. A. Engelbrecht", "J. C. Schoeman"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-09", "url": "https://arxiv.org/abs/2412.06333", "pdf_url": "https://arxiv.org/pdf/2412.06333v3", "arxiv_id": "2412.06333", "doi": "10.1007/s10458-025-09709-5", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.0}
{"id": "86734b04f21694b25d3d9e43ca903d72b14fb352b42b8459f3fabd8123411b6e", "sources": ["arxiv", "semantic_scholar"], "title": "Asynchronous LLM Function Calling", "abstract": "Large language models (LLMs) use function calls to interface with external tools and data source. However, the current approach to LLM function calling is inherently synchronous, where each call blocks LLM inference, limiting LLM operation and concurrent function execution. In this work, we propose AsyncLM, a system for asynchronous LLM function calling. AsyncLM improves LLM's operational efficiency by enabling LLMs to generate and execute function calls concurrently. Instead of waiting for each call's completion, AsyncLM introduces an interrupt mechanism to asynchronously notify the LLM in-flight when function calls return. We design an in-context protocol for function calls and interrupts, provide fine-tuning strategy to adapt LLMs to the interrupt semantics, and implement these mechanisms efficiently on LLM inference process. We demonstrate that AsyncLM can reduce end-to-end task completion latency from 1.6x-5.4x compared to synchronous function calling on a set of benchmark tasks in the Berkeley function calling leaderboard (BFCL). Furthermore, we discuss how interrupt mechanisms can be extended to enable novel human-LLM or LLM-LLM interactions.", "authors": ["In Gim", "Seung-seob Lee", "Lin Zhong"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-09", "url": "https://arxiv.org/abs/2412.07017", "pdf_url": "https://arxiv.org/pdf/2412.07017v1", "arxiv_id": "2412.07017", "doi": "10.48550/arXiv.2412.07017", "citation_count": 22, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3404}
{"id": "30eb0d01ca491ffdaf7500ffa623bd375867ff0f6cfbf99129fb2b8022cd3180", "sources": ["arxiv", "semantic_scholar"], "title": "AgentAlign: Misalignment-Adapted Multi-Agent Perception for Resilient Inter-Agent Sensor Correlations", "abstract": "Cooperative perception has attracted wide attention given its capability to leverage shared information across connected automated vehicles (CAVs) and smart infrastructures to address sensing occlusion and range limitation issues. However, existing research overlooks the fragile multi-sensor correlations in multi-agent settings, as the heterogeneous agent sensor measurements are highly susceptible to environmental factors, leading to weakened inter-agent sensor interactions. The varying operational conditions and other real-world factors inevitably introduce multifactorial noise and consequentially lead to multi-sensor misalignment, making the deployment of multi-agent multi-modality perception particularly challenging in the real world. In this paper, we propose AgentAlign, a real-world heterogeneous agent cross-modality feature alignment framework, to effectively address these multi-modality misalignment issues. Our method introduces a cross-modality feature alignment space (CFAS) and heterogeneous agent feature alignment (HAFA) mechanism to harmonize multi-modality features across various agents dynamically. Additionally, we present a novel V2XSet-noise dataset that simulates realistic sensor imperfections under diverse environmental conditions, facilitating a systematic evaluation of our approach's robustness. Extensive experiments on the V2X-Real and V2XSet-Noise benchmarks demonstrate that our framework achieves state-of-the-art performance, underscoring its potential for real-world applications in cooperative autonomous driving. The controllable V2XSet-Noise dataset and generation pipeline will be released in the future.", "authors": ["Zonglin Meng", "Yun Zhang", "Zhaoliang Zheng", "Zhihao Zhao", "Jiaqi Ma"], "categories": ["cs.CV", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-09", "url": "https://arxiv.org/abs/2412.06142", "pdf_url": "https://arxiv.org/pdf/2412.06142v1", "arxiv_id": "2412.06142", "doi": "10.48550/arXiv.2412.06142", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "fbd34a998c18313f3185100561c92ee6d71efc35201622fe3bab805b16920fbd", "sources": ["arxiv", "semantic_scholar"], "title": "A Collaborative Multi-Agent Approach to Retrieval-Augmented Generation Across Diverse Data", "abstract": "Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by incorporating external, domain-specific data into the generative process. While LLMs are highly capable, they often rely on static, pre-trained datasets, limiting their ability to integrate dynamic or private data. Traditional RAG systems typically use a single-agent architecture to handle query generation, data retrieval, and response synthesis. However, this approach becomes inefficient when dealing with diverse data sources, such as relational databases, document stores, and graph databases, often leading to performance bottlenecks and reduced accuracy. This paper proposes a multi-agent RAG system to address these limitations. Specialized agents, each optimized for a specific data source, handle query generation for relational, NoSQL, and document-based systems. These agents collaborate within a modular framework, with query execution delegated to an environment designed for compatibility across various database types. This distributed approach enhances query efficiency, reduces token overhead, and improves response accuracy by ensuring that each agent focuses on its specialized task. The proposed system is scalable and adaptable, making it ideal for generative AI workflows that require integration with diverse, dynamic, or private data sources. By leveraging specialized agents and a modular execution environment, the system provides an efficient and robust solution for handling complex, heterogeneous data environments in generative AI applications.", "authors": ["Aniruddha Salve", "Saba Attar", "Mahesh Deshmukh", "Sayali Shivpuje", "Arnab Mitra Utsab"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-08", "url": "https://arxiv.org/abs/2412.05838", "pdf_url": "https://arxiv.org/pdf/2412.05838v1", "arxiv_id": "2412.05838", "doi": "10.48550/arXiv.2412.05838", "citation_count": 15, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "40a3a27781f2248e86aaab6139fe4fa3d234f3308bc10c3c4282b27980fd14d7", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative SQL Generation for Segmented Databases By Using Multi-functional LLM Agents", "abstract": "Text-to-SQL task aims to automatically yield SQL queries according to user text questions. To address this problem, we propose a Cooperative SQL Generation framework based on Multi-functional Agents (CSMA) through information interaction among large language model (LLM) based agents who own part of the database schema seperately. Inspired by the collaboration in human teamwork, CSMA consists of three stages: 1) Question-related schema collection, 2) Question-corresponding SQL query generation, and 3) SQL query correctness check. In the first stage, agents analyze their respective schema and communicate with each other to collect the schema information relevant to the question. In the second stage, agents try to generate the corresponding SQL query for the question using the collected information. In the third stage, agents check if the SQL query is created correctly according to their known information. This interaction-based method makes the question-relevant part of database schema from each agent to be used for SQL generation and check. Experiments on the Spider and Bird benckmark demonstrate that CSMA achieves a high performance level comparable to the state-of-the-arts, meanwhile holding the private data in these individual agents.", "authors": ["Zhiguang Wu", "Fengbin Zhu", "Xuequn Shang", "Yupei Zhang", "Pan Zhou"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-08", "url": "https://arxiv.org/abs/2412.05850", "pdf_url": "https://arxiv.org/pdf/2412.05850v1", "arxiv_id": "2412.05850", "doi": "10.48550/arXiv.2412.05850", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "624753fccab0fa78036358550bb6a5fc24ad6968d82d3e954b0dfe65a5ef1eca", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Effective GenAI Multi-Agent Collaboration: Design and Evaluation for Enterprise Applications", "abstract": "AI agents powered by large language models (LLMs) have shown strong capabilities in problem solving. Through combining many intelligent agents, multi-agent collaboration has emerged as a promising approach to tackle complex, multi-faceted problems that exceed the capabilities of single AI agents. However, designing the collaboration protocols and evaluating the effectiveness of these systems remains a significant challenge, especially for enterprise applications. This report addresses these challenges by presenting a comprehensive evaluation of coordination and routing capabilities in a novel multi-agent collaboration framework. We evaluate two key operational modes: (1) a coordination mode enabling complex task completion through parallel communication and payload referencing, and (2) a routing mode for efficient message forwarding between agents. We benchmark on a set of handcrafted scenarios from three enterprise domains, which are publicly released with the report. For coordination capabilities, we demonstrate the effectiveness of inter-agent communication and payload referencing mechanisms, achieving end-to-end goal success rates of 90%. Our analysis yields several key findings: multi-agent collaboration enhances goal success rates by up to 70% compared to single-agent approaches in our benchmarks; payload referencing improves performance on code-intensive tasks by 23%; latency can be substantially reduced with a routing mechanism that selectively bypasses agent orchestration. These findings offer valuable guidance for enterprise deployments of multi-agent systems and advance the development of scalable, efficient multi-agent collaboration frameworks.", "authors": ["Raphael Shu", "Nilaksh Das", "Michelle Yuan", "Monica Sunkara", "Yi Zhang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-06", "url": "https://arxiv.org/abs/2412.05449", "pdf_url": "https://arxiv.org/pdf/2412.05449v1", "arxiv_id": "2412.05449", "doi": "10.48550/arXiv.2412.05449", "citation_count": 28, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3656}
{"id": "8bb82e83de27470fbf023ac654b9b8b7a3971fa2d743201cac2c31abb4fbc756", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing LLMs for Impression Generation in Radiology Reports through a Multi-Agent System", "abstract": "This study introduces \"RadCouncil,\" a multi-agent Large Language Model (LLM) framework designed to enhance the generation of impressions in radiology reports from the finding section. RadCouncil comprises three specialized agents: 1) a \"Retrieval\" Agent that identifies and retrieves similar reports from a vector database, 2) a \"Radiologist\" Agent that generates impressions based on the finding section of the given report plus the exemplar reports retrieved by the Retrieval Agent, and 3) a \"Reviewer\" Agent that evaluates the generated impressions and provides feedback. The performance of RadCouncil was evaluated using both quantitative metrics (BLEU, ROUGE, BERTScore) and qualitative criteria assessed by GPT-4, using chest X-ray as a case study. Experiment results show improvements in RadCouncil over the single-agent approach across multiple dimensions, including diagnostic accuracy, stylistic concordance, and clarity. This study highlights the potential of utilizing multiple interacting LLM agents, each with a dedicated task, to enhance performance in specialized medical tasks and the development of more robust and adaptable healthcare AI solutions.", "authors": ["Fang Zeng", "Zhiliang Lyu", "Quanzheng Li", "Xiang Li"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-06", "url": "https://arxiv.org/abs/2412.06828", "pdf_url": "https://arxiv.org/pdf/2412.06828v1", "arxiv_id": "2412.06828", "doi": "10.48550/arXiv.2412.06828", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "b69beb5ad6f07fa6e30a9f4cc98c46e52818772d9c8bea5a3077fb9d6b8b81b4", "sources": ["arxiv", "semantic_scholar"], "title": "TOOL-ED: Enhancing Empathetic Response Generation with the Tool Calling Capability of LLM", "abstract": "Empathetic conversation is a crucial characteristic in daily conversations between individuals. Nowadays, Large Language models (LLMs) have shown outstanding performance in generating empathetic responses. Knowledge bases like COMET can assist LLMs in mitigating illusions and enhancing the understanding of users' intentions and emotions. However, models remain heavily reliant on fixed knowledge bases and unrestricted incorporation of external knowledge can introduce noise. Tool learning is a flexible end-to-end approach that assists LLMs in handling complex problems. In this paper, we propose Emotional Knowledge Tool Calling (EKTC) framework, which encapsulates the commonsense knowledge bases as empathetic tools, enabling LLMs to integrate external knowledge flexibly through tool calling. In order to adapt the models to the new task, we construct a novel dataset TOOL-ED based on the EMPATHETICMPATHETIC DIALOGUE (ED) dataset. We validate EKTC on the ED dataset, and the experimental results demonstrate that our framework can enhance the ability of LLMs to generate empathetic responses effectively.", "authors": ["Huiying Cao", "Yiqun Zhang", "Shi Feng", "Xiaocui Yang", "Daling Wang", "Yifei Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-04", "url": "https://arxiv.org/abs/2412.03096", "pdf_url": "https://arxiv.org/pdf/2412.03096v2", "arxiv_id": "2412.03096", "doi": "10.48550/arXiv.2412.03096", "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Computational Linguistics", "quality_score": 0.2698}
{"id": "95c30166b578e875109dc220e0f1dca7bb9bab01524401769714c3928444f28b", "sources": ["arxiv", "semantic_scholar"], "title": "MALT: Improving Reasoning with Multi-Agent LLM Training", "abstract": "Large Language Models (LLMs) often produce answers with a single chain-of-thought, which restricts their ability to explore reasoning paths or self-correct flawed outputs in complex tasks. In this paper, we introduce MALT (Multi-Agent LLM Training), a novel post-training strategy that divides the reasoning process into generation, verification, and refinement steps using a sequential pipeline of heterogeneous agents. During data generation, each agent is repeatedly sampled to form a multi-agent search tree, where final outputs are graded against ground-truth data. We then apply value iteration to propagate reward signals back to each role-conditioned model, automatically producing multi-agent post-training data without human or teacher-model supervision. Our off-policy approach allows each agent to specialize by learning from correct and incorrect trajectories, ultimately improving the end-to-end reasoning chain. On MATH, GSM8K, and CSQA, MALT surpasses the same baseline LLM with a relative improvement of 15.66%, 7.42%, and 9.40% respectively, making it an important advance towards multi-agent cooperative training.", "authors": ["Sumeet Ramesh Motwani", "Chandler Smith", "Rocktim Jyoti Das", "Rafael Rafailov", "Ivan Laptev", "Philip H. S. Torr", "Fabio Pizzati", "Ronald Clark", "Christian Schroeder de Witt"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-02", "url": "https://arxiv.org/abs/2412.01928", "pdf_url": "https://arxiv.org/pdf/2412.01928v3", "arxiv_id": "2412.01928", "doi": "10.48550/arXiv.2412.01928", "citation_count": 56, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.439}
{"id": "4d3be7df8dbe5d7f5bc65e7fc42f512f74b7a965fb2ac56a1d4829acdc5ab317", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaboration in Incident Response with Large Language Models", "abstract": "Incident response (IR) is a critical aspect of cybersecurity, requiring rapid decision-making and coordinated efforts to address cyberattacks effectively. Leveraging large language models (LLMs) as intelligent agents offers a novel approach to enhancing collaboration and efficiency in IR scenarios. This paper explores the application of LLM-based multi-agent collaboration using the Backdoors & Breaches framework, a tabletop game designed for cybersecurity training. We simulate real-world IR dynamics through various team structures, including centralized, decentralized, and hybrid configurations. By analyzing agent interactions and performance across these setups, we provide insights into optimizing multi-agent collaboration for incident response. Our findings highlight the potential of LLMs to enhance decision-making, improve adaptability, and streamline IR processes, paving the way for more effective and coordinated responses to cyber threats.", "authors": ["Zefang Liu"], "categories": ["cs.CL", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2024-12-01", "url": "https://arxiv.org/abs/2412.00652", "pdf_url": "https://arxiv.org/pdf/2412.00652v2", "arxiv_id": "2412.00652", "doi": "10.48550/arXiv.2412.00652", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "52a1ea354b87b4dc83f76962940f4213a60d1e76a743d5a4e687b3bdc171ae9a", "sources": ["arxiv", "semantic_scholar"], "title": "MAG-V: A Multi-Agent Framework for Synthetic Data Generation and Verification", "abstract": "Extending the capabilities of Large Language Models (LLMs) with functions or tools for environment interaction has led to the emergence of the agent paradigm. In industry, training an LLM is not always feasible because of the scarcity of domain data, legal holds on proprietary customer data, rapidly changing business requirements, and the need to prototype new assistants. Agents provide an elegant solution to the above by relying on the zero-shot reasoning abilities of the underlying LLM and utilizing tools to explore and reason over customer data and respond to user requests. However, there are two concerns here: (I) acquiring large scale customer queries for agent testing is time-consuming, and (II) high reliance on the tool call sequence (or trajectory) followed by the agent to respond to user queries may lead to unexpected or incorrect behavior. To address this, we propose MAG-V, a multi-agent framework to first generate a dataset of questions that mimic customer queries; and second, reverse-engineer alternate questions from the responses for trajectory verification. Initial results indicate that our synthetic data can improve agent performance on actual customer queries. Furthermore, our trajectory verification methodology, inspired by distant supervision and using traditional machine learning (ML) models, outperforms a GPT-4o judge baseline by 11% accuracy and matches the performance of a GPT-4 judge on our constructed dataset. Overall, our approach is a step towards unifying diverse task agents into a cohesive framework for achieving an aligned objective.", "authors": ["Saptarshi Sengupta", "Harsh Vashistha", "Kristal Curtis", "Akshay Mallipeddi", "Abhinav Mathur", "Joseph Ross", "Liang Gou"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-28", "url": "https://arxiv.org/abs/2412.04494", "pdf_url": "https://arxiv.org/pdf/2412.04494v2", "arxiv_id": "2412.04494", "doi": "10.48550/arXiv.2412.04494", "citation_count": 16, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3076}
{"id": "3e03d1757ae4f80f1ebffafcdcdecaf5110f5aea288158f95f9030bb925b0ae7", "sources": ["arxiv", "semantic_scholar"], "title": "Large Language Model-Brained GUI Agents: A Survey", "abstract": "GUIs have long been central to human-computer interaction, providing an intuitive and visually-driven way to access and interact with digital systems. The advent of LLMs, particularly multimodal models, has ushered in a new era of GUI automation. They have demonstrated exceptional capabilities in natural language understanding, code generation, and visual processing. This has paved the way for a new generation of LLM-brained GUI agents capable of interpreting complex GUI elements and autonomously executing actions based on natural language instructions. These agents represent a paradigm shift, enabling users to perform intricate, multi-step tasks through simple conversational commands. Their applications span across web navigation, mobile app interactions, and desktop automation, offering a transformative user experience that revolutionizes how individuals interact with software. This emerging field is rapidly advancing, with significant progress in both research and industry. To provide a structured understanding of this trend, this paper presents a comprehensive survey of LLM-brained GUI agents, exploring their historical evolution, core components, and advanced techniques. We address research questions such as existing GUI agent frameworks, the collection and utilization of data for training specialized GUI agents, the development of large action models tailored for GUI tasks, and the evaluation metrics and benchmarks necessary to assess their effectiveness. Additionally, we examine emerging applications powered by these agents. Through a detailed analysis, this survey identifies key research gaps and outlines a roadmap for future advancements in the field. By consolidating foundational knowledge and state-of-the-art developments, this work aims to guide both researchers and practitioners in overcoming challenges and unlocking the full potential of LLM-brained GUI agents.", "authors": ["Chaoyun Zhang", "Shilin He", "Jiaxu Qian", "Bowen Li", "Liqun Li", "Si Qin", "Yu Kang", "Minghua Ma", "Guyue Liu", "Qingwei Lin", "Saravan Rajmohan", "Dongmei Zhang", "Qi Zhang"], "categories": ["cs.AI", "cs.CL", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-27", "url": "https://arxiv.org/abs/2411.18279", "pdf_url": "https://arxiv.org/pdf/2411.18279v12", "arxiv_id": "2411.18279", "doi": "10.48550/arXiv.2411.18279", "citation_count": 172, "influential_citation_count": 6, "has_code": true, "code_url": "https://github.com/vyokky/LLM-Brained-GUI-Agents-Survey", "venue": "arXiv.org", "quality_score": 0.5595}
{"id": "33977c19dec724b5e735cca883c9ba3b85ef4834fae99e4c19c95b2f45577e76", "sources": ["arxiv", "semantic_scholar"], "title": "Exploration of LLM Multi-Agent Application Implementation Based on LangGraph+CrewAI", "abstract": "With the rapid development of large model technology, the application of agent technology in various fields is becoming increasingly widespread, profoundly changing people's work and lifestyles. In complex and dynamic systems, multi-agents achieve complex tasks that are difficult for a single agent to complete through division of labor and collaboration among agents. This paper discusses the integrated application of LangGraph and CrewAI. LangGraph improves the efficiency of information transmission through graph architecture, while CrewAI enhances team collaboration capabilities and system performance through intelligent task allocation and resource management. The main research contents of this paper are: (1) designing the architecture of agents based on LangGraph for precise control; (2) enhancing the capabilities of agents based on CrewAI to complete a variety of tasks. This study aims to delve into the application of LangGraph and CrewAI in multi-agent systems, providing new perspectives for the future development of agent technology, and promoting technological progress and application innovation in the field of large model intelligent agents.", "authors": ["Zhihua Duan", "Jialin Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-27", "url": "https://arxiv.org/abs/2411.18241", "pdf_url": "https://arxiv.org/pdf/2411.18241v1", "arxiv_id": "2411.18241", "doi": "10.48550/arXiv.2411.18241", "citation_count": 50, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4269}
{"id": "a4a3468138a63665118b08cb5e20aa688e1a0c7039f310dd89a47a708a8abfa6", "sources": ["arxiv", "semantic_scholar"], "title": "CATP-LLM: Empowering Large Language Models for Cost-Aware Tool Planning", "abstract": "Utilizing large language models (LLMs) for tool planning has emerged as a promising avenue for developing general AI systems, where LLMs automatically schedule external tools (e.g., vision models) to tackle complex tasks based on task descriptions. To push this paradigm toward practical applications, it is crucial for LLMs to consider tool execution costs (e.g., execution time) for tool planning. Unfortunately, prior studies overlook the tool execution costs, leading to the generation of expensive plans whose costs outweigh their benefits in terms of task performance. To fill this gap, we propose the Cost-Aware Tool Planning with LLMs (CATP-LLM) framework, which for the first time provides a coherent design to empower LLMs for cost-aware tool planning. Specifically, To facilitate efficient concurrent tool execution and cost reduction, we design a tool planning language to enhance the LLM for creating multi-branch non-sequential plans. Moreover, we propose a cost-aware offline reinforcement learning algorithm to fine-tune the LLM to optimize the performance-cost trade-off in tool planning. In the lack of public cost-related datasets, we further present OpenCATP, the first dataset for cost-aware planning, which comprises 11,100 evaluation samples from diverse tasks. Extensive experiments show that CATP-LLM outperforms GPT-4 even when using Llama2-7B as its backbone, with the average improvement of 1.5%-93.9% in terms of plan quality. Codes and dataset are available at: https://github.com/duowuyms/OpenCATP-LLM.", "authors": ["Duo Wu", "Jinghe Wang", "Yuan Meng", "Yanning Zhang", "Le Sun", "Zhi Wang"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-25", "url": "https://arxiv.org/abs/2411.16313", "pdf_url": "https://arxiv.org/pdf/2411.16313v3", "arxiv_id": "2411.16313", "doi": "10.1109/ICCV51701.2025.00814", "citation_count": 10, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/duowuyms/OpenCATP-LLM", "venue": "IEEE International Conference on Computer Vision", "quality_score": 0.2603}
{"id": "f6e7f845d51c44fb3c890b74c84b01b5ddf042b1958cff85515e342b64fd7c71", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Multi-Agent Consensus through Third-Party LLM Integration: Analyzing Uncertainty and Mitigating Hallucinations in Large Language Models", "abstract": "Large Language Models (LLMs) still face challenges when dealing with complex reasoning tasks, often resulting in hallucinations, which limit the practical application of LLMs. To alleviate this issue, this paper proposes a new method that integrates different LLMs to expand the knowledge boundary, reduce dependence on a single model, and promote in-depth debate among agents. The main contributions include: 1) Introducing third-party LLMs to adjust the attention weights of agents through uncertainty estimation and confidence analysis, optimizing consensus formation in multi-agent systems; 2) Experiments on arithmetic datasets have validated the effectiveness of the method, surpassing traditional multi-agent baselines. This research provides a new perspective for large models to alleviate hallucination phenomena when dealing with complex tasks.", "authors": ["Zhihua Duan", "Jialin Wang"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-25", "url": "https://arxiv.org/abs/2411.16189", "pdf_url": "https://arxiv.org/pdf/2411.16189v1", "arxiv_id": "2411.16189", "doi": "10.1109/ICAACE65325.2025.11019272", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2258}
{"id": "73a63559365fe70bc3975e930cf15d2ffccb3077854c29c17a174ca2ef27fdf8", "sources": ["arxiv", "semantic_scholar"], "title": "DrugAgent: Automating AI-aided Drug Discovery Programming through LLM Multi-Agent Collaboration", "abstract": "Recent progress in Large Language Models (LLMs) has drawn attention to their potential for accelerating drug discovery. However, a central problem remains: translating theoretical ideas into robust implementations in the highly specialized context of pharmaceutical research. This limitation prevents practitioners from making full use of the latest AI developments in drug discovery. To address this challenge, we introduce DrugAgent, a multi-agent framework that automates machine learning (ML) programming for drug discovery tasks. DrugAgent employs an LLM Planner that formulates high-level ideas and an LLM Instructor that identifies and integrates domain knowledge when implementing those ideas. We present case studies on three representative drug discovery tasks. Our results show that DrugAgent consistently outperforms leading baselines, including a relative improvement of 4.92% in ROC-AUC compared to ReAct for drug-target interaction (DTI). DrugAgent is publicly available at https://anonymous.4open.science/r/drugagent-5C42/.", "authors": ["Sizhe Liu", "Yizhou Lu", "Siyu Chen", "Xiyang Hu", "Jieyu Zhao", "Yingzhou Lu", "Yue Zhao"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-24", "url": "https://arxiv.org/abs/2411.15692", "pdf_url": "https://arxiv.org/pdf/2411.15692v2", "arxiv_id": "2411.15692", "doi": "10.48550/arXiv.2411.15692", "citation_count": 41, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4058}
{"id": "19d20a26f0abbb9cb6314eb1e151eafd4223f6ef4d014ba9c7a7a15b57e31288", "sources": ["arxiv"], "title": "Less is More: Optimizing Function Calling for LLM Execution on Edge Devices", "abstract": "The advanced function-calling capabilities of foundation models open up new possibilities for deploying agents to perform complex API tasks. However, managing large amounts of data and interacting with numerous APIs makes function calling hardware-intensive and costly, especially on edge devices. Current Large Language Models (LLMs) struggle with function calling at the edge because they cannot handle complex inputs or manage multiple tools effectively. This results in low task-completion accuracy, increased delays, and higher power consumption. In this work, we introduce Less-is-More, a novel fine-tuning-free function-calling scheme for dynamic tool selection. Our approach is based on the key insight that selectively reducing the number of tools available to LLMs significantly improves their function-calling performance, execution time, and power efficiency on edge devices. Experimental results with state-of-the-art LLMs on edge hardware show agentic success rate improvements, with execution time reduced by up to 70% and power consumption by up to 40%.", "authors": ["Varatheepan Paramanayakam", "Andreas Karatzas", "Iraklis Anagnostopoulos", "Dimitrios Stamoulis"], "categories": ["cs.PF", "cs.DC", "cs.LG"], "fields_of_study": [], "published_date": "2024-11-23", "url": "https://arxiv.org/abs/2411.15399", "pdf_url": "https://arxiv.org/pdf/2411.15399v1", "arxiv_id": "2411.15399", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0}
{"id": "f6412ac6eb969e71154f09894ce2dc614ea58180d6c73482e2962efa22bc9ea3", "sources": ["arxiv", "semantic_scholar"], "title": "Regulator-Manufacturer AI Agents Modeling: Mathematical Feedback-Driven Multi-Agent LLM Framework", "abstract": "The increasing complexity of regulatory updates from global authorities presents significant challenges for medical device manufacturers, necessitating agile strategies to sustain compliance and maintain market access. Concurrently, regulatory bodies must effectively monitor manufacturers' responses and develop strategic surveillance plans. This study employs a multi-agent modeling approach, enhanced with Large Language Models (LLMs), to simulate regulatory dynamics and examine the adaptive behaviors of key actors, including regulatory bodies, manufacturers, and competitors. These agents operate within a simulated environment governed by regulatory flow theory, capturing the impacts of regulatory changes on compliance decisions, market adaptation, and innovation strategies. Our findings illuminate the influence of regulatory shifts on industry behaviour and identify strategic opportunities for improving regulatory practices, optimizing compliance, and fostering innovation. By leveraging the integration of multi-agent systems and LLMs, this research provides a novel perspective and offers actionable insights for stakeholders navigating the evolving regulatory landscape of the medical device industry.", "authors": ["Yu Han", "Zekun Guo"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-22", "url": "https://arxiv.org/abs/2411.15356", "pdf_url": "https://arxiv.org/pdf/2411.15356v2", "arxiv_id": "2411.15356", "doi": "10.48550/arXiv.2411.15356", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "7363d67ac7c2c47c0f67576b6feefc22226c15f1db7a40807dd29350a1dc7179", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based Multi-Agent Systems: Techniques and Business Perspectives", "abstract": "In the era of (multi-modal) large language models, most operational processes can be reformulated and reproduced using LLM agents. The LLM agents can perceive, control, and get feedback from the environment so as to accomplish the given tasks in an autonomous manner. Besides the environment-interaction property, the LLM agents can call various external tools to ease the task completion process. The tools can be regarded as a predefined operational process with private or real-time knowledge that does not exist in the parameters of LLMs. As a natural trend of development, the tools for calling are becoming autonomous agents, thus the full intelligent system turns out to be a LLM-based Multi-Agent System (LaMAS). Compared to the previous single-LLM-agent system, LaMAS has the advantages of i) dynamic task decomposition and organic specialization, ii) higher flexibility for system changing, iii) proprietary data preserving for each participating entity, and iv) feasibility of monetization for each entity. This paper discusses the technical and business landscapes of LaMAS. To support the ecosystem of LaMAS, we provide a preliminary version of such LaMAS protocol considering technical requirements, data privacy, and business incentives. As such, LaMAS would be a practical solution to achieve artificial collective intelligence in the near future.", "authors": ["Yingxuan Yang", "Qiuying Peng", "Jun Wang", "Ying Wen", "Weinan Zhang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-21", "url": "https://arxiv.org/abs/2411.14033", "pdf_url": "https://arxiv.org/pdf/2411.14033v2", "arxiv_id": "2411.14033", "doi": null, "citation_count": 31, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3763}
{"id": "16ad9d4a534b20e57daa6d3348f45c96c54c38179e8b2487a178e0c59dc50af3", "sources": ["arxiv", "semantic_scholar"], "title": "Emergent Structure in Multi-agent Systems Using Geometric Embeddings", "abstract": "This work investigates the self-organization of multi-agent systems into closed trajectories, a common requirement in unmanned aerial vehicle (UAV) surveillance tasks. In such scenarios, smooth, unbiased control signals save energy and mitigate mechanical strain. We propose a decentralized control system architecture that produces a globally stable emergent structure from local observations only; there is no requirement for agents to share a global plan or follow prescribed trajectories. Central to our approach is the formulation of an injective virtual embedding induced by rotations from the actual agent positions. This embedding serves as a structure-preserving map around which all agent stabilize their relative positions and permits the use of well-established linear control techniques. We construct the embedding such that it is topologically equivalent to the desired trajectory (i.e., a homeomorphism), thereby preserving the stability characteristics. We demonstrate the versatility of this approach through implementation on a swarm of Quanser QDrone quadcopters. Results demonstrate the quadcopters self-organize into the desired trajectory while maintaining even separation.", "authors": ["Dimitria Silveria", "Kleber Cabral", "Peter Jardine", "Sidney Givigi"], "categories": ["eess.SY", "cs.RO", "math.GT"], "fields_of_study": ["Computer Science", "Engineering", "Mathematics"], "published_date": "2024-11-17", "url": "https://arxiv.org/abs/2411.11142", "pdf_url": "https://arxiv.org/pdf/2411.11142v1", "arxiv_id": "2411.11142", "doi": "10.1109/ISSE63315.2024.10741086", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Information Security Solutions Europe", "quality_score": 0.0753}
{"id": "8a8d5d8de65e31822c0c70d3a48dd195c65a2ca0722c190dcd5562b6b6a83915", "sources": ["arxiv", "semantic_scholar"], "title": "BudgetMLAgent: A Cost-Effective LLM Multi-Agent system for Automating Machine Learning Tasks", "abstract": "Large Language Models (LLMs) excel in diverse applications including generation of code snippets, but often struggle with generating code for complex Machine Learning (ML) tasks. Although existing LLM single-agent based systems give varying performance depending on the task complexity, they purely rely on larger and expensive models such as GPT-4. Our investigation reveals that no-cost and low-cost models such as Gemini-Pro, Mixtral and CodeLlama perform far worse than GPT-4 in a single-agent setting. With the motivation of developing a cost-efficient LLM based solution for solving ML tasks, we propose an LLM Multi-Agent based system which leverages combination of experts using profiling, efficient retrieval of past observations, LLM cascades, and ask-the-expert calls. Through empirical analysis on ML engineering tasks in the MLAgentBench benchmark, we demonstrate the effectiveness of our system, using no-cost models, namely Gemini as the base LLM, paired with GPT-4 in cascade and expert to serve occasional ask-the-expert calls for planning. With 94.2\\% reduction in the cost (from \\$0.931 per run cost averaged over all tasks for GPT-4 single agent system to \\$0.054), our system is able to yield better average success rate of 32.95\\% as compared to GPT-4 single-agent system yielding 22.72\\% success rate averaged over all the tasks of MLAgentBench.", "authors": ["Shubham Gandhi", "Manasi Patwardhan", "Lovekesh Vig", "Gautam Shroff"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-12", "url": "https://arxiv.org/abs/2411.07464", "pdf_url": "https://arxiv.org/pdf/2411.07464v2", "arxiv_id": "2411.07464", "doi": "10.1145/3703412.3703416", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on AI-ML-Systems", "quality_score": 0.301}
{"id": "27248ca5e5ab07d1132f16831a485a339e45218095e9cf42a40796ee4f964541", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Stochastic Bandits Robust to Adversarial Corruptions", "abstract": "We study the problem of multi-agent multi-armed bandits with adversarial corruption in a heterogeneous setting, where each agent accesses a subset of arms. The adversary can corrupt the reward observations for all agents. Agents share these corrupted rewards with each other, and the objective is to maximize the cumulative total reward of all agents (and not be misled by the adversary). We propose a multi-agent cooperative learning algorithm that is robust to adversarial corruptions. For this newly devised algorithm, we demonstrate that an adversary with an unknown corruption budget $C$ only incurs an additive $O((L / L_{\\min}) C)$ term to the standard regret of the model in non-corruption settings, where $L$ is the total number of agents, and $L_{\\min}$ is the minimum number of agents with mutual access to an arm. As a side-product, our algorithm also improves the state-of-the-art regret bounds when reducing to both the single-agent and homogeneous multi-agent scenarios, tightening multiplicative $K$ (the number of arms) and $L$ (the number of agents) factors, respectively.", "authors": ["Fatemeh Ghaffari", "Xuchuang Wang", "Jinhang Zuo", "Mohammad Hajiesmaili"], "categories": ["cs.LG", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2024-11-12", "url": "https://arxiv.org/abs/2411.08167", "pdf_url": "https://arxiv.org/pdf/2411.08167v1", "arxiv_id": "2411.08167", "doi": "10.48550/arXiv.2411.08167", "citation_count": 3, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "3b4568be96adc2e02b28b8512c3694cbb34f7e65f292f0c59eb20be810924e33", "sources": ["arxiv", "semantic_scholar"], "title": "Using Generative AI and Multi-Agents to Provide Automatic Feedback", "abstract": "This study investigates the use of generative AI and multi-agent systems to provide automatic feedback in educational contexts, particularly for student constructed responses in science assessments. The research addresses a key gap in the field by exploring how multi-agent systems, called AutoFeedback, can improve the quality of GenAI-generated feedback, overcoming known issues such as over-praise and over-inference that are common in single-agent large language models (LLMs). The study developed a multi-agent system consisting of two AI agents: one for generating feedback and another for validating and refining it. The system was tested on a dataset of 240 student responses, and its performance was compared to that of a single-agent LLM. Results showed that AutoFeedback significantly reduced the occurrence of over-praise and over-inference errors, providing more accurate and pedagogically sound feedback. The findings suggest that multi-agent systems can offer a more reliable solution for generating automated feedback in educational settings, highlighting their potential for scalable and personalized learning support. These results have important implications for educators and researchers seeking to leverage AI in formative assessments, offering a pathway to more effective feedback mechanisms that enhance student learning outcomes.", "authors": ["Shuchen Guo", "Ehsan Latif", "Yifan Zhou", "Xuan Huang", "Xiaoming Zhai"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-11", "url": "https://arxiv.org/abs/2411.07407", "pdf_url": "https://arxiv.org/pdf/2411.07407v1", "arxiv_id": "2411.07407", "doi": "10.48550/arXiv.2411.07407", "citation_count": 35, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "e259a1366d6a959f113ddbb52cccf70764eaca8e58c7c59cfe763f4320c68e2d", "sources": ["arxiv", "semantic_scholar"], "title": "A Multi-Agent Approach for REST API Testing with Semantic Graphs and LLM-Driven Inputs", "abstract": "As modern web services increasingly rely on REST APIs, their thorough testing has become crucial. Furthermore, the advent of REST API documentation languages, such as the OpenAPI Specification, has led to the emergence of many black-box REST API testing tools. However, these tools often focus on individual test elements in isolation (e.g., APIs, parameters, values), resulting in lower coverage and less effectiveness in fault detection. To address these limitations, we present AutoRestTest, the first black-box tool to adopt a dependency-embedded multi-agent approach for REST API testing that integrates multi-agent reinforcement learning (MARL) with a semantic property dependency graph (SPDG) and Large Language Models (LLMs). Our approach treats REST API testing as a separable problem, where four agents -- API, dependency, parameter, and value agents -- collaborate to optimize API exploration. LLMs handle domain-specific value generation, the SPDG model simplifies the search space for dependencies using a similarity score between API operations, and MARL dynamically optimizes the agents' behavior. Our evaluation of AutoRestTest on 12 real-world REST services shows that it outperforms the four leading black-box REST API testing tools, including those assisted by RESTGPT (which generates realistic test inputs using LLMs), in terms of code coverage, operation coverage, and fault detection. Notably, AutoRestTest is the only tool able to trigger an internal server error in the Spotify service. Our ablation study illustrates that each component of AutoRestTest -- the SPDG, the LLM, and the agent-learning mechanism -- contributes to its overall effectiveness.", "authors": ["Myeongsoo Kim", "Tyler Stennett", "Saurabh Sinha", "Alessandro Orso"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-11", "url": "https://arxiv.org/abs/2411.07098", "pdf_url": "https://arxiv.org/pdf/2411.07098v2", "arxiv_id": "2411.07098", "doi": "10.1109/ICSE55347.2025.00179", "citation_count": 26, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "International Conference on Software Engineering", "quality_score": 0.3578}
{"id": "96a358c56a7d3c2bdc67c7ad66cdaabbb2f13580cbd922cb71844a4f5ea83ea1", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Collective Dynamics of Multi-Agent Systems using Event-based Vision", "abstract": "This paper proposes a novel problem: vision-based perception to learn and predict the collective dynamics of multi-agent systems, specifically focusing on interaction strength and convergence time. Multi-agent systems are defined as collections of more than ten interacting agents that exhibit complex group behaviors. Unlike prior studies that assume knowledge of agent positions, we focus on deep learning models to directly predict collective dynamics from visual data, captured as frames or events. Due to the lack of relevant datasets, we create a simulated dataset using a state-of-the-art flocking simulator, coupled with a vision-to-event conversion framework. We empirically demonstrate the effectiveness of event-based representation over traditional frame-based methods in predicting these collective behaviors. Based on our analysis, we present event-based vision for Multi-Agent dynamic Prediction (evMAP), a deep learning architecture designed for real-time, accurate understanding of interaction strength and collective behavior emergence in multi-agent systems.", "authors": ["Minah Lee", "Uday Kamal", "Saibal Mukhopadhyay"], "categories": ["cs.MA", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-11", "url": "https://arxiv.org/abs/2411.07039", "pdf_url": "https://arxiv.org/pdf/2411.07039v1", "arxiv_id": "2411.07039", "doi": "10.48550/arXiv.2411.07039", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Learning for Dynamics & Control", "quality_score": 0.0}
{"id": "7007a5706a66694ce90e8e1e4b3273b1366904a909fa217902ff1d87de256770", "sources": ["arxiv", "semantic_scholar"], "title": "Think Smart, Act SMARL! Analyzing Probabilistic Logic Shields for Multi-Agent Reinforcement Learning", "abstract": "Safe reinforcement learning (RL) is crucial for real-world applications, and multi-agent interactions introduce additional safety challenges. While Probabilistic Logic Shields (PLS) has been a powerful proposal to enforce safety in single-agent RL, their generalizability to multi-agent settings remains unexplored. In this paper, we address this gap by conducting extensive analyses of PLS within decentralized, multi-agent environments, and in doing so, propose $\\textbf{Shielded Multi-Agent Reinforcement Learning (SMARL)}$ as a general framework for steering MARL towards norm-compliant outcomes. Our key contributions are: (1) a novel Probabilistic Logic Temporal Difference (PLTD) update for shielded, independent Q-learning, which incorporates probabilistic constraints directly into the value update process; (2) a probabilistic logic policy gradient method for shielded PPO with formal safety guarantees for MARL; and (3) comprehensive evaluation across symmetric and asymmetrically shielded $n$-player game-theoretic benchmarks, demonstrating fewer constraint violations and significantly better cooperation under normative constraints. These results position SMARL as an effective mechanism for equilibrium selection, paving the way toward safer, socially aligned multi-agent systems.", "authors": ["Satchit Chatterji", "Erman Acar"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-07", "url": "https://arxiv.org/abs/2411.04867", "pdf_url": "https://arxiv.org/pdf/2411.04867v3", "arxiv_id": "2411.04867", "doi": "10.3233/FAIA251103", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.0}
{"id": "ba79b0cde783eed1409417e59db8e59d3e2b27129df13b8836549bf0fcfb2fb1", "sources": ["arxiv", "semantic_scholar"], "title": "Semantic-Aware Resource Management for C-V2X Platooning via Multi-Agent Reinforcement Learning", "abstract": "Semantic communication transmits the extracted features of information rather than raw data, significantly reducing redundancy, which is crucial for addressing spectrum and energy challenges in 6G networks. In this paper, we introduce semantic communication into a cellular vehicle-to-everything (C-V2X)- based autonomous vehicle platoon system for the first time, aiming to achieve efficient management of communication resources in a dynamic environment. Firstly, we construct a mathematical model for semantic communication in platoon systems, in which the DeepSC model and MU-DeepSC model are used to semantically encode and decode unimodal and multi-modal data, respectively. Then, we propose the quality of experience (QoE) metric based on semantic similarity and semantic rate. Meanwhile, we consider the success rate of semantic information transmission (SRS) metric to ensure the fairness of channel resource allocation. Next, the optimization problem is posed with the aim of maximizing the QoE in vehicle-to-vehicle (V2V) links while improving SRS. To solve this mixed integer nonlinear programming problem (MINLP) and adapt to time-varying channel conditions, the paper proposes a distributed semantic-aware multi-modal resource allocation (SAMRA) algorithm based on multi-agent reinforcement learning (MARL), referred to as SAMRAMARL. The algorithm can dynamically allocate channels and power and determine semantic symbol length based on the contextual importance of the transmitted information, ensuring efficient resource utilization. Finally, extensive simulations have demonstrated that SAMRAMARL outperforms existing methods, achieving significant gains in QoE, SRS, and communication delay in C-V2X platooning scenarios.", "authors": ["Wenjun Zhang", "Qiong Wu", "Pingyi Fan", "Kezhi Wang", "Nan Cheng", "Wen Chen", "Khaled B. Letaief"], "categories": ["cs.LG", "cs.MA", "cs.NI", "eess.SP"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2024-11-07", "url": "https://arxiv.org/abs/2411.04672", "pdf_url": "https://arxiv.org/pdf/2411.04672v2", "arxiv_id": "2411.04672", "doi": "10.48550/arXiv.2411.04672", "citation_count": 13, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/qiongwu86/Semantic-Aware-Resource-Management-for-C-V2X-Platooning-via-Multi-Agent-Reinforcement-Learning", "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "7dba94464f9f1d17a1f49143b007f69fb74e48264282f15d34a7379e45cd3395", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agents are Social Groups: Investigating Social Influence of Multiple Agents in Human-Agent Interactions", "abstract": "Multi-agent systems - systems with multiple independent AI agents working together to achieve a common goal - are becoming increasingly prevalent in daily life. Drawing inspiration from the phenomenon of human group social influence, we investigate whether a group of AI agents can create social pressure on users to agree with them, potentially changing their stance on a topic. We conducted a study in which participants discussed social issues with either a single or multiple AI agents, and where the agents either agreed or disagreed with the user's stance on the topic. We found that conversing with multiple agents (holding conversation content constant) increased the social pressure felt by participants, and caused a greater shift in opinion towards the agents' stances on each topic. Our study shows the potential advantages of multi-agent systems over single-agent platforms in causing opinion change. We discuss design implications for possible multi-agent systems that promote social good, as well as the potential for malicious actors to use these systems to manipulate public opinion.", "authors": ["Tianqi Song", "Yugin Tan", "Zicheng Zhu", "Yibin Feng", "Yi-Chieh Lee"], "categories": ["cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-07", "url": "https://arxiv.org/abs/2411.04578", "pdf_url": "https://arxiv.org/pdf/2411.04578v2", "arxiv_id": "2411.04578", "doi": "10.1145/3757633", "citation_count": 23, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3495}
{"id": "3270b77c99b45b42d91f045744663f3c056a969bd442c7909807e4f646114d9f", "sources": ["arxiv", "semantic_scholar"], "title": "CPIG: Leveraging Consistency Policy with Intention Guidance for Multi-agent Exploration", "abstract": "Efficient exploration is crucial in cooperative multi-agent reinforcement learning (MARL), especially in sparse-reward settings. However, due to the reliance on the unimodal policy, existing methods are prone to falling into the local optima, hindering the effective exploration of better policies. Furthermore, in sparse-reward settings, each agent tends to receive a scarce reward, which poses significant challenges to inter-agent cooperation. This not only increases the difficulty of policy learning but also degrades the overall performance of multi-agent tasks. To address these issues, we propose a Consistency Policy with Intention Guidance (CPIG), with two primary components: (a) introducing a multimodal policy to enhance the agent's exploration capability, and (b) sharing the intention among agents to foster agent cooperation. For component (a), CPIG incorporates a Consistency model as the policy, leveraging its multimodal nature and stochastic characteristics to facilitate exploration. Regarding component (b), we introduce an Intention Learner to deduce the intention on the global state from each agent's local observation. This intention then serves as a guidance for the Consistency Policy, promoting cooperation among agents. The proposed method is evaluated in multi-agent particle environments (MPE) and multi-agent MuJoCo (MAMuJoCo). Empirical results demonstrate that our method not only achieves comparable performance to various baselines in dense-reward environments but also significantly enhances performance in sparse-reward settings, outperforming state-of-the-art (SOTA) algorithms by 20%.", "authors": ["Yuqian Fu", "Yuanheng Zhu", "Haoran Li", "Zijie Zhao", "Jiajun Chai", "Dongbin Zhao"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-06", "url": "https://arxiv.org/abs/2411.03603", "pdf_url": "https://arxiv.org/pdf/2411.03603v2", "arxiv_id": "2411.03603", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0}
{"id": "459adb69f895037a32bbac5e9b3d02f47780ed9d92083660579cc0cb0a3bf96a", "sources": ["arxiv", "semantic_scholar"], "title": "SMoA: Improving Multi-agent Large Language Models with Sparse Mixture-of-Agents", "abstract": "While multi-agent systems have been shown to significantly enhance the performance of Large Language Models (LLMs) across various tasks and applications, the dense interaction between scaling agents potentially hampers their efficiency and diversity. To address these challenges, we draw inspiration from the sparse mixture-of-agents (SMoE) and propose a sparse mixture-of-agents (SMoA) framework to improve the efficiency and diversity of multi-agent LLMs. Unlike completely connected structures, SMoA introduces novel Response Selection and Early Stopping mechanisms to sparsify information flows among individual LLM agents, striking a balance between performance and efficiency. Additionally, inspired by the expert diversity principle in SMoE frameworks for workload balance between experts, we assign distinct role descriptions to each LLM agent, fostering diverse and divergent thinking. Extensive experiments on reasoning, alignment, and fairness benchmarks demonstrate that SMoA achieves performance comparable to traditional mixture-of-agents approaches but with significantly lower computational costs. Further analysis reveals that SMoA is more stable, has a greater capacity to scale, and offers considerable potential through hyper-parameter optimization. Code and data will be available at: https://github.com/David-Li0406/SMoA.", "authors": ["Dawei Li", "Zhen Tan", "Peijia Qian", "Yifan Li", "Kumar Satvik Chaudhary", "Lijie Hu", "Jiayi Shen"], "categories": ["cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-05", "url": "https://arxiv.org/abs/2411.03284", "pdf_url": "https://arxiv.org/pdf/2411.03284v1", "arxiv_id": "2411.03284", "doi": "10.48550/arXiv.2411.03284", "citation_count": 31, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/David-Li0406/SMoA", "venue": "Pacific-Asia Conference on Knowledge Discovery and Data Mining", "quality_score": 0.3763}
{"id": "be3c4e2a5a7764642cd8e97d0f7ae657ad2739af46d158b3c371f525e50a0d58", "sources": ["arxiv", "semantic_scholar"], "title": "Spontaneous Emergence of Agent Individuality through Social Interactions in LLM-Based Communities", "abstract": "We study the emergence of agency from scratch by using Large Language Model (LLM)-based agents. In previous studies of LLM-based agents, each agent's characteristics, including personality and memory, have traditionally been predefined. We focused on how individuality, such as behavior, personality, and memory, can be differentiated from an undifferentiated state. The present LLM agents engage in cooperative communication within a group simulation, exchanging context-based messages in natural language. By analyzing this multi-agent simulation, we report valuable new insights into how social norms, cooperation, and personality traits can emerge spontaneously. This paper demonstrates that autonomously interacting LLM-powered agents generate hallucinations and hashtags to sustain communication, which, in turn, increases the diversity of words within their interactions. Each agent's emotions shift through communication, and as they form communities, the personalities of the agents emerge and evolve accordingly. This computational modeling approach and its findings will provide a new method for analyzing collective artificial intelligence.", "authors": ["Ryosuke Takata", "Atsushi Masumori", "Takashi Ikegami"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-05", "url": "https://arxiv.org/abs/2411.03252", "pdf_url": "https://arxiv.org/pdf/2411.03252v1", "arxiv_id": "2411.03252", "doi": "10.48550/arXiv.2411.03252", "citation_count": 11, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2698}
{"id": "ff03891ae9fadbce1e4775bcd9085bfa9ea819c07b725760216e632bb18b03ea", "sources": ["arxiv", "semantic_scholar"], "title": "Guiding Multi-agent Multi-task Reinforcement Learning by a Hierarchical Framework with Logical Reward Shaping", "abstract": "Multi-agent hierarchical reinforcement learning (MAHRL) has been studied as an effective means to solve intelligent decision problems in complex and large-scale environments. However, most current MAHRL algorithms follow the traditional way of using reward functions in reinforcement learning, which limits their use to a single task. This study aims to design a multi-agent cooperative algorithm with logic reward shaping (LRS), which uses a more flexible way of setting the rewards, allowing for the effective completion of multi-tasks. LRS uses Linear Temporal Logic (LTL) to express the internal logic relation of subtasks within a complex task. Then, it evaluates whether the subformulae of the LTL expressions are satisfied based on a designed reward structure. This helps agents to learn to effectively complete tasks by adhering to the LTL expressions, thus enhancing the interpretability and credibility of their decisions. To enhance coordination and cooperation among multiple agents, a value iteration technique is designed to evaluate the actions taken by each agent. Based on this evaluation, a reward function is shaped for coordination, which enables each agent to evaluate its status and complete the remaining subtasks through experiential learning. Experiments have been conducted on various types of tasks in the Minecraft-like environment. The results demonstrate that the proposed algorithm can improve the performance of multi-agents when learning to complete multi-tasks.", "authors": ["Chanjuan Liu", "Jinmiao Cong", "Bingcai Chen", "Yaochu Jin", "Enqiang Zhu"], "categories": ["cs.AI", "cs.LO"], "fields_of_study": ["Medicine", "Computer Science"], "published_date": "2024-11-02", "url": "https://arxiv.org/abs/2411.01184", "pdf_url": "https://arxiv.org/pdf/2411.01184v1", "arxiv_id": "2411.01184", "doi": "10.1109/TCYB.2025.3631239", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Cybernetics", "quality_score": 0.1193}
{"id": "2a8617704cb9dabc40120e467f85146c88724c283f22209e39eda6c3d4e692d4", "sources": ["arxiv", "semantic_scholar"], "title": "DARD: A Multi-Agent Approach for Task-Oriented Dialog Systems", "abstract": "Task-oriented dialogue systems are essential for applications ranging from customer service to personal assistants and are widely used across various industries. However, developing effective multi-domain systems remains a significant challenge due to the complexity of handling diverse user intents, entity types, and domain-specific knowledge across several domains. In this work, we propose DARD (Domain Assigned Response Delegation), a multi-agent conversational system capable of successfully handling multi-domain dialogs. DARD leverages domain-specific agents, orchestrated by a central dialog manager agent. Our extensive experiments compare and utilize various agent modeling approaches, combining the strengths of smaller fine-tuned models (Flan-T5-large & Mistral-7B) with their larger counterparts, Large Language Models (LLMs) (Claude Sonnet 3.0). We provide insights into the strengths and limitations of each approach, highlighting the benefits of our multi-agent framework in terms of flexibility and composability. We evaluate DARD using the well-established MultiWOZ benchmark, achieving state-of-the-art performance by improving the dialogue inform rate by 6.6% and the success rate by 4.1% over the best-performing existing approaches. Additionally, we discuss various annotator discrepancies and issues within the MultiWOZ dataset and its evaluation system.", "authors": ["Aman Gupta", "Anirudh Ravichandran", "Ziji Zhang", "Swair Shah", "Anurag Beniwal", "Narayanan Sadagopan"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-11-01", "url": "https://arxiv.org/abs/2411.00427", "pdf_url": "https://arxiv.org/pdf/2411.00427v1", "arxiv_id": "2411.00427", "doi": "10.48550/arXiv.2411.00427", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "6a803a8e958c28981c4754f2bb8c3f8f54f47125d54e891221671b8721adcaf8", "sources": ["arxiv", "semantic_scholar"], "title": "EMOS: Embodiment-aware Heterogeneous Multi-robot Operating System with LLM Agents", "abstract": "Heterogeneous multi-robot systems (HMRS) have emerged as a powerful approach for tackling complex tasks that single robots cannot manage alone. Current large-language-model-based multi-agent systems (LLM-based MAS) have shown success in areas like software development and operating systems, but applying these systems to robot control presents unique challenges. In particular, the capabilities of each agent in a multi-robot system are inherently tied to the physical composition of the robots, rather than predefined roles. To address this issue, we introduce a novel multi-agent framework designed to enable effective collaboration among heterogeneous robots with varying embodiments and capabilities, along with a new benchmark named Habitat-MAS. One of our key designs is $\\textit{Robot Resume}$: Instead of adopting human-designed role play, we propose a self-prompted approach, where agents comprehend robot URDF files and call robot kinematics tools to generate descriptions of their physics capabilities to guide their behavior in task planning and action execution. The Habitat-MAS benchmark is designed to assess how a multi-agent framework handles tasks that require embodiment-aware reasoning, which includes 1) manipulation, 2) perception, 3) navigation, and 4) comprehensive multi-floor object rearrangement. The experimental results indicate that the robot's resume and the hierarchical design of our multi-agent system are essential for the effective operation of the heterogeneous multi-robot system within this intricate problem context.", "authors": ["Junting Chen", "Checheng Yu", "Xunzhe Zhou", "Tianqi Xu", "Yao Mu", "Mengkang Hu", "Wenqi Shao", "Yikai Wang", "Guohao Li", "Lin Shao"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-30", "url": "https://arxiv.org/abs/2410.22662", "pdf_url": "https://arxiv.org/pdf/2410.22662v2", "arxiv_id": "2410.22662", "doi": "10.48550/arXiv.2410.22662", "citation_count": 25, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.3537}
{"id": "b36d09fd724bf93b781ee1581288666fb9bfc6105ba60a374f6b74df64768303", "sources": ["arxiv", "semantic_scholar"], "title": "ACC-Collab: An Actor-Critic Approach to Multi-Agent LLM Collaboration", "abstract": "Large language models (LLMs) have demonstrated a remarkable ability to serve as general-purpose tools for various language-based tasks. Recent works have demonstrated that the efficacy of such models can be improved through iterative dialog between multiple models. While these paradigms show promise in improving model efficacy, most works in this area treat collaboration as an emergent behavior, rather than a learned behavior. In doing so, current multi-agent frameworks rely on collaborative behaviors to have been sufficiently trained into off-the-shelf models. To address this limitation, we propose ACC-Collab, an Actor-Critic based learning framework to produce a two-agent team (an actor-agent and a critic-agent) specialized in collaboration. We demonstrate that ACC-Collab outperforms SotA multi-agent techniques on a wide array of benchmarks.", "authors": ["Andrew Estornell", "Jean-Francois Ton", "Yuanshun Yao", "Yang Liu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-30", "url": "https://arxiv.org/abs/2411.00053", "pdf_url": "https://arxiv.org/pdf/2411.00053v3", "arxiv_id": "2411.00053", "doi": null, "citation_count": 22, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.3404}
{"id": "16ab61f89e266d1d333be4b409215587bc50ecbb3df848b8c193134e4b0ad94b", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Large Language Models for Conversational Task-Solving", "abstract": "In an era where single large language models have dominated the landscape of artificial intelligence for years, multi-agent systems arise as new protagonists in conversational task-solving. While previous studies have showcased their potential in reasoning tasks and creative endeavors, an analysis of their limitations concerning the conversational paradigms and the impact of individual agents is missing. It remains unascertained how multi-agent discussions perform across tasks of varying complexity and how the structure of these conversations influences the process. To fill that gap, this work systematically evaluates multi-agent systems across various discussion paradigms, assessing their strengths and weaknesses in both generative tasks and question-answering tasks. Alongside the experiments, I propose a taxonomy of 20 multi-agent research studies from 2022 to 2024, followed by the introduction of a framework for deploying multi-agent LLMs in conversational task-solving. I demonstrate that while multi-agent systems excel in complex reasoning tasks, outperforming a single model by leveraging expert personas, they fail on basic tasks. Concretely, I identify three challenges that arise: 1) While longer discussions enhance reasoning, agents fail to maintain conformity to strict task requirements, which leads to problem drift, making shorter conversations more effective for basic tasks. 2) Prolonged discussions risk alignment collapse, raising new safety concerns for these systems. 3) I showcase discussion monopolization through long generations, posing the problem of fairness in decision-making for tasks like summarization. This work uncovers both the potential and challenges that arise with multi-agent interaction and varying conversational paradigms, providing insights into how future research could improve the efficiency, performance, and safety of multi-agent LLMs.", "authors": ["Jonas Becker"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-30", "url": "https://arxiv.org/abs/2410.22932", "pdf_url": "https://arxiv.org/pdf/2410.22932v2", "arxiv_id": "2410.22932", "doi": "10.48550/arXiv.2410.22932", "citation_count": 29, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3693}
{"id": "44384c6798fa523d31ab13800b3966b32f0f576ac38cbaca2805297c5619121a", "sources": ["arxiv", "semantic_scholar"], "title": "Inverse Attention Agents for Multi-Agent Systems", "abstract": "A major challenge for Multi-Agent Systems is enabling agents to adapt dynamically to diverse environments in which opponents and teammates may continually change. Agents trained using conventional methods tend to excel only within the confines of their training cohorts; their performance drops significantly when confronting unfamiliar agents. To address this shortcoming, we introduce Inverse Attention Agents that adopt concepts from the Theory of Mind (ToM) implemented algorithmically using an attention mechanism trained in an end-to-end manner. Crucial to determining the final actions of these agents, the weights in their attention model explicitly represent attention to different goals. We furthermore propose an inverse attention network that deduces the ToM of agents based on observations and prior actions. The network infers the attentional states of other agents, thereby refining the attention weights to adjust the agent's final action. We conduct experiments in a continuous environment, tackling demanding tasks encompassing cooperation, competition, and a blend of both. They demonstrate that the inverse attention network successfully infers the attention of other agents, and that this information improves agent performance. Additional human experiments show that, compared to baseline agent models, our inverse attention agents exhibit superior cooperation with humans and better emulate human behaviors.", "authors": ["Qian Long", "Ruoyan Li", "Minglu Zhao", "Tao Gao", "Demetri Terzopoulos"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-29", "url": "https://arxiv.org/abs/2410.21794", "pdf_url": "https://arxiv.org/pdf/2410.21794v2", "arxiv_id": "2410.21794", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.0753}
{"id": "e63972bdf9e0880b3f23283d33ac0c5b0490435a297cd38d0b7f38d9313389fc", "sources": ["arxiv", "semantic_scholar"], "title": "Synergizing LLM Agents and Knowledge Graph for Socioeconomic Prediction in LBSN", "abstract": "The fast development of location-based social networks (LBSNs) has led to significant changes in society, resulting in popular studies of using LBSN data for socioeconomic prediction, e.g., regional population and commercial activity estimation. Existing studies design various graphs to model heterogeneous LBSN data, and further apply graph representation learning methods for socioeconomic prediction. However, these approaches heavily rely on heuristic ideas and expertise to extract task-relevant knowledge from diverse data, which may not be optimal for specific tasks. Additionally, they tend to overlook the inherent relationships between different indicators, limiting the prediction accuracy. Motivated by the remarkable abilities of large language models (LLMs) in commonsense reasoning, embedding, and multi-agent collaboration, in this work, we synergize LLM agents and knowledge graph for socioeconomic prediction. We first construct a location-based knowledge graph (LBKG) to integrate multi-sourced LBSN data. Then we leverage the reasoning power of LLM agent to identify relevant meta-paths in the LBKG for each type of socioeconomic prediction task, and design a semantic-guided attention module for knowledge fusion with meta-paths. Moreover, we introduce a cross-task communication mechanism to further enhance performance by enabling knowledge sharing across tasks at both LLM agent and KG levels. On the one hand, the LLM agents for different tasks collaborate to generate more diverse and comprehensive meta-paths. On the other hand, the embeddings from different tasks are adaptively merged for better socioeconomic prediction. Experiments on two datasets demonstrate the effectiveness of the synergistic design between LLM and KG, providing insights for information sharing across socioeconomic prediction tasks.", "authors": ["Zhilun Zhou", "Jingyang Fan", "Yu Liu", "Fengli Xu", "Depeng Jin", "Yong Li"], "categories": ["cs.CL", "cs.AI", "cs.LG", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-29", "url": "https://arxiv.org/abs/2411.00028", "pdf_url": "https://arxiv.org/pdf/2411.00028v2", "arxiv_id": "2411.00028", "doi": "10.48550/arXiv.2411.00028", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "8ec95f99de5f547e267754fe2c14bc84d177f7bb173d0af169bcbd5f0d8d2805", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Financial Question Answering with a Multi-Agent Reflection Framework", "abstract": "While Large Language Models (LLMs) have shown impressive capabilities in numerous Natural Language Processing (NLP) tasks, they still struggle with financial question answering (QA), particularly when numerical reasoning is required. Recently, LLM-based multi-agent frameworks have demonstrated remarkable effectiveness in multi-step reasoning, which is crucial for financial QA tasks as it involves extracting relevant information from tables and text and then performing numerical reasoning on the extracted data to infer answers. In this study, we propose a multi-agent framework incorporating a critic agent that reflects on the reasoning steps and final answers for each question. Additionally, we enhance our system by adding multiple critic agents, each focusing on a specific aspect of the answer. Our results indicate that this framework significantly improves performance compared to single-agent reasoning, with an average performance increase of 15% for the LLaMA3-8B model and 5% for the LLaMA3-70B model. Furthermore, our framework performs on par with, and in some cases surpasses, larger single-agent LLMs such as LLaMA3.1-405B and GPT-4o-mini, though it falls slightly short compared to Claude-3.5 Sonnet. Overall, our framework presents an effective solution to enhance open-source LLMs for financial QA tasks, offering a cost-effective alternative to larger models like Claude-3.5 Sonnet.", "authors": ["Sorouralsadat Fatemi", "Yuheng Hu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-29", "url": "https://arxiv.org/abs/2410.21741", "pdf_url": "https://arxiv.org/pdf/2410.21741v1", "arxiv_id": "2410.21741", "doi": "10.1145/3677052.3698686", "citation_count": 23, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": "International Conference on AI in Finance", "quality_score": 0.3451}
{"id": "fb4ba3c38c4cd4248a1128d5b105fa5536f725b2bb6af89bc32975fbdce54685", "sources": ["arxiv", "semantic_scholar"], "title": "MARCO: Multi-Agent Real-time Chat Orchestration", "abstract": "Large language model advancements have enabled the development of multi-agent frameworks to tackle complex, real-world problems such as to automate tasks that require interactions with diverse tools, reasoning, and human collaboration. We present MARCO, a Multi-Agent Real-time Chat Orchestration framework for automating tasks using LLMs. MARCO addresses key challenges in utilizing LLMs for complex, multi-step task execution. It incorporates robust guardrails to steer LLM behavior, validate outputs, and recover from errors that stem from inconsistent output formatting, function and parameter hallucination, and lack of domain knowledge. Through extensive experiments we demonstrate MARCO's superior performance with 94.48% and 92.74% accuracy on task execution for Digital Restaurant Service Platform conversations and Retail conversations datasets respectively along with 44.91% improved latency and 33.71% cost reduction. We also report effects of guardrails in performance gain along with comparisons of various LLM models, both open-source and proprietary. The modular and generic design of MARCO allows it to be adapted for automating tasks across domains and to execute complex usecases through multi-turn interactions.", "authors": ["Anubhav Shrimal", "Stanley Kanagaraj", "Kriti Biswas", "Swarnalatha Raghuraman", "Anish Nediyanchath", "Yi Zhang", "Promod Yenigalla"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-29", "url": "https://arxiv.org/abs/2410.21784", "pdf_url": "https://arxiv.org/pdf/2410.21784v1", "arxiv_id": "2410.21784", "doi": "10.18653/v1/2024.emnlp-industry.102", "citation_count": 5, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.1945}
{"id": "66833e6dbd33651510f752fe5134e094d5d07877dc7eca54d149432290e1551f", "sources": ["arxiv", "semantic_scholar"], "title": "Can We Trust AI Agents? A Case Study of an LLM-Based Multi-Agent System for Ethical AI", "abstract": "AI-based systems, including Large Language Models (LLM), impact millions by supporting diverse tasks but face issues like misinformation, bias, and misuse. AI ethics is crucial as new technologies and concerns emerge, but objective, practical guidance remains debated. This study examines the use of LLMs for AI ethics in practice, assessing how LLM trustworthiness-enhancing techniques affect software development in this context. Using the Design Science Research (DSR) method, we identify techniques for LLM trustworthiness: multi-agents, distinct roles, structured communication, and multiple rounds of debate. We design a multi-agent prototype LLM-MAS, where agents engage in structured discussions on real-world AI ethics issues from the AI Incident Database. We evaluate the prototype across three case scenarios using thematic analysis, hierarchical clustering, comparative (baseline) studies, and running source code. The system generates approximately 2,000 lines of code per case, compared to only 80 lines in baseline trials. Discussions reveal terms like bias detection, transparency, accountability, user consent, GDPR compliance, fairness evaluation, and EU AI Act compliance, showing this prototype ability to generate extensive source code and documentation addressing often overlooked AI ethics issues. However, practical challenges in source code integration and dependency management may limit its use by practitioners.", "authors": ["José Antonio Siqueira de Cerqueira", "Mamia Agbese", "Rebekah Rousi", "Nannan Xi", "Juho Hamari", "Pekka Abrahamsson"], "categories": ["cs.CY", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-25", "url": "https://arxiv.org/abs/2411.08881", "pdf_url": "https://arxiv.org/pdf/2411.08881v2", "arxiv_id": "2411.08881", "doi": null, "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2785}
{"id": "908397863f3838a0467e4e1b80a4051313cda74462402911442690c2b401cca3", "sources": ["arxiv", "semantic_scholar"], "title": "MaCTG: Multi-Agent Collaborative Thought Graph for Automatic Programming", "abstract": "With the rapid advancement of Large Language Models (LLMs), LLM-based approaches have demonstrated strong problem-solving capabilities across various domains. However, in automatic programming, a single LLM is typically limited to function-level code generation, while multi-agent systems composed of multiple LLMs often suffer from inefficient task planning. This lack of structured coordination can lead to cascading hallucinations, where accumulated errors across agents result in suboptimal workflows and excessive computational costs. To overcome these challenges, we introduce MaCTG (Multi-Agent Collaborative Thought Graph), a novel multi-agent framework that employs a dynamic graph structure to facilitate precise task allocation and controlled collaboration among LLM agents. MaCTG autonomously assigns agent roles based on programming requirements, dynamically refines task distribution through context-aware adjustments, and systematically verifies and integrates project-level code, effectively reducing hallucination errors and improving overall accuracy. MaCTG enhances cost-effectiveness by implementing a hybrid LLM deployment, where proprietary models handle complex reasoning, while open-source models are used for routine coding and validation tasks. To evaluate MaCTG's effectiveness, we applied it to traditional image processing auto-programming tasks, achieving a state-of-the-art accuracy of 83.33%. Additionally, by leveraging its hybrid LLM configuration, MaCTG significantly reduced operational costs by 89.09% compared to existing multi-agent frameworks, demonstrating its efficiency, scalability, and real-world applicability.", "authors": ["Zixiao Zhao", "Jing Sun", "Zhe Hou", "Zhiyuan Wei", "Cheng-Hao Cai", "Miao Qiao", "Jin Song Dong"], "categories": ["cs.SE", "cs.CV", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-25", "url": "https://arxiv.org/abs/2410.19245", "pdf_url": "https://arxiv.org/pdf/2410.19245v2", "arxiv_id": "2410.19245", "doi": null, "citation_count": 3, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.1505}
{"id": "0dbf3ef183bbadae5ed0c991ddbb57139e919f863d8e80b1f0aa248030fe95dc", "sources": ["arxiv", "semantic_scholar"], "title": "Offline-to-Online Multi-Agent Reinforcement Learning with Offline Value Function Memory and Sequential Exploration", "abstract": "Offline-to-Online Reinforcement Learning has emerged as a powerful paradigm, leveraging offline data for initialization and online fine-tuning to enhance both sample efficiency and performance. However, most existing research has focused on single-agent settings, with limited exploration of the multi-agent extension, i.e., Offline-to-Online Multi-Agent Reinforcement Learning (O2O MARL). In O2O MARL, two critical challenges become more prominent as the number of agents increases: (i) the risk of unlearning pre-trained Q-values due to distributional shifts during the transition from offline-to-online phases, and (ii) the difficulty of efficient exploration in the large joint state-action space. To tackle these challenges, we propose a novel O2O MARL framework called Offline Value Function Memory with Sequential Exploration (OVMSE). First, we introduce the Offline Value Function Memory (OVM) mechanism to compute target Q-values, preserving knowledge gained during offline training, ensuring smoother transitions, and enabling efficient fine-tuning. Second, we propose a decentralized Sequential Exploration (SE) strategy tailored for O2O MARL, which effectively utilizes the pre-trained offline policy for exploration, thereby significantly reducing the joint state-action space to be explored. Extensive experiments on the StarCraft Multi-Agent Challenge (SMAC) demonstrate that OVMSE significantly outperforms existing baselines, achieving superior sample efficiency and overall performance.", "authors": ["Hai Zhong", "Xun Wang", "Zhuoran Li", "Longbo Huang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-25", "url": "https://arxiv.org/abs/2410.19450", "pdf_url": "https://arxiv.org/pdf/2410.19450v3", "arxiv_id": "2410.19450", "doi": "10.48550/arXiv.2410.19450", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1505}
{"id": "0447979c5eca0c45b0e95c1843746b42be56f719a413ac9dd3c8eb0485deca6e", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Reinforcement Learning with Selective State-Space Models", "abstract": "The Transformer model has demonstrated success across a wide range of domains, including in Multi-Agent Reinforcement Learning (MARL) where the Multi-Agent Transformer (MAT) has emerged as a leading algorithm in the field. However, a significant drawback of Transformer models is their quadratic computational complexity relative to input size, making them computationally expensive when scaling to larger inputs. This limitation restricts MAT's scalability in environments with many agents. Recently, State-Space Models (SSMs) have gained attention due to their computational efficiency, but their application in MARL remains unexplored. In this work, we investigate the use of Mamba, a recent SSM, in MARL and assess whether it can match the performance of MAT while providing significant improvements in efficiency. We introduce a modified version of MAT that incorporates standard and bi-directional Mamba blocks, as well as a novel \"cross-attention\" Mamba block. Extensive testing shows that our Multi-Agent Mamba (MAM) matches the performance of MAT across multiple standard multi-agent environments, while offering superior scalability to larger agent scenarios. This is significant for the MARL community, because it indicates that SSMs could replace Transformers without compromising performance, whilst also supporting more effective scaling to higher numbers of agents. Our project page is available at https://sites.google.com/view/multi-agent-mamba .", "authors": ["Jemma Daniel", "Ruan de Kock", "Louay Ben Nessir", "Sasha Abramowitz", "Omayma Mahjoub", "Wiem Khlifi", "Claude Formanek", "Arnu Pretorius"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-25", "url": "https://arxiv.org/abs/2410.19382", "pdf_url": "https://arxiv.org/pdf/2410.19382v2", "arxiv_id": "2410.19382", "doi": "10.48550/arXiv.2410.19382", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1193}
{"id": "b054fa8e5d623daa2fab393768a0cc4b47216bb28b99bce056bb47a4aa561163", "sources": ["arxiv", "semantic_scholar"], "title": "Benchmarking Floworks against OpenAI & Anthropic: A Novel Framework for Enhanced LLM Function Calling", "abstract": "Large Language Models (LLMs) have shown remarkable capabilities in various domains, yet their economic impact has been limited by challenges in tool use and function calling. This paper introduces ThorV2, a novel architecture that significantly enhances LLMs' function calling abilities. We develop a comprehensive benchmark focused on HubSpot CRM operations to evaluate ThorV2 against leading models from OpenAI and Anthropic. Our results demonstrate that ThorV2 outperforms existing models in accuracy, reliability, latency, and cost efficiency for both single and multi-API calling tasks. We also show that ThorV2 is far more reliable and scales better to multistep tasks compared to traditional models. Our work offers the tantalizing possibility of more accurate function-calling compared to today's best-performing models using significantly smaller LLMs. These advancements have significant implications for the development of more capable AI assistants and the broader application of LLMs in real-world scenarios.", "authors": ["Nirav Bhan", "Shival Gupta", "Sai Manaswini", "Ritik Baba", "Narun Yadav", "Hillori Desai", "Yash Choudhary", "Aman Pawar", "Sarthak Shrivastava", "Sudipta Biswas"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-23", "url": "https://arxiv.org/abs/2410.17950", "pdf_url": "https://arxiv.org/pdf/2410.17950v1", "arxiv_id": "2410.17950", "doi": "10.48550/arXiv.2410.17950", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "0b04f48e49228c7fcca7a3fa4c3318ece4d6f748d7b147ec1cbe3e09d6e5e62e", "sources": ["arxiv", "semantic_scholar"], "title": "Breaking ReAct Agents: Foot-in-the-Door Attack Will Get You In", "abstract": "Following the advancement of large language models (LLMs), the development of LLM-based autonomous agents has become increasingly prevalent. As a result, the need to understand the security vulnerabilities of these agents has become a critical task. We examine how ReAct agents can be exploited using a straightforward yet effective method we refer to as the foot-in-the-door attack. Our experiments show that indirect prompt injection attacks, prompted by harmless and unrelated requests (such as basic calculations) can significantly increase the likelihood of the agent performing subsequent malicious actions. Our results show that once a ReAct agents thought includes a specific tool or action, the likelihood of executing this tool in the subsequent steps increases significantly, as the agent seldom re-evaluates its actions. Consequently, even random, harmless requests can establish a foot-in-the-door, allowing an attacker to embed malicious instructions into the agents thought process, making it more susceptible to harmful directives. To mitigate this vulnerability, we propose implementing a simple reflection mechanism that prompts the agent to reassess the safety of its actions during execution, which can help reduce the success of such attacks.", "authors": ["Itay Nakash", "George Kour", "Guy Uziel", "Ateret Anaby-Tavor"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-22", "url": "https://arxiv.org/abs/2410.16950", "pdf_url": "https://arxiv.org/pdf/2410.16950v1", "arxiv_id": "2410.16950", "doi": "10.48550/arXiv.2410.16950", "citation_count": 25, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.3537}
{"id": "705ff40d93578d66aa4afeb7d67bb8987ef67eac17cd73a7f50862468ebccea6", "sources": ["arxiv", "semantic_scholar"], "title": "Episodic Future Thinking Mechanism for Multi-agent Reinforcement Learning", "abstract": "Understanding cognitive processes in multi-agent interactions is a primary goal in cognitive science. It can guide the direction of artificial intelligence (AI) research toward social decision-making in multi-agent systems, which includes uncertainty from character heterogeneity. In this paper, we introduce an episodic future thinking (EFT) mechanism for a reinforcement learning (RL) agent, inspired by cognitive processes observed in animals. To enable future thinking functionality, we first develop a multi-character policy that captures diverse characters with an ensemble of heterogeneous policies. Here, the character of an agent is defined as a different weight combination on reward components, representing distinct behavioral preferences. The future thinking agent collects observation-action trajectories of the target agents and uses the pre-trained multi-character policy to infer their characters. Once the character is inferred, the agent predicts the upcoming actions of target agents and simulates the potential future scenario. This capability allows the agent to adaptively select the optimal action, considering the predicted future scenario in multi-agent interactions. To evaluate the proposed mechanism, we consider the multi-agent autonomous driving scenario with diverse driving traits and multiple particle environments. Simulation results demonstrate that the EFT mechanism with accurate character inference leads to a higher reward than existing multi-agent solutions. We also confirm that the effect of reward improvement remains valid across societies with different levels of character diversity.", "authors": ["Dongsu Lee", "Minhae Kwon"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-22", "url": "https://arxiv.org/abs/2410.17373", "pdf_url": "https://arxiv.org/pdf/2410.17373v1", "arxiv_id": "2410.17373", "doi": "10.48550/arXiv.2410.17373", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.2113}
{"id": "cd7ba998fd34c9a98a04287d1c9680ce1b3a8e2199f71b44366b9cd3e5f8df5c", "sources": ["arxiv", "semantic_scholar"], "title": "Personalized Recommendation Systems using Multimodal, Autonomous, Multi Agent Systems", "abstract": "This paper describes a highly developed personalised recommendation system using multimodal, autonomous, multi-agent systems. The system focuses on the incorporation of futuristic AI tech and LLMs like Gemini-1.5- pro and LLaMA-70B to improve customer service experiences especially within e-commerce. Our approach uses multi agent, multimodal systems to provide best possible recommendations to its users. The system is made up of three agents as a whole. The first agent recommends products appropriate for answering the given question, while the second asks follow-up questions based on images that belong to these recommended products and is followed up with an autonomous search by the third agent. It also features a real-time data fetch, user preferences-based recommendations and is adaptive learning. During complicated queries the application processes with Symphony, and uses the Groq API to answer quickly with low response times. It uses a multimodal way to utilize text and images comprehensively, so as to optimize product recommendation and customer interaction.", "authors": ["Param Thakkar", "Anushka Yadav"], "categories": ["cs.IR", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-22", "url": "https://arxiv.org/abs/2410.19855", "pdf_url": "https://arxiv.org/pdf/2410.19855v1", "arxiv_id": "2410.19855", "doi": "10.48550/arXiv.2410.19855", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "d1bc3ad0d9b382b15c056f0850c5fcc123b41aaf13671cec0ec3e36fbdf39369", "sources": ["arxiv", "semantic_scholar"], "title": "VipAct: Visual-Perception Enhancement via Specialized VLM Agent Collaboration and Tool-use", "abstract": "While vision-language models (VLMs) have demonstrated remarkable performance across various tasks combining textual and visual information, they continue to struggle with fine-grained visual perception tasks that require detailed pixel-level analysis. Effectively eliciting comprehensive reasoning from VLMs on such intricate visual elements remains an open challenge. In this paper, we present VipAct, an agent framework that enhances VLMs by integrating multi-agent collaboration and vision expert models, enabling more precise visual understanding and comprehensive reasoning. VipAct consists of an orchestrator agent, which manages task requirement analysis, planning, and coordination, along with specialized agents that handle specific tasks such as image captioning and vision expert models that provide high-precision perceptual information. This multi-agent approach allows VLMs to better perform fine-grained visual perception tasks by synergizing planning, reasoning, and tool use. We evaluate VipAct on benchmarks featuring a diverse set of visual perception tasks, with experimental results demonstrating significant performance improvements over state-of-the-art baselines across all tasks. Furthermore, comprehensive ablation studies reveal the critical role of multi-agent collaboration in eliciting more detailed System-2 reasoning and highlight the importance of image input for task planning. Additionally, our error analysis identifies patterns of VLMs' inherent limitations in visual perception, providing insights into potential future improvements. VipAct offers a flexible and extensible framework, paving the way for more advanced visual perception systems across various real-world applications.", "authors": ["Zhehao Zhang", "Ryan Rossi", "Tong Yu", "Franck Dernoncourt", "Ruiyi Zhang", "Jiuxiang Gu", "Sungchul Kim", "Xiang Chen", "Zichao Wang", "Nedim Lipka"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-21", "url": "https://arxiv.org/abs/2410.16400", "pdf_url": "https://arxiv.org/pdf/2410.16400v2", "arxiv_id": "2410.16400", "doi": "10.48550/arXiv.2410.16400", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.2865}
{"id": "c36aec50531385c76025d07de9916c06918490aed4bc6c283591f75aebf5f5f8", "sources": ["arxiv", "semantic_scholar"], "title": "NetSafe: Exploring the Topological Safety of Multi-agent Networks", "abstract": "Large language models (LLMs) have empowered nodes within multi-agent networks with intelligence, showing growing applications in both academia and industry. However, how to prevent these networks from generating malicious information remains unexplored with previous research on single LLM's safety be challenging to transfer. In this paper, we focus on the safety of multi-agent networks from a topological perspective, investigating which topological properties contribute to safer networks. To this end, we propose a general framework, NetSafe along with an iterative RelCom interaction to unify existing diverse LLM-based agent frameworks, laying the foundation for generalized topological safety research. We identify several critical phenomena when multi-agent networks are exposed to attacks involving misinformation, bias, and harmful information, termed as Agent Hallucination and Aggregation Safety. Furthermore, we find that highly connected networks are more susceptible to the spread of adversarial attacks, with task performance in a Star Graph Topology decreasing by 29.7%. Besides, our proposed static metrics aligned more closely with real-world dynamic evaluations than traditional graph-theoretic metrics, indicating that networks with greater average distances from attackers exhibit enhanced safety. In conclusion, our work introduces a new topological perspective on the safety of LLM-based multi-agent networks and discovers several unreported phenomena, paving the way for future research to explore the safety of such networks.", "authors": ["Miao Yu", "Shilong Wang", "Guibin Zhang", "Junyuan Mao", "Chenlong Yin", "Qijiong Liu", "Qingsong Wen", "Kun Wang", "Yang Wang"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-21", "url": "https://arxiv.org/abs/2410.15686", "pdf_url": "https://arxiv.org/pdf/2410.15686v1", "arxiv_id": "2410.15686", "doi": "10.48550/arXiv.2410.15686", "citation_count": 32, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3796}
{"id": "9798446a7322218964e947393f1633bb51580bc44df5e66328f28aeb7cff65ee", "sources": ["arxiv", "semantic_scholar"], "title": "Imprompter: Tricking LLM Agents into Improper Tool Use", "abstract": "Large Language Model (LLM) Agents are an emerging computing paradigm that blends generative machine learning with tools such as code interpreters, web browsing, email, and more generally, external resources. These agent-based systems represent an emerging shift in personal computing. We contribute to the security foundations of agent-based systems and surface a new class of automatically computed obfuscated adversarial prompt attacks that violate the confidentiality and integrity of user resources connected to an LLM agent. We show how prompt optimization techniques can find such prompts automatically given the weights of a model. We demonstrate that such attacks transfer to production-level agents. For example, we show an information exfiltration attack on Mistral's LeChat agent that analyzes a user's conversation, picks out personally identifiable information, and formats it into a valid markdown command that results in leaking that data to the attacker's server. This attack shows a nearly 80% success rate in an end-to-end evaluation. We conduct a range of experiments to characterize the efficacy of these attacks and find that they reliably work on emerging agent-based systems like Mistral's LeChat, ChatGLM, and Meta's Llama. These attacks are multimodal, and we show variants in the text-only and image domains.", "authors": ["Xiaohan Fu", "Shuheng Li", "Zihan Wang", "Yihao Liu", "Rajesh K. Gupta", "Taylor Berg-Kirkpatrick", "Earlence Fernandes"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-19", "url": "https://arxiv.org/abs/2410.14923", "pdf_url": "https://arxiv.org/pdf/2410.14923v2", "arxiv_id": "2410.14923", "doi": "10.48550/arXiv.2410.14923", "citation_count": 73, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/Reapor-Yurnero/imprompter", "venue": "arXiv.org", "quality_score": 0.4673}
{"id": "3a55cd19e6830bed351f170f95b7d9fb390716c1ad74bcfda729f41133b7e572", "sources": ["arxiv", "semantic_scholar"], "title": "Toolshed: Scale Tool-Equipped Agents with Advanced RAG-Tool Fusion and Tool Knowledge Bases", "abstract": "Recent advancements in tool-equipped Agents (LLMs) have enabled complex tasks like secure database interactions and multi-agent code development. However, scaling tool capacity beyond agent reasoning or model limits remains a challenge. In this paper, we address these challenges by introducing Toolshed Knowledge Bases, a tool knowledge base (vector database) designed to store enhanced tool representations and optimize tool selection for large-scale tool-equipped Agents. Additionally, we propose Advanced RAG-Tool Fusion, a novel ensemble of tool-applied advanced retrieval-augmented generation (RAG) techniques across the pre-retrieval, intra-retrieval, and post-retrieval phases, without requiring model fine-tuning. During pre-retrieval, tool documents are enhanced with key information and stored in the Toolshed Knowledge Base. Intra-retrieval focuses on query planning and transformation to increase retrieval accuracy. Post-retrieval refines the retrieved tool documents and enables self-reflection. Furthermore, by varying both the total number of tools (tool-M) an Agent has access to and the tool selection threshold (top-k), we address trade-offs between retrieval accuracy, agent performance, and token cost. Our approach achieves 46%, 56%, and 47% absolute improvements on the ToolE single-tool, ToolE multi-tool and Seal-Tools benchmark datasets, respectively (Recall@5).", "authors": ["Elias Lumer", "Vamse Kumar Subbiah", "James A. Burke", "Pradeep Honaganahalli Basavaraju", "Austin Huber"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-18", "url": "https://arxiv.org/abs/2410.14594", "pdf_url": "https://arxiv.org/pdf/2410.14594v2", "arxiv_id": "2410.14594", "doi": "10.48550/arXiv.2410.14594", "citation_count": 24, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Agents and Artificial Intelligence", "quality_score": 0.3495}
{"id": "699306bbcd7827388fc834ac38a8b0c89aba2432336ac1a71562bb841f58c6fe", "sources": ["arxiv", "semantic_scholar"], "title": "MeNTi: Bridging Medical Calculator and LLM Agent with Nested Tool Calling", "abstract": "Integrating tools into Large Language Models (LLMs) has facilitated the widespread application. Despite this, in specialized downstream task contexts, reliance solely on tools is insufficient to fully address the complexities of the real world. This particularly restricts the effective deployment of LLMs in fields such as medicine. In this paper, we focus on the downstream tasks of medical calculators, which use standardized tests to assess an individual's health status. We introduce MeNTi, a universal agent architecture for LLMs. MeNTi integrates a specialized medical toolkit and employs meta-tool and nested calling mechanisms to enhance LLM tool utilization. Specifically, it achieves flexible tool selection and nested tool calling to address practical issues faced in intricate medical scenarios, including calculator selection, slot filling, and unit conversion. To assess the capabilities of LLMs for quantitative assessment throughout the clinical process of calculator scenarios, we introduce CalcQA. This benchmark requires LLMs to use medical calculators to perform calculations and assess patient health status. CalcQA is constructed by professional physicians and includes 100 case-calculator pairs, complemented by a toolkit of 281 medical tools. The experimental results demonstrate significant performance improvements with our framework. This research paves new directions for applying LLMs in demanding scenarios of medicine.", "authors": ["Yakun Zhu", "Shaohang Wei", "Xu Wang", "Kui Xue", "Xiaofan Zhang", "Shaoting Zhang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-17", "url": "https://arxiv.org/abs/2410.13610", "pdf_url": "https://arxiv.org/pdf/2410.13610v3", "arxiv_id": "2410.13610", "doi": "10.48550/arXiv.2410.13610", "citation_count": 13, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/shzyk/MENTI", "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.2865}
{"id": "ac98ab7c852ba749c1ff95ca8f13e9124548d7dc93e63a9ada6b17c7aaa57f95", "sources": ["arxiv", "semantic_scholar"], "title": "Rapid and Automated Alloy Design with Graph Neural Network-Powered LLM-Driven Multi-Agent Systems", "abstract": "A multi-agent AI model is used to automate the discovery of new metallic alloys, integrating multimodal data and external knowledge including insights from physics via atomistic simulations. Our multi-agent system features three key components: (a) a suite of LLMs responsible for tasks such as reasoning and planning, (b) a group of AI agents with distinct roles and expertise that dynamically collaborate, and (c) a newly developed graph neural network (GNN) model for rapid retrieval of key physical properties. A set of LLM-driven AI agents collaborate to automate the exploration of the vast design space of MPEAs, guided by predictions from the GNN. We focus on the NbMoTa family of body-centered cubic (bcc) alloys, modeled using an ML-based interatomic potential, and target two key properties: the Peierls barrier and solute/screw dislocation interaction energy. Our GNN model accurately predicts these atomic-scale properties, providing a faster alternative to costly brute-force calculations and reducing the computational burden on multi-agent systems for physics retrieval. This AI system revolutionizes materials discovery by reducing reliance on human expertise and overcoming the limitations of direct all-atom simulations. By synergizing the predictive power of GNNs with the dynamic collaboration of LLM-based agents, the system autonomously navigates vast alloy design spaces, identifying trends in atomic-scale material properties and predicting macro-scale mechanical strength, as demonstrated by several computational experiments. This approach accelerates the discovery of advanced alloys and holds promise for broader applications in other complex systems, marking a significant step forward in automated materials design.", "authors": ["Alireza Ghafarollahi", "Markus J. Buehler"], "categories": ["cond-mat.mtrl-sci", "cond-mat.dis-nn", "cond-mat.mes-hall", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2024-10-17", "url": "https://arxiv.org/abs/2410.13768", "pdf_url": "https://arxiv.org/pdf/2410.13768v1", "arxiv_id": "2410.13768", "doi": "10.48550/arXiv.2410.13768", "citation_count": 9, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "cdd8792eba71418921ce4742ada764cdd350d1d413d724f829df9686154d80e6", "sources": ["arxiv", "semantic_scholar"], "title": "AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents", "abstract": "Autonomy via agents using large language models (LLMs) for personalized, standardized tasks boosts human efficiency. Automating web tasks (like booking hotels within a budget) is increasingly sought after. Fulfilling practical needs, the web agent also serves as an important proof-of-concept example for various agent grounding scenarios, with its success promising advancements in many future applications. Prior research often handcrafts web agent strategies (e.g., prompting templates, multi-agent systems, search methods, etc.) and the corresponding in-context examples, which may not generalize well across all real-world scenarios. On the other hand, there has been limited study on the misalignment between a web agent's observation/action representation and the pre-training data of the LLM it's based on. This discrepancy is especially notable when LLMs are primarily trained for language completion rather than tasks involving embodied navigation actions and symbolic web elements. Our study enhances an LLM-based web agent by simply refining its observation and action space to better align with the LLM's capabilities. This approach enables our base agent to significantly outperform previous methods on a wide variety of web tasks. Specifically, on WebArena, a benchmark featuring general-purpose web interaction tasks, our agent AgentOccam surpasses the previous state-of-the-art and concurrent work by 9.8 (+29.4%) and 5.9 (+15.8%) absolute points respectively, and boosts the success rate by 26.6 points (+161%) over similar plain web agents with its observation and action space alignment. We achieve this without using in-context examples, new agent roles, online feedback or search strategies. AgentOccam's simple design highlights LLMs' impressive zero-shot performance on web tasks, and underlines the critical role of carefully tuning observation and action spaces for LLM-based agents.", "authors": ["Ke Yang", "Yao Liu", "Sapana Chaudhary", "Rasool Fakoor", "Pratik Chaudhari", "George Karypis", "Huzefa Rangwala"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-17", "url": "https://arxiv.org/abs/2410.13825", "pdf_url": "https://arxiv.org/pdf/2410.13825v2", "arxiv_id": "2410.13825", "doi": "10.48550/arXiv.2410.13825", "citation_count": 104, "influential_citation_count": 17, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.6276}
{"id": "a0dee6cfbee0b411bdee0881428cd0059b2a77d7f806400d7af0cec17cc59023", "sources": ["arxiv", "semantic_scholar"], "title": "MedAide: Information Fusion and Anatomy of Medical Intents via LLM-based Agent Collaboration", "abstract": "In healthcare intelligence, the ability to fuse heterogeneous, multi-intent information from diverse clinical sources is fundamental to building reliable decision-making systems. Large Language Model (LLM)-driven information interaction systems currently showing potential promise in the healthcare domain. Nevertheless, they often suffer from information redundancy and coupling when dealing with complex medical intents, leading to severe hallucinations and performance bottlenecks. To this end, we propose MedAide, an LLM-based medical multi-agent collaboration framework designed to enable intent-aware information fusion and coordinated reasoning across specialized healthcare domains. Specifically, we introduce a regularization-guided module that combines syntactic constraints with retrieval augmented generation to decompose complex queries into structured representations, facilitating fine-grained clinical information fusion and intent resolution. Additionally, a dynamic intent prototype matching module is proposed to utilize dynamic prototype representation with a semantic similarity matching mechanism to achieve adaptive recognition and updating of the agent's intent in multi-round healthcare dialogues. Ultimately, we design a rotation agent collaboration mechanism that introduces dynamic role rotation and decision-level information fusion across specialized medical agents. Extensive experiments are conducted on four medical benchmarks with composite intents. Experimental results from automated metrics and expert doctor evaluations show that MedAide outperforms current LLMs and improves their medical proficiency and strategic reasoning.", "authors": ["Dingkang Yang", "Jinjie Wei", "Mingcheng Li", "Jiyao Liu", "Lihao Liu", "Ming Hu", "Junjun He", "Yakun Ju", "Wei Zhou", "Yang Liu", "Lihua Zhang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-16", "url": "https://arxiv.org/abs/2410.12532", "pdf_url": "https://arxiv.org/pdf/2410.12532v3", "arxiv_id": "2410.12532", "doi": "10.1016/j.inffus.2025.103743", "citation_count": 22, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Information Fusion", "quality_score": 0.3404}
{"id": "d9fe9fad1052180ea1e836394c381302a23c53e94581904dc65325c40a3364c6", "sources": ["arxiv", "semantic_scholar"], "title": "Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning", "abstract": "Large Language Models (LLMs) have exhibited significant potential in performing diverse tasks, including the ability to call functions or use external tools to enhance their performance. While current research on function calling by LLMs primarily focuses on single-turn interactions, this paper addresses the overlooked necessity for LLMs to engage in multi-turn function calling--critical for handling compositional, real-world queries that require planning with functions but not only use functions. To facilitate this, we introduce an approach, BUTTON, which generates synthetic compositional instruction tuning data via bottom-up instruction construction and top-down trajectory generation. In the bottom-up phase, we generate simple atomic tasks based on real-world scenarios and build compositional tasks using heuristic strategies based on atomic tasks. Corresponding function definitions are then synthesized for these compositional tasks. The top-down phase features a multi-agent environment where interactions among simulated humans, assistants, and tools are utilized to gather multi-turn function calling trajectories. This approach ensures task compositionality and allows for effective function and trajectory generation by examining atomic tasks within compositional tasks. We produce a dataset BUTTONInstruct comprising 8k data points and demonstrate its effectiveness through extensive experiments across various LLMs.", "authors": ["Mingyang Chen", "Haoze Sun", "Tianpeng Li", "Fan Yang", "Hao Liang", "Keer Lu", "Bin Cui", "Wentao Zhang", "Zenan Zhou", "Weipeng Chen"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-16", "url": "https://arxiv.org/abs/2410.12952", "pdf_url": "https://arxiv.org/pdf/2410.12952v2", "arxiv_id": "2410.12952", "doi": "10.48550/arXiv.2410.12952", "citation_count": 37, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.3949}
{"id": "2270fde3d92d7400e2481c2fab959337d4894f8b6f134c0c4be2ed25283cacea", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Corridor Generating Algorithm", "abstract": "In this paper, we propose the Multi-Agent Corridor Generating Algorithm (MACGA) for solving the Multi-agent Pathfinding (MAPF) problem, where a group of agents need to find non-colliding paths to their target locations. Existing approaches struggle to solve dense MAPF instances. In MACGA, the agents build \\emph{corridors}, which are sequences of connected vertices, from current locations towards agents' goals, and evacuate other agents out of the corridors to avoid collisions and deadlocks. We also present the MACGA+PIBT algorithm, which integrates the well-known rule-based PIBT algorithm into MACGA to improve runtime and solution quality. The proposed algorithms run in polynomial time and have a reachability property, i.e., every agent is guaranteed to reach its goal location at some point. We demonstrate experimentally that MACGA and MACGA+PIBT outperform baseline algorithms in terms of success rate, runtime, and makespan across diverse MAPF benchmark grids.", "authors": ["Arseniy Pertzovsky", "Roni Stern", "Roie Zivan", "Ariel Felner"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-16", "url": "https://arxiv.org/abs/2410.12397", "pdf_url": "https://arxiv.org/pdf/2410.12397v2", "arxiv_id": "2410.12397", "doi": "10.24963/ijcai.2025/28", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.1193}
{"id": "23240656dbda30451466860b1370519b099445038fc619cd154583762883ffac", "sources": ["arxiv", "semantic_scholar"], "title": "Aegis:An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering", "abstract": "Functional safety is a critical aspect of automotive engineering, encompassing all phases of a vehicle's lifecycle, including design, development, production, operation, and decommissioning. This domain involves highly knowledge-intensive tasks. This paper introduces Aegis: An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering. Aegis is specifically designed to support complex functional safety tasks within the automotive sector. It is tailored to perform Hazard Analysis and Risk Assessment(HARA), document Functional Safety Requirements(FSR), and plan test cases for Automatic Emergency Braking(AEB) systems. The most advanced version, Aegis-Max, leverages Retrieval-Augmented Generation(RAG) and reflective mechanisms to enhance its capability in managing complex, knowledge-intensive tasks. Additionally, targeted prompt refinement by professional functional safety practitioners can significantly optimize Aegis's performance in the functional safety domain. This paper demonstrates the potential of Aegis to improve the efficiency and effectiveness of functional safety processes in automotive engineering.", "authors": ["Lu Shi", "Bin Qi", "Jiarui Luo", "Yang Zhang", "Zhanzhao Liang", "Zhaowei Gao", "Wenke Deng", "Lin Sun"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-16", "url": "https://arxiv.org/abs/2410.12475", "pdf_url": "https://arxiv.org/pdf/2410.12475v2", "arxiv_id": "2410.12475", "doi": "10.48550/arXiv.2410.12475", "citation_count": 14, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.294}
{"id": "9c66df0ff101961bbce1a79c5de138c9ca89064060004c9f2ad17befed8a19b5", "sources": ["arxiv", "semantic_scholar"], "title": "Using Protected Attributes to Consider Fairness in Multi-Agent Systems", "abstract": "Fairness in Multi-Agent Systems (MAS) has been extensively studied, particularly in reward distribution among agents in scenarios such as goods allocation, resource division, lotteries, and bargaining systems. Fairness in MAS depends on various factors, including the system's governing rules, the behaviour of the agents, and their characteristics. Yet, fairness in human society often involves evaluating disparities between disadvantaged and privileged groups, guided by principles of Equality, Diversity, and Inclusion (EDI). Taking inspiration from the work on algorithmic fairness, which addresses bias in machine learning-based decision-making, we define protected attributes for MAS as characteristics that should not disadvantage an agent in terms of its expected rewards. We adapt fairness metrics from the algorithmic fairness literature -- namely, demographic parity, counterfactual fairness, and conditional statistical parity -- to the multi-agent setting, where self-interested agents interact within an environment. These metrics allow us to evaluate the fairness of MAS, with the ultimate aim of designing MAS that do not disadvantage agents based on protected attributes.", "authors": ["Gabriele La Malfa", "Jie M. Zhang", "Michael Luck", "Elizabeth Black"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-16", "url": "https://arxiv.org/abs/2410.12889", "pdf_url": "https://arxiv.org/pdf/2410.12889v1", "arxiv_id": "2410.12889", "doi": "10.48550/arXiv.2410.12889", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0753}
{"id": "6ceded0c01e13f9b017c1c99da18f1760c77596a93baee1c05a04f2d6a377b58", "sources": ["arxiv", "semantic_scholar"], "title": "Compositional Shielding and Reinforcement Learning for Multi-Agent Systems", "abstract": "Deep reinforcement learning has emerged as a powerful tool for obtaining high-performance policies. However, the safety of these policies has been a long-standing issue. One promising paradigm to guarantee safety is a shield, which shields a policy from making unsafe actions. However, computing a shield scales exponentially in the number of state variables. This is a particular concern in multi-agent systems with many agents. In this work, we propose a novel approach for multi-agent shielding. We address scalability by computing individual shields for each agent. The challenge is that typical safety specifications are global properties, but the shields of individual agents only ensure local properties. Our key to overcome this challenge is to apply assume-guarantee reasoning. Specifically, we present a sound proof rule that decomposes a (global, complex) safety specification into (local, simple) obligations for the shields of the individual agents. Moreover, we show that applying the shields during reinforcement learning significantly improves the quality of the policies obtained for a given training budget. We demonstrate the effectiveness and scalability of our multi-agent shielding framework in two case studies, reducing the computation time from hours to seconds and achieving fast learning convergence.", "authors": ["Asger Horn Brorholt", "Kim Guldstrand Larsen", "Christian Schilling"], "categories": ["cs.LO", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-14", "url": "https://arxiv.org/abs/2410.10460", "pdf_url": "https://arxiv.org/pdf/2410.10460v2", "arxiv_id": "2410.10460", "doi": "10.48550/arXiv.2410.10460", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1747}
{"id": "5176195815a68747fe9381f198a6796243347984ce576ae03c61dd6014e77788", "sources": ["arxiv", "semantic_scholar"], "title": "CAMPHOR: Collaborative Agents for Multi-input Planning and High-Order Reasoning On Device", "abstract": "While server-side Large Language Models (LLMs) demonstrate proficiency in function calling and complex reasoning, deploying Small Language Models (SLMs) directly on devices brings opportunities to improve latency and privacy but also introduces unique challenges for accuracy and memory. We introduce CAMPHOR, an innovative on-device SLM multi-agent framework designed to handle multiple user inputs and reason over personal context locally, ensuring privacy is maintained. CAMPHOR employs a hierarchical architecture where a high-order reasoning agent decomposes complex tasks and coordinates expert agents responsible for personal context retrieval, tool interaction, and dynamic plan generation. By implementing parameter sharing across agents and leveraging prompt compression, we significantly reduce model size, latency, and memory usage. To validate our approach, we present a novel dataset capturing multi-agent task trajectories centered on personalized mobile assistant use-cases. Our experiments reveal that fine-tuned SLM agents not only surpass closed-source LLMs in task completion F1 by~35\\% but also eliminate the need for server-device communication, all while enhancing privacy.", "authors": ["Yicheng Fu", "Raviteja Anantha", "Jianpeng Cheng"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-12", "url": "https://arxiv.org/abs/2410.09407", "pdf_url": "https://arxiv.org/pdf/2410.09407v1", "arxiv_id": "2410.09407", "doi": "10.48550/arXiv.2410.09407", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2258}
{"id": "9a9fae426aa8d2331ff51499f30deeff7529b1fd2933dff06331538a3d5a045b", "sources": ["arxiv", "semantic_scholar"], "title": "AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents", "abstract": "The robustness of LLMs to jailbreak attacks, where users design prompts to circumvent safety measures and misuse model capabilities, has been studied primarily for LLMs acting as simple chatbots. Meanwhile, LLM agents -- which use external tools and can execute multi-stage tasks -- may pose a greater risk if misused, but their robustness remains underexplored. To facilitate research on LLM agent misuse, we propose a new benchmark called AgentHarm. The benchmark includes a diverse set of 110 explicitly malicious agent tasks (440 with augmentations), covering 11 harm categories including fraud, cybercrime, and harassment. In addition to measuring whether models refuse harmful agentic requests, scoring well on AgentHarm requires jailbroken agents to maintain their capabilities following an attack to complete a multi-step task. We evaluate a range of leading LLMs, and find (1) leading LLMs are surprisingly compliant with malicious agent requests without jailbreaking, (2) simple universal jailbreak templates can be adapted to effectively jailbreak agents, and (3) these jailbreaks enable coherent and malicious multi-step agent behavior and retain model capabilities. To enable simple and reliable evaluation of attacks and defenses for LLM-based agents, we publicly release AgentHarm at https://huggingface.co/datasets/ai-safety-institute/AgentHarm.", "authors": ["Maksym Andriushchenko", "Alexandra Souly", "Mateusz Dziemian", "Derek Duenas", "Maxwell Lin", "Justin Wang", "Dan Hendrycks", "Andy Zou", "Zico Kolter", "Matt Fredrikson", "Eric Winsor", "Jerome Wynne", "Yarin Gal", "Xander Davies"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-11", "url": "https://arxiv.org/abs/2410.09024", "pdf_url": "https://arxiv.org/pdf/2410.09024v3", "arxiv_id": "2410.09024", "doi": "10.48550/arXiv.2410.09024", "citation_count": 277, "influential_citation_count": 23, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.6901}
{"id": "9844b23e05badcfea6fbfa0ce603354922186bab97d03dd3aa45d853265cf7c3", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Actor-Critics in Autonomous Cyber Defense", "abstract": "The need for autonomous and adaptive defense mechanisms has become paramount in the rapidly evolving landscape of cyber threats. Multi-Agent Deep Reinforcement Learning (MADRL) presents a promising approach to enhancing the efficacy and resilience of autonomous cyber operations. This paper explores the application of Multi-Agent Actor-Critic algorithms which provides a general form in Multi-Agent learning to cyber defense, leveraging the collaborative interactions among multiple agents to detect, mitigate, and respond to cyber threats. We demonstrate each agent is able to learn quickly and counter act on the threats autonomously using MADRL in simulated cyber-attack scenarios. The results indicate that MADRL can significantly enhance the capability of autonomous cyber defense systems, paving the way for more intelligent cybersecurity strategies. This study contributes to the growing body of knowledge on leveraging artificial intelligence for cybersecurity and sheds light for future research and development in autonomous cyber operations.", "authors": ["Mingjun Wang", "Remington Dechene"], "categories": ["cs.CR", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-11", "url": "https://arxiv.org/abs/2410.09134", "pdf_url": "https://arxiv.org/pdf/2410.09134v2", "arxiv_id": "2410.09134", "doi": "10.48550/arXiv.2410.09134", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "d34b9ea231dc91b4cb20a8ae9d1e9afc98ec61f1f79f966cf6ec169ab8fdb541", "sources": ["arxiv", "semantic_scholar"], "title": "Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System", "abstract": "Large Language Model (LLM) based multi-agent systems (MAS) show remarkable potential in collaborative problem-solving, yet they still face critical challenges: low communication efficiency, poor scalability, and a lack of effective parameter-updating optimization methods. We present Optima, a novel framework that addresses these issues by significantly enhancing both communication efficiency and task effectiveness in LLM-based MAS through LLM training. Optima employs an iterative generate, rank, select, and train paradigm with a reward function balancing task performance, token efficiency, and communication readability. We explore various RL algorithms, including Supervised Fine-Tuning, Direct Preference Optimization, and their hybrid approaches, providing insights into their effectiveness-efficiency trade-offs. We integrate Monte Carlo Tree Search-inspired techniques for DPO data generation, treating conversation turns as tree nodes to explore diverse interaction paths. Evaluated on common multi-agent tasks, including information-asymmetric question answering and complex reasoning, Optima shows consistent and substantial improvements over single-agent baselines and vanilla MAS based on Llama 3 8B, achieving up to 2.8x performance gain with less than 10\\% tokens on tasks requiring heavy information exchange. Moreover, Optima's efficiency gains open new possibilities for leveraging inference-compute more effectively, leading to improved inference-time scaling laws. By addressing fundamental challenges in LLM-based MAS, Optima shows the potential towards scalable, efficient, and effective MAS (https://chenweize1998.github.io/optima-project-page).", "authors": ["Weize Chen", "Jiarui Yuan", "Chen Qian", "Cheng Yang", "Zhiyuan Liu", "Maosong Sun"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-10", "url": "https://arxiv.org/abs/2410.08115", "pdf_url": "https://arxiv.org/pdf/2410.08115v2", "arxiv_id": "2410.08115", "doi": "10.48550/arXiv.2410.08115", "citation_count": 43, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4109}
{"id": "bbd790e0fd5ef0b9153e25f79893c6ab0d5615d4f7acbc7ff7e48948191031e3", "sources": ["arxiv", "semantic_scholar"], "title": "Prompt Infection: LLM-to-LLM Prompt Injection within Multi-Agent Systems", "abstract": "As Large Language Models (LLMs) grow increasingly powerful, multi-agent systems are becoming more prevalent in modern AI applications. Most safety research, however, has focused on vulnerabilities in single-agent LLMs. These include prompt injection attacks, where malicious prompts embedded in external content trick the LLM into executing unintended or harmful actions, compromising the victim's application. In this paper, we reveal a more dangerous vector: LLM-to-LLM prompt injection within multi-agent systems. We introduce Prompt Infection, a novel attack where malicious prompts self-replicate across interconnected agents, behaving much like a computer virus. This attack poses severe threats, including data theft, scams, misinformation, and system-wide disruption, all while propagating silently through the system. Our extensive experiments demonstrate that multi-agent systems are highly susceptible, even when agents do not publicly share all communications. To address this, we propose LLM Tagging, a defense mechanism that, when combined with existing safeguards, significantly mitigates infection spread. This work underscores the urgent need for advanced security measures as multi-agent LLM systems become more widely adopted.", "authors": ["Donghyun Lee", "Mo Tiwari"], "categories": ["cs.MA", "cs.AI", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-09", "url": "https://arxiv.org/abs/2410.07283", "pdf_url": "https://arxiv.org/pdf/2410.07283v1", "arxiv_id": "2410.07283", "doi": "10.48550/arXiv.2410.07283", "citation_count": 106, "influential_citation_count": 8, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5073}
{"id": "b7ce2470b40fe5ae588743580eb19d1856764749dff93b04ccda7598ae0c44f4", "sources": ["arxiv", "semantic_scholar"], "title": "I Want to Break Free! Persuasion and Anti-Social Behavior of LLMs in Multi-Agent Settings with Social Hierarchy", "abstract": "As LLM-based agents become increasingly autonomous and will more freely interact with each other, studying the interplay among them becomes crucial to anticipate emergent phenomena and potential risks. In this work, we provide an in-depth analysis of the interactions among agents within a simulated hierarchical social environment, drawing inspiration from the Stanford Prison Experiment. Leveraging 2,400 conversations across six LLMs (i.e., LLama3, Orca2, Command-r, Mixtral, Mistral2, and gpt4.1) and 240 experimental scenarios, we analyze persuasion and anti-social behavior between a guard and a prisoner agent with differing objectives. We first document model-specific conversational failures in this multi-agent power dynamic context, thereby narrowing our analytic sample to 1,600 conversations. Among models demonstrating successful interaction, we find that goal setting significantly influences persuasiveness but not anti-social behavior. Moreover, agent personas, especially the guard's, substantially impact both successful persuasion by the prisoner and the manifestation of anti-social actions. Notably, we observe the emergence of anti-social conduct even in absence of explicit negative personality prompts. These results have important implications for the development of interactive LLM agents and the ongoing discussion of their societal impact.", "authors": ["Gian Maria Campedelli", "Nicolò Penzo", "Massimo Stefan", "Roberto Dessì", "Marco Guerini", "Bruno Lepri", "Jacopo Staiano"], "categories": ["cs.CL", "cs.AI", "cs.CY", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-09", "url": "https://arxiv.org/abs/2410.07109", "pdf_url": "https://arxiv.org/pdf/2410.07109v3", "arxiv_id": "2410.07109", "doi": "10.48550/arXiv.2410.07109", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2603}
{"id": "4c58bd1f04be23fd6cd2a717f83ee0287693f7afac3323327ac153c06af66522", "sources": ["arxiv", "semantic_scholar"], "title": "Coevolving with the Other You: Fine-Tuning LLM with Sequential Cooperative Multi-Agent Reinforcement Learning", "abstract": "Reinforcement learning (RL) has emerged as a pivotal technique for fine-tuning large language models (LLMs) on specific tasks. However, prevailing RL fine-tuning methods predominantly rely on PPO and its variants. Though these algorithms are effective in general RL settings, they often exhibit suboptimal performance and vulnerability to distribution collapse when applied to the fine-tuning of LLMs. In this paper, we propose CORY, extending the RL fine-tuning of LLMs to a sequential cooperative multi-agent reinforcement learning framework, to leverage the inherent coevolution and emergent capabilities of multi-agent systems. In CORY, the LLM to be fine-tuned is initially duplicated into two autonomous agents: a pioneer and an observer. The pioneer generates responses based on queries, while the observer generates responses using both the queries and the pioneer's responses. The two agents are trained together. During training, the agents exchange roles periodically, fostering cooperation and coevolution between them. Experiments evaluate CORY's performance by fine-tuning GPT-2 and Llama-2 under subjective and objective reward functions on the IMDB Review and GSM8K datasets, respectively. Results show that CORY outperforms PPO in terms of policy optimality, resistance to distribution collapse, and training robustness, thereby underscoring its potential as a superior methodology for refining LLMs in real-world applications.", "authors": ["Hao Ma", "Tianyi Hu", "Zhiqiang Pu", "Boyin Liu", "Xiaolin Ai", "Yanyan Liang", "Min Chen"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-08", "url": "https://arxiv.org/abs/2410.06101", "pdf_url": "https://arxiv.org/pdf/2410.06101v2", "arxiv_id": "2410.06101", "doi": "10.48550/arXiv.2410.06101", "citation_count": 42, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.4084}
{"id": "d210a57b656689515cdda0c93f84fc630a2e8b36469fab7ae2abc85acec36549", "sources": ["arxiv", "semantic_scholar"], "title": "Scalable and Accurate Graph Reasoning with LLM-based Multi-Agents", "abstract": "Recent research has explored the use of Large Language Models (LLMs) for tackling complex graph reasoning tasks. However, due to the intricacies of graph structures and the inherent limitations of LLMs in handling long text, current approaches often fail to deliver satisfactory accuracy, even on small-scale graphs and simple tasks. To address these challenges, we introduce GraphAgent-Reasoner, a fine-tuning-free framework that utilizes a multi-agent collaboration strategy for explicit and precise graph reasoning. Inspired by distributed graph computation theory, our framework decomposes graph problems into smaller, node-centric tasks that are distributed among multiple agents. The agents collaborate to solve the overall problem, significantly reducing the amount of information and complexity handled by a single LLM, thus enhancing the accuracy of graph reasoning. By simply increasing the number of agents, GraphAgent-Reasoner can efficiently scale to accommodate larger graphs with over 1,000 nodes. Evaluated on the GraphInstruct dataset, our framework demonstrates near-perfect accuracy on polynomial-time graph reasoning tasks, significantly outperforming the best available models, both closed-source and fine-tuned open-source variants. Our framework also demonstrates the capability to handle real-world graph reasoning applications such as webpage importance analysis.", "authors": ["Yuwei Hu", "Runlin Lei", "Xinyi Huang", "Zhewei Wei", "Yongchao Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-07", "url": "https://arxiv.org/abs/2410.05130", "pdf_url": "https://arxiv.org/pdf/2410.05130v3", "arxiv_id": "2410.05130", "doi": "10.48550/arXiv.2410.05130", "citation_count": 19, "influential_citation_count": 3, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "a39ca02398b64209256c063ee71f4ef25f7fdc6183eaf66d1ef0ca8208278fc5", "sources": ["arxiv", "semantic_scholar"], "title": "YOLO-MARL: You Only LLM Once for Multi-Agent Reinforcement Learning", "abstract": "Advancements in deep multi-agent reinforcement learning (MARL) have positioned it as a promising approach for decision-making in cooperative games. However, it still remains challenging for MARL agents to learn cooperative strategies for some game environments. Recently, large language models (LLMs) have demonstrated emergent reasoning capabilities, making them promising candidates for enhancing coordination among the agents. However, due to the model size of LLMs, it can be expensive to frequently infer LLMs for actions that agents can take. In this work, we propose You Only LLM Once for MARL (YOLO-MARL), a novel framework that leverages the high-level task planning capabilities of LLMs to improve the policy learning process of multi-agents in cooperative games. Notably, for each game environment, YOLO-MARL only requires one time interaction with LLMs in the proposed strategy generation, state interpretation and planning function generation modules, before the MARL policy training process. This avoids the ongoing costs and computational time associated with frequent LLMs API calls during training. Moreover, trained decentralized policies based on normal-sized neural networks operate independently of the LLM. We evaluate our method across two different environments and demonstrate that YOLO-MARL outperforms traditional MARL algorithms.", "authors": ["Yuan Zhuang", "Yi Shen", "Zhili Zhang", "Yuxiao Chen", "Fei Miao"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-05", "url": "https://arxiv.org/abs/2410.03997", "pdf_url": "https://arxiv.org/pdf/2410.03997v2", "arxiv_id": "2410.03997", "doi": "10.1109/IROS60139.2025.11245945", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.2603}
{"id": "3c9d92cf9cec5971a171c1e92ef7380140c7edf4d03d8367bdcec474896e0dca", "sources": ["arxiv", "semantic_scholar"], "title": "AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML", "abstract": "Automated machine learning (AutoML) accelerates AI development by automating tasks in the development pipeline, such as optimal model search and hyperparameter tuning. Existing AutoML systems often require technical expertise to set up complex tools, which is in general time-consuming and requires a large amount of human effort. Therefore, recent works have started exploiting large language models (LLM) to lessen such burden and increase the usability of AutoML frameworks via a natural language interface, allowing non-expert users to build their data-driven solutions. These methods, however, are usually designed only for a particular process in the AI development pipeline and do not efficiently use the inherent capacity of the LLMs. This paper proposes AutoML-Agent, a novel multi-agent framework tailored for full-pipeline AutoML, i.e., from data retrieval to model deployment. AutoML-Agent takes user's task descriptions, facilitates collaboration between specialized LLM agents, and delivers deployment-ready models. Unlike existing work, instead of devising a single plan, we introduce a retrieval-augmented planning strategy to enhance exploration to search for more optimal plans. We also decompose each plan into sub-tasks (e.g., data preprocessing and neural network design) each of which is solved by a specialized agent we build via prompting executing in parallel, making the search process more efficient. Moreover, we propose a multi-stage verification to verify executed results and guide the code generation LLM in implementing successful solutions. Extensive experiments on seven downstream tasks using fourteen datasets show that AutoML-Agent achieves a higher success rate in automating the full AutoML process, yielding systems with good performance throughout the diverse domains.", "authors": ["Patara Trirat", "Wonyong Jeong", "Sung Ju Hwang"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02958", "pdf_url": "https://arxiv.org/pdf/2410.02958v2", "arxiv_id": "2410.02958", "doi": "10.48550/arXiv.2410.02958", "citation_count": 82, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.4798}
{"id": "a2841ff70d5a1c855b40d2ef43e2947ce98e80123aa453ef2ace4a9ea1146de4", "sources": ["arxiv", "semantic_scholar"], "title": "Agent-Oriented Planning in Multi-Agent Systems", "abstract": "Through the collaboration of multiple LLM-empowered agents possessing diverse expertise and tools, multi-agent systems achieve impressive progress in solving real-world problems. Given the user queries, the meta-agents, serving as the brain within multi-agent systems, are required to decompose the queries into multiple sub-tasks that can be allocated to suitable agents capable of solving them, so-called agent-oriented planning. In this study, we identify three critical design principles of agent-oriented planning, including solvability, completeness, and non-redundancy, to ensure that each sub-task can be effectively resolved, resulting in satisfactory responses to user queries. These principles further inspire us to propose AOP, a novel framework for agent-oriented planning in multi-agent systems, leveraging a fast task decomposition and allocation process followed by an effective and efficient evaluation via a reward model. According to the evaluation results, the meta-agent is also responsible for promptly making necessary adjustments to sub-tasks and scheduling. Besides, we integrate a feedback loop into AOP to further enhance the effectiveness and robustness of such a problem-solving process. Extensive experiments demonstrate the advancement of AOP in solving real-world problems compared to both single-agent systems and existing planning strategies for multi-agent systems. The source code is available at https://github.com/lalaliat/Agent-Oriented-Planning", "authors": ["Ao Li", "Yuexiang Xie", "Songze Li", "Fugee Tsung", "Bolin Ding", "Yaliang Li"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02189", "pdf_url": "https://arxiv.org/pdf/2410.02189v2", "arxiv_id": "2410.02189", "doi": "10.48550/arXiv.2410.02189", "citation_count": 41, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/lalaliat/Agent-Oriented-Planning", "venue": "International Conference on Learning Representations", "quality_score": 0.4058}
{"id": "825c7ddc0d4c77d3fe65449349cc4cbab6bf189e0d39403959ec1092a355f417", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Implicit Bias Detection and Mitigation in Multi-Agent LLM Interactions", "abstract": "As Large Language Models (LLMs) continue to evolve, they are increasingly being employed in numerous studies to simulate societies and execute diverse social tasks. However, LLMs are susceptible to societal biases due to their exposure to human-generated data. Given that LLMs are being used to gain insights into various societal aspects, it is essential to mitigate these biases. To that end, our study investigates the presence of implicit gender biases in multi-agent LLM interactions and proposes two strategies to mitigate these biases. We begin by creating a dataset of scenarios where implicit gender biases might arise, and subsequently develop a metric to assess the presence of biases. Our empirical analysis reveals that LLMs generate outputs characterized by strong implicit bias associations (>= 50\\% of the time). Furthermore, these biases tend to escalate following multi-agent interactions. To mitigate them, we propose two strategies: self-reflection with in-context examples (ICE); and supervised fine-tuning. Our research demonstrates that both methods effectively mitigate implicit biases, with the ensemble of fine-tuning and self-reflection proving to be the most successful.", "authors": ["Angana Borah", "Rada Mihalcea"], "categories": ["cs.CL", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02584", "pdf_url": "https://arxiv.org/pdf/2410.02584v1", "arxiv_id": "2410.02584", "doi": "10.48550/arXiv.2410.02584", "citation_count": 55, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.437}
{"id": "17270d1417b4576c91ac6c5c43219842db0df3d2d38b9c35319ade2fb44ce84f", "sources": ["arxiv", "semantic_scholar"], "title": "Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems", "abstract": "Recent advancements in large language model (LLM)-powered agents have shown that collective intelligence can significantly outperform individual capabilities, largely attributed to the meticulously designed inter-agent communication topologies. Though impressive in performance, existing multi-agent pipelines inherently introduce substantial token overhead, as well as increased economic costs, which pose challenges for their large-scale deployments. In response to this challenge, we propose an economical, simple, and robust multi-agent communication framework, termed $\\texttt{AgentPrune}$, which can seamlessly integrate into mainstream multi-agent systems and prunes redundant or even malicious communication messages. Technically, $\\texttt{AgentPrune}$ is the first to identify and formally define the \\textit{communication redundancy} issue present in current LLM-based multi-agent pipelines, and efficiently performs one-shot pruning on the spatial-temporal message-passing graph, yielding a token-economic and high-performing communication topology. Extensive experiments across six benchmarks demonstrate that $\\texttt{AgentPrune}$ \\textbf{(I)} achieves comparable results as state-of-the-art topologies at merely $\\$5.6$ cost compared to their $\\$43.7$, \\textbf{(II)} integrates seamlessly into existing multi-agent frameworks with $28.1\\%\\sim72.8\\%\\downarrow$ token reduction, and \\textbf{(III)} successfully defend against two types of agent-based adversarial attacks with $3.5\\%\\sim10.8\\%\\uparrow$ performance boost.", "authors": ["Guibin Zhang", "Yanwei Yue", "Zhixun Li", "Sukwon Yun", "Guancheng Wan", "Kun Wang", "Dawei Cheng", "Jeffrey Xu Yu", "Tianlong Chen"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02506", "pdf_url": "https://arxiv.org/pdf/2410.02506v1", "arxiv_id": "2410.02506", "doi": "10.48550/arXiv.2410.02506", "citation_count": 104, "influential_citation_count": 12, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.557}
{"id": "9f7a9dabecc3bd296e7b9da7fac4656c293306cada7b2d72d4df87c2adbbc6bb", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Emergence of Interaction Patterns across Independent RL Agents in Multi-Agent Environments", "abstract": "Many real-world problems, such as controlling swarms of drones and urban traffic, naturally lend themselves to modeling as multi-agent reinforcement learning (RL) problems. However, existing multi-agent RL methods often suffer from scalability challenges, primarily due to the introduction of communication among agents. Consequently, a key challenge lies in adapting the success of deep learning in single-agent RL to the multi-agent setting. In response to this challenge, we propose an approach that fundamentally reimagines multi-agent environments. Unlike conventional methods that model each agent individually with separate networks, our approach, the Bottom Up Network (BUN), adopts a unique perspective. BUN treats the collective of multi-agents as a unified entity while employing a specialized weight initialization strategy that promotes independent learning. Furthermore, we dynamically establish connections among agents using gradient information, enabling coordination when necessary while maintaining these connections as limited and sparse to effectively manage the computational budget. Our extensive empirical evaluations across a variety of cooperative multi-agent scenarios, including tasks such as cooperative navigation and traffic control, consistently demonstrate BUN's superiority over baseline methods with substantially reduced computational costs.", "authors": ["Vasanth Reddy Baddam", "Suat Gumussoy", "Almuatazbellah Boker", "Hoda Eldardiry"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02516", "pdf_url": "https://arxiv.org/pdf/2410.02516v1", "arxiv_id": "2410.02516", "doi": "10.48550/arXiv.2410.02516", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "aa8ad0049cda16d8bbb4a0ff6f6904c6dd531ce306e4cf089ba9d33d9a9a09d0", "sources": ["arxiv", "semantic_scholar"], "title": "Stop Wandering, Find the Keys: LLMs Discriminate Key States for Efficient Multi-Agent Exploration", "abstract": "With expansive state-action spaces, efficient multi-agent exploration remains a longstanding challenge in reinforcement learning. Although pursuing novelty, diversity, or uncertainty attracts increasing attention, redundant efforts brought by exploration without proper guidance choices poses a practical issue for the community. This paper introduces a systematic approach, termed LEMAE, choosing to channel informative task-relevant guidance from a knowledgeable Large Language Model (LLM) for Efficient Multi-Agent Exploration. Specifically, we ground linguistic knowledge from LLM into symbolic key states, that are critical for task fulfillment, in a discriminative manner at low LLM inference costs. To unleash the power of key states, we design Subspace-based Hindsight Intrinsic Reward (SHIR) to guide agents toward key states by increasing reward density. Additionally, we build the Key State Memory Tree (KSMT) to track transitions between key states in a specific task for organized exploration. Benefiting from diminishing redundant explorations, LEMAE outperforms existing SOTA approaches on the challenging benchmarks (e.g., SMAC and MPE) by a large margin, achieving a 10x acceleration in certain scenarios.", "authors": ["Yun Qu", "Boyuan Wang", "Yuhang Jiang", "Jianzhun Shao", "Yixiu Mao", "Heming Zou", "Chang Liu", "Cheems Wang", "Meiqin Liu", "Xiangyang Ji"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-03", "url": "https://arxiv.org/abs/2410.02511", "pdf_url": "https://arxiv.org/pdf/2410.02511v2", "arxiv_id": "2410.02511", "doi": "10.1007/s11432-025-4978-2", "citation_count": 13, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "SCIENCE CHINA Information Sciences 2026", "quality_score": 0.2865}
{"id": "9c83d2abcec4ec736a0343d6e7a18019a17793d3b9e54a6b55612552e1750f7f", "sources": ["arxiv", "semantic_scholar"], "title": "RGD: Multi-LLM Based Agent Debugger via Refinement and Generation Guidance", "abstract": "Large Language Models (LLMs) have shown incredible potential in code generation tasks, and recent research in prompt engineering have enhanced LLMs' understanding of textual information. However, ensuring the accuracy of generated code often requires extensive testing and validation by programmers. While LLMs can typically generate code based on task descriptions, their accuracy remains limited, especially for complex tasks that require a deeper understanding of both the problem statement and the code generation process. This limitation is primarily due to the LLMs' need to simultaneously comprehend text and generate syntactically and semantically correct code, without having the capability to automatically refine the code. In real-world software development, programmers rarely produce flawless code in a single attempt based on the task description alone, they rely on iterative feedback and debugging to refine their programs. Inspired by this process, we introduce a novel architecture of LLM-based agents for code generation and automatic debugging: Refinement and Guidance Debugging (RGD). The RGD framework is a multi-LLM-based agent debugger that leverages three distinct LLM agents-Guide Agent, Debug Agent, and Feedback Agent. RGD decomposes the code generation task into multiple steps, ensuring a clearer workflow and enabling iterative code refinement based on self-reflection and feedback. Experimental results demonstrate that RGD exhibits remarkable code generation capabilities, achieving state-of-the-art performance with a 9.8% improvement on the HumanEval dataset and a 16.2% improvement on the MBPP dataset compared to the state-of-the-art approaches and traditional direct prompting approaches. We highlight the effectiveness of the RGD framework in enhancing LLMs' ability to generate and refine code autonomously.", "authors": ["Haolin Jin", "Zechao Sun", "Huaming Chen"], "categories": ["cs.SE", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-02", "url": "https://arxiv.org/abs/2410.01242", "pdf_url": "https://arxiv.org/pdf/2410.01242v2", "arxiv_id": "2410.01242", "doi": "10.1109/ICA63002.2024.00037", "citation_count": 14, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Agents", "quality_score": 0.294}
{"id": "b2202c2ba451f8f205da277da46027f2d2a7aa523b9cdcbcb4e326416033506e", "sources": ["arxiv", "semantic_scholar"], "title": "Zodiac: A Cardiologist-Level LLM Framework for Multi-Agent Diagnostics", "abstract": "Large language models (LLMs) have demonstrated remarkable progress in healthcare. However, a significant gap remains regarding LLMs' professionalism in domain-specific clinical practices, limiting their application in real-world diagnostics. In this work, we introduce ZODIAC, an LLM-powered framework with cardiologist-level professionalism designed to engage LLMs in cardiological diagnostics. ZODIAC assists cardiologists by extracting clinically relevant characteristics from patient data, detecting significant arrhythmias, and generating preliminary reports for the review and refinement by cardiologists. To achieve cardiologist-level professionalism, ZODIAC is built on a multi-agent collaboration framework, enabling the processing of patient data across multiple modalities. Each LLM agent is fine-tuned using real-world patient data adjudicated by cardiologists, reinforcing the model's professionalism. ZODIAC undergoes rigorous clinical validation with independent cardiologists, evaluated across eight metrics that measure clinical effectiveness and address security concerns. Results show that ZODIAC outperforms industry-leading models, including OpenAI's GPT-4o, Meta's Llama-3.1-405B, and Google's Gemini-pro, as well as medical-specialist LLMs like Microsoft's BioGPT. ZODIAC demonstrates the transformative potential of specialized LLMs in healthcare by delivering domain-specific solutions that meet the stringent demands of medical practice. Notably, ZODIAC has been successfully integrated into electrocardiography (ECG) devices, exemplifying the growing trend of embedding LLMs into Software-as-Medical-Device (SaMD).", "authors": ["Yuan Zhou", "Peng Zhang", "Mengya Song", "Alice Zheng", "Yiwen Lu", "Zhiheng Liu", "Yong Chen", "Zhaohan Xi"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-10-02", "url": "https://arxiv.org/abs/2410.02026", "pdf_url": "https://arxiv.org/pdf/2410.02026v1", "arxiv_id": "2410.02026", "doi": "10.48550/arXiv.2410.02026", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "b9efbfaaa1d122a932d42b642739e9c95b17500dd5c7fa57d8a92cf63aacd5eb", "sources": ["arxiv", "semantic_scholar"], "title": "MARLadona -- Towards Cooperative Team Play Using Multi-Agent Reinforcement Learning", "abstract": "Robot soccer, in its full complexity, poses an unsolved research challenge. Current solutions heavily rely on engineered heuristic strategies, which lack robustness and adaptability. Deep reinforcement learning has gained significant traction in various complex robotics tasks such as locomotion, manipulation, and competitive games (e.g., AlphaZero, OpenAI Five), making it a promising solution to the robot soccer problem. This paper introduces MARLadona. A decentralized multi-agent reinforcement learning (MARL) training pipeline capable of producing agents with sophisticated team play behavior, bridging the shortcomings of heuristic methods. Furthermore, we created an open-source multi-agent soccer environment. Utilizing our MARL framework and a modified global entity encoder (GEE) as our core architecture, our approach achieves a 66.8% win rate against HELIOS agent, which employs a state-of-the-art heuristic strategy. In addition, we provided an in-depth analysis of the policy behavior and interpreted the agent's intention using the critic network.", "authors": ["Zichong Li", "Filip Bjelonic", "Victor Klemm", "Marco Hutter"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-30", "url": "https://arxiv.org/abs/2409.20326", "pdf_url": "https://arxiv.org/pdf/2409.20326v3", "arxiv_id": "2409.20326", "doi": "10.1109/ICRA55743.2025.11128256", "citation_count": 12, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "IEEE International Conference on Robotics and Automation", "quality_score": 0.2785}
{"id": "a79a65555d9a554e273d1070ca2d1396109bf8efb9fcae735bd14eea3858cb17", "sources": ["arxiv", "semantic_scholar"], "title": "Safe Decentralized Multi-Agent Control using Black-Box Predictors, Conformal Decision Policies, and Control Barrier Functions", "abstract": "We address the challenge of safe control in decentralized multi-agent robotic settings, where agents use uncertain black-box models to predict other agents' trajectories. We use the recently proposed conformal decision theory to adapt the restrictiveness of control barrier functions-based safety constraints based on observed prediction errors. We use these constraints to synthesize controllers that balance between the objectives of safety and task accomplishment, despite the prediction errors. We provide an upper bound on the average over time of the value of a monotonic function of the difference between the safety constraint based on the predicted trajectories and the constraint based on the ground truth ones. We validate our theory through experimental results showing the performance of our controllers when navigating a robot in the multi-agent scenes in the Stanford Drone Dataset.", "authors": ["Sacha Huriot", "Hussein Sibai"], "categories": ["eess.SY", "cs.MA", "cs.RO"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2024-09-27", "url": "https://arxiv.org/abs/2409.18862", "pdf_url": "https://arxiv.org/pdf/2409.18862v4", "arxiv_id": "2409.18862", "doi": "10.1109/ICRA55743.2025.11128015", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Robotics and Automation", "quality_score": 0.2258}
{"id": "bcee55e9dc2073a55639f054a2a4f6fcae2c04c6b9b77c4651160742f1ac9974", "sources": ["arxiv", "semantic_scholar"], "title": "AmpAgent: An LLM-based Multi-Agent System for Multi-stage Amplifier Schematic Design from Literature for Process and Performance Porting", "abstract": "Multi-stage amplifiers are widely applied in analog circuits. However, their large number of components, complex transfer functions, and intricate pole-zero distributions necessitate extensive manpower for derivation and param sizing to ensure their stability. In order to achieve efficient derivation of the transfer function and simplify the difficulty of circuit design, we propose AmpAgent: a multi-agent system based on large language models (LLMs) for efficiently designing such complex amplifiers from literature with process and performance porting. AmpAgent is composed of three agents: Literature Analysis Agent, Mathematics Reasoning Agent and Device Sizing Agent. They are separately responsible for retrieving key information (e.g. formulas and transfer functions) from the literature, decompose the whole circuit's design problem by deriving the key formulas, and address the decomposed problem iteratively. AmpAgent was employed in the schematic design of seven types of multi-stage amplifiers with different compensation techniques. In terms of design efficiency, AmpAgent has reduced the number of iterations by 1.32$ \\sim $4${\\times}$ and execution time by 1.19$ \\sim $2.99${\\times}$ compared to conventional optimization algorithms, with a success rate increased by 1.03$ \\sim $6.79${\\times}$. In terms of circuit performance, it has improved by 1.63$ \\sim $27.25${\\times}$ compared to the original literature. The findings suggest that LLMs could play a crucial role in the field of complex analog circuit schematic design, as well as process and performance porting.", "authors": ["Chengjie Liu", "Weiyu Chen", "Anlan Peng", "Yuan Du", "Li Du", "Jun Yang"], "categories": ["cs.ET", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2024-09-23", "url": "https://arxiv.org/abs/2409.14739", "pdf_url": "https://arxiv.org/pdf/2409.14739v1", "arxiv_id": "2409.14739", "doi": "10.48550/arXiv.2409.14739", "citation_count": 38, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3978}
{"id": "c619337ad4c4c691dd0592ce816943ad1942907be864286f575c446a630f258e", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Vulcan: An Information-Driven Multi-Agent Path Finding Approach", "abstract": "Scientists often search for phenomena of interest while exploring new environments. Autonomous vehicles are deployed to explore such areas where human-operated vehicles would be costly or dangerous. Online control of autonomous vehicles for information-gathering is called adaptive sampling and can be framed as a POMDP that uses information gain as its principal objective. While prior work focuses largely on single-agent scenarios, this paper confronts challenges unique to multi-agent adaptive sampling, such as avoiding redundant observations, preventing vehicle collision, and facilitating path planning under limited communication. We start with Multi-Agent Path Finding (MAPF) methods, which address collision avoidance by decomposing the MAPF problem into a series of single-agent path planning problems. We then present information-driven MAPF which addresses multi-agent information gain under limited communication. First, we introduce an admissible heuristic that relaxes mutual information gain to an additive function that can be evaluated as a set of independent single agent path planning problems. Second, we extend our approach to a distributed system that is robust to limited communication. When all agents are in range, the group plans jointly to maximize information. When some agents move out of range, communicating subgroups are formed and the subgroups plan independently. Since redundant observations are less likely when vehicles are far apart, this approach only incurs a small loss in information gain, resulting in an approach that gracefully transitions from full to partial communication. We evaluate our method against other adaptive sampling strategies across various scenarios, including real-world robotic applications. Our method was able to locate up to 200% more unique phenomena in certain scenarios, and each agent located its first unique phenomenon faster by up to 50%.", "authors": ["Jake Olkin", "Viraj Parimi", "Brian Williams"], "categories": ["cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-19", "url": "https://arxiv.org/abs/2409.13065", "pdf_url": "https://arxiv.org/pdf/2409.13065v1", "arxiv_id": "2409.13065", "doi": "10.1109/IROS58592.2024.10801571", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.1747}
{"id": "21c58c3873c16902a91fc9085faf72b637d38113fc2e8ccf7b4344f234a1d698", "sources": ["arxiv", "semantic_scholar"], "title": "Strategic Collusion of LLM Agents: Market Division in Multi-Commodity Competitions", "abstract": "Machine-learning technologies are seeing increased deployment in real-world market scenarios. In this work, we explore the strategic behaviors of large language models (LLMs) when deployed as autonomous agents in multi-commodity markets, specifically within Cournot competition frameworks. We examine whether LLMs can independently engage in anti-competitive practices such as collusion or, more specifically, market division. Our findings demonstrate that LLMs can effectively monopolize specific commodities by dynamically adjusting their pricing and resource allocation strategies, thereby maximizing profitability without direct human input or explicit collusion commands. These results pose unique challenges and opportunities for businesses looking to integrate AI into strategic roles and for regulatory bodies tasked with maintaining fair and competitive markets. The study provides a foundation for further exploration into the ramifications of deferring high-stakes decisions to LLM-based agents.", "authors": ["Ryan Y. Lin", "Siddhartha Ojha", "Kevin Cai", "Maxwell F. Chen"], "categories": ["cs.GT", "cs.AI", "cs.CL", "q-fin.CP"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2024-09-19", "url": "https://arxiv.org/abs/2410.00031", "pdf_url": "https://arxiv.org/pdf/2410.00031v2", "arxiv_id": "2410.00031", "doi": "10.48550/arXiv.2410.00031", "citation_count": 19, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "78893c37cedef28899d38c827e543ba6557685a253c0d661bc88e451921f87f2", "sources": ["arxiv", "semantic_scholar"], "title": "Improving LLM Reasoning with Multi-Agent Tree-of-Thought Validator Agent", "abstract": "Multi-agent strategies have emerged as a promising approach to enhance the reasoning abilities of Large Language Models (LLMs) by assigning specialized roles in the problem-solving process. Concurrently, Tree of Thoughts (ToT) methods have shown potential in improving reasoning for complex question-answering tasks by exploring diverse reasoning paths. A critical limitation in multi-agent reasoning is the 'Reasoner' agent's shallow exploration of reasoning paths. While ToT strategies could help mitigate this problem, they may generate flawed reasoning branches, which could harm the trustworthiness of the final answer. To leverage the strengths of both multi-agent reasoning and ToT strategies, we introduce a novel approach combining ToT-based Reasoner agents with a Thought Validator agent. Multiple Reasoner agents operate in parallel, employing ToT to explore diverse reasoning paths. The Thought Validator then scrutinizes these paths, considering a Reasoner's conclusion only if its reasoning is valid. This method enables a more robust voting strategy by discarding faulty reasoning paths, enhancing the system's ability to tackle tasks requiring systematic and trustworthy reasoning. Our method demonstrates superior performance compared to existing techniques when evaluated on the GSM8K dataset, outperforming the standard ToT strategy by an average 5.6% across four LLMs. The code and related content can be found in: https://github.com/SecureAIAutonomyLab/MA-ToT", "authors": ["Fatemeh Haji", "Mazal Bethany", "Maryam Tabar", "Jason Chiang", "Anthony Rios", "Peyman Najafirad"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-17", "url": "https://arxiv.org/abs/2409.11527", "pdf_url": "https://arxiv.org/pdf/2409.11527v2", "arxiv_id": "2409.11527", "doi": "10.48550/arXiv.2409.11527", "citation_count": 14, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/SecureAIAutonomyLab/MA-ToT", "venue": "arXiv.org", "quality_score": 0.294}
{"id": "0f6a3b3370fe1dc47617b75822329668c8450640b7f4597fa4068ad941bba32a", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Agent-UMF: LLM-based Agent Unified Modeling Framework for Seamless Design of Multi Active/Passive Core-Agent Architectures", "abstract": "In an era where vast amounts of data are collected and processed from diverse sources, there is a growing demand for sophisticated AI systems capable of intelligently fusing and analyzing this information. To address these challenges, researchers have turned towards integrating tools into LLM-powered agents to enhance the overall information fusion process. However, the conjunction of these technologies and the proposed enhancements in several state-of-the-art works followed a non-unified software architecture, resulting in a lack of modularity and terminological inconsistencies among researchers. To address these issues, we propose a novel LLM-based Agent Unified Modeling Framework (LLM-Agent-UMF) that establishes a clear foundation for agent development from both functional and software architectural perspectives, developed and evaluated using the Architecture Tradeoff and Risk Analysis Framework (ATRAF). Our framework clearly distinguishes between the different components of an LLM-based agent, setting LLMs and tools apart from a new element, the core-agent, which plays the role of central coordinator. This pivotal entity comprises five modules: planning, memory, profile, action, and security -- the latter often neglected in previous works. By classifying core-agents into passive and active types based on their authoritative natures, we propose various multi-core agent architectures that combine unique characteristics of distinctive agents to tackle complex tasks more efficiently. We evaluate our framework by applying it to thirteen state-of-the-art agents, thereby demonstrating its alignment with their functionalities and clarifying overlooked architectural aspects. Moreover, we thoroughly assess five architecture variants of our framework by designing new agent architectures that combine characteristics of state-of-the-art agents to address specific goals. ...", "authors": ["Amine Ben Hassouna", "Hana Chaari", "Ines Belhaj"], "categories": ["cs.SE", "cs.AI", "cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-17", "url": "https://arxiv.org/abs/2409.11393", "pdf_url": "https://arxiv.org/pdf/2409.11393v3", "arxiv_id": "2409.11393", "doi": "10.1016/j.inffus.2025.103865", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Information Fusion", "quality_score": 0.25}
{"id": "cceac8dd077970061b5e98b7c3e8eef25e39d077a79345f62af2ddbe9a056a43", "sources": ["arxiv", "semantic_scholar"], "title": "AutoSafeCoder: A Multi-Agent Framework for Securing LLM Code Generation through Static Analysis and Fuzz Testing", "abstract": "Recent advancements in automatic code generation using large language models (LLMs) have brought us closer to fully automated secure software development. However, existing approaches often rely on a single agent for code generation, which struggles to produce secure, vulnerability-free code. Traditional program synthesis with LLMs has primarily focused on functional correctness, often neglecting critical dynamic security implications that happen during runtime. To address these challenges, we propose AutoSafeCoder, a multi-agent framework that leverages LLM-driven agents for code generation, vulnerability analysis, and security enhancement through continuous collaboration. The framework consists of three agents: a Coding Agent responsible for code generation, a Static Analyzer Agent identifying vulnerabilities, and a Fuzzing Agent performing dynamic testing using a mutation-based fuzzing approach to detect runtime errors. Our contribution focuses on ensuring the safety of multi-agent code generation by integrating dynamic and static testing in an iterative process during code generation by LLM that improves security. Experiments using the SecurityEval dataset demonstrate a 13% reduction in code vulnerabilities compared to baseline LLMs, with no compromise in functionality.", "authors": ["Ana Nunez", "Nafis Tanveer Islam", "Sumit Kumar Jha", "Peyman Najafirad"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-16", "url": "https://arxiv.org/abs/2409.10737", "pdf_url": "https://arxiv.org/pdf/2409.10737v2", "arxiv_id": "2409.10737", "doi": "10.48550/arXiv.2409.10737", "citation_count": 44, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4133}
{"id": "8633918b3b6f3caacfba93486108e006ed0419c0191ef7d2cacf1f7a15dd424c", "sources": ["arxiv", "semantic_scholar"], "title": "Windows Agent Arena: Evaluating Multi-Modal OS Agents at Scale", "abstract": "Large language models (LLMs) show remarkable potential to act as computer agents, enhancing human productivity and software accessibility in multi-modal tasks that require planning and reasoning. However, measuring agent performance in realistic environments remains a challenge since: (i) most benchmarks are limited to specific modalities or domains (e.g. text-only, web navigation, Q&A, coding) and (ii) full benchmark evaluations are slow (on order of magnitude of days) given the multi-step sequential nature of tasks. To address these challenges, we introduce the Windows Agent Arena: a reproducible, general environment focusing exclusively on the Windows operating system (OS) where agents can operate freely within a real Windows OS and use the same wide range of applications, tools, and web browsers available to human users when solving tasks. We adapt the OSWorld framework (Xie et al., 2024) to create 150+ diverse Windows tasks across representative domains that require agent abilities in planning, screen understanding, and tool usage. Our benchmark is scalable and can be seamlessly parallelized in Azure for a full benchmark evaluation in as little as 20 minutes. To demonstrate Windows Agent Arena's capabilities, we also introduce a new multi-modal agent, Navi. Our agent achieves a success rate of 19.5% in the Windows domain, compared to 74.5% performance of an unassisted human. Navi also demonstrates strong performance on another popular web-based benchmark, Mind2Web. We offer extensive quantitative and qualitative analysis of Navi's performance, and provide insights into the opportunities for future research in agent development and data generation using Windows Agent Arena. Webpage: https://microsoft.github.io/WindowsAgentArena Code: https://github.com/microsoft/WindowsAgentArena", "authors": ["Rogerio Bonatti", "Dan Zhao", "Francesco Bonacci", "Dillon Dupont", "Sara Abdali", "Yinheng Li", "Yadong Lu", "Justin Wagle", "Kazuhito Koishida", "Arthur Bucker", "Lawrence Jang", "Zack Hui"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-12", "url": "https://arxiv.org/abs/2409.08264", "pdf_url": "https://arxiv.org/pdf/2409.08264v2", "arxiv_id": "2409.08264", "doi": "10.48550/arXiv.2409.08264", "citation_count": 158, "influential_citation_count": 22, "has_code": true, "code_url": "https://github.com/microsoft/WindowsAgentArena", "venue": "International Conference on Machine Learning", "quality_score": 0.6809}
{"id": "f1c360852fe5512844736ef1316c14cf6beae71ed7134bfa900ac9e6ed08d90d", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based multi-agent poetry generation in non-cooperative environments", "abstract": "Despite substantial progress of large language models (LLMs) for automatic poetry generation, the generated poetry lacks diversity while the training process differs greatly from human learning. Under the rationale that the learning process of the poetry generation systems should be more human-like and their output more diverse and novel, we introduce a framework based on social learning where we emphasize non-cooperative interactions besides cooperative interactions to encourage diversity. Our experiments are the first attempt at LLM-based multi-agent systems in non-cooperative environments for poetry generation employing both TRAINING-BASED agents (GPT-2) and PROMPTING-BASED agents (GPT-3 and GPT-4). Our evaluation based on 96k generated poems shows that our framework benefits the poetry generation process for TRAINING-BASED agents resulting in 1) a 3.0-3.7 percentage point (pp) increase in diversity and a 5.6-11.3 pp increase in novelty according to distinct and novel n-grams. The generated poetry from TRAINING-BASED agents also exhibits group divergence in terms of lexicons, styles and semantics. PROMPTING-BASED agents in our framework also benefit from non-cooperative environments and a more diverse ensemble of models with non-homogeneous agents has the potential to further enhance diversity, with an increase of 7.0-17.5 pp according to our experiments. However, PROMPTING-BASED agents show a decrease in lexical diversity over time and do not exhibit the group-based divergence intended in the social network. Our paper argues for a paradigm shift in creative tasks such as automatic poetry generation to include social learning processes (via LLM-based agent modeling) similar to human interaction.", "authors": ["Ran Zhang", "Steffen Eger"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-05", "url": "https://arxiv.org/abs/2409.03659", "pdf_url": "https://arxiv.org/pdf/2409.03659v2", "arxiv_id": "2409.03659", "doi": "10.48550/arXiv.2409.03659", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Journal of Language Modelling", "quality_score": 0.3138}
{"id": "8c875e611dbfd31e537dee802f7fe726fdf5b3465d1a25e0da2529dd6616b261", "sources": ["arxiv", "semantic_scholar"], "title": "Emergent Language: A Survey and Taxonomy", "abstract": "The field of emergent language represents a novel area of research within the domain of artificial intelligence, particularly within the context of multi-agent reinforcement learning. Although the concept of studying language emergence is not new, early approaches were primarily concerned with explaining human language formation, with little consideration given to its potential utility for artificial agents. In contrast, studies based on reinforcement learning aim to develop communicative capabilities in agents that are comparable to or even superior to human language. Thus, they extend beyond the learned statistical representations that are common in natural language processing research. This gives rise to a number of fundamental questions, from the prerequisites for language emergence to the criteria for measuring its success. This paper addresses these questions by providing a comprehensive review of 181 scientific publications on emergent language in artificial intelligence. Its objective is to serve as a reference for researchers interested in or proficient in the field. Consequently, the main contributions are the definition and overview of the prevailing terminology, the analysis of existing evaluation methods and metrics, and the description of the identified research gaps.", "authors": ["Jannik Peters", "Constantin Waubert de Puiseau", "Hasan Tercan", "Arya Gopikrishnan", "Gustavo Adolpho Lucas De Carvalho", "Christian Bitter", "Tobias Meisen"], "categories": ["cs.MA", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-04", "url": "https://arxiv.org/abs/2409.02645", "pdf_url": "https://arxiv.org/pdf/2409.02645v2", "arxiv_id": "2409.02645", "doi": "10.1007/s10458-025-09691-y", "citation_count": 19, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.3253}
{"id": "0920e4d5adc17592f6960cf97348cc16125bdfd07d2b567774b5bc9c24d63cf9", "sources": ["arxiv", "semantic_scholar"], "title": "Focus Agent: LLM-Powered Virtual Focus Group", "abstract": "In the domain of Human-Computer Interaction, focus groups represent a widely utilised yet resource-intensive methodology, often demanding the expertise of skilled moderators and meticulous preparatory efforts. This study introduces the ``Focus Agent,'' a Large Language Model (LLM) powered framework that simulates both the focus group (for data collection) and acts as a moderator in a focus group setting with human participants. To assess the data quality derived from the Focus Agent, we ran five focus group sessions with a total of 23 human participants as well as deploying the Focus Agent to simulate these discussions with AI participants. Quantitative analysis indicates that Focus Agent can generate opinions similar to those of human participants. Furthermore, the research exposes some improvements associated with LLMs acting as moderators in focus group discussions that include human participants.", "authors": ["Taiyu Zhang", "Xuesong Zhang", "Robbe Cools", "Adalberto L. Simeone"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-09-03", "url": "https://arxiv.org/abs/2409.01907", "pdf_url": "https://arxiv.org/pdf/2409.01907v1", "arxiv_id": "2409.01907", "doi": "10.1145/3652988.3673918", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Intelligent Virtual Agents", "quality_score": 0.25}
{"id": "5cde13375d7e7ef5a4abf595a8e1c29b1febd91ce691f16e830f08e4f41f44b6", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Multi-agent Multi-machine Tending by Mobile Robots", "abstract": "Robotics can help address the growing worker shortage challenge of the manufacturing industry. As such, machine tending is a task collaborative robots can tackle that can also highly boost productivity. Nevertheless, existing robotics systems deployed in that sector rely on a fixed single-arm setup, whereas mobile robots can provide more flexibility and scalability. In this work, we introduce a multi-agent multi-machine tending learning framework by mobile robots based on Multi-agent Reinforcement Learning (MARL) techniques with the design of a suitable observation and reward. Moreover, an attention-based encoding mechanism is developed and integrated into Multi-agent Proximal Policy Optimization (MAPPO) algorithm to boost its performance for machine tending scenarios. Our model (AB-MAPPO) outperformed MAPPO in this new challenging scenario in terms of task success, safety, and resources utilization. Furthermore, we provided an extensive ablation study to support our various design decisions.", "authors": ["Abdalwhab Abdalwhab", "Giovanni Beltrame", "Samira Ebrahimi Kahou", "David St-Onge"], "categories": ["cs.RO", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-29", "url": "https://arxiv.org/abs/2408.16875", "pdf_url": "https://arxiv.org/pdf/2408.16875v3", "arxiv_id": "2408.16875", "doi": "10.48550/arXiv.2408.16875", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "c4601dee697fad2f46d2dbec092c5bf13629bb87749b752645be7e3b6085747a", "sources": ["arxiv", "semantic_scholar"], "title": "BattleAgentBench: A Benchmark for Evaluating Cooperation and Competition Capabilities of Language Models in Multi-Agent Systems", "abstract": "Large Language Models (LLMs) are becoming increasingly powerful and capable of handling complex tasks, e.g., building single agents and multi-agent systems. Compared to single agents, multi-agent systems have higher requirements for the collaboration capabilities of language models. Many benchmarks are proposed to evaluate their collaborative abilities. However, these benchmarks lack fine-grained evaluations of LLM collaborative capabilities. Additionally, multi-agent collaborative and competitive scenarios are ignored in existing works. To address these two problems, we propose a benchmark, called BattleAgentBench, which defines seven sub-stages of three varying difficulty levels and conducts a fine-grained evaluation of language models in terms of single-agent scenario navigation capabilities, paired-agent task execution abilities, and multi-agent collaboration and competition capabilities. We conducted extensive evaluations on leading four closed-source and seven open-source models. Experimental results indicate that API-based models perform excellently on simple tasks but open-source small models struggle with simple tasks. Regarding difficult tasks that require collaborative and competitive abilities, although API-based models have demonstrated some collaborative capabilities, there is still enormous room for improvement.", "authors": ["Wei Wang", "Dan Zhang", "Tao Feng", "Boyan Wang", "Jie Tang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-28", "url": "https://arxiv.org/abs/2408.15971", "pdf_url": "https://arxiv.org/pdf/2408.15971v1", "arxiv_id": "2408.15971", "doi": "10.48550/arXiv.2408.15971", "citation_count": 21, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3356}
{"id": "3be90444957e56b7f1dd11b700dcb0defb87c46e4cadb409f7ce3e591d600d8a", "sources": ["arxiv", "semantic_scholar"], "title": "GenOnet: Generative Open xG Network Simulation with Multi-Agent LLM and ns-3", "abstract": "The move toward Sixth-Generation (6G) networks relies on open interfaces and protocols for seamless interoperability across devices, vendors, and technologies. In this context, open 6G development involves multiple disciplines and requires advanced simulation approaches for testing. In this demo paper, we propose a generative simulation approach based on a multi-agent Large Language Model (LLM) and Network Simulator 3 (ns-3), called Generative Open xG Network Simulation (GenOnet), to effectively generate, debug, execute, and interpret simulated Open Fifth-Generation (5G) environments. The first version of GenOnet application represents a specialized adaptation of the OpenAI GPT models. It incorporates supplementary tools, agents, 5G standards, and seamless integration with ns-3 simulation capabilities, supporting both C++ variants and Python implementations. This release complies with the latest Open Radio Access Network (O-RAN) and 3GPP standards.", "authors": ["Farhad Rezazadeh", "Amir Ashtari Gargari", "Sandra Lagén", "Josep Mangues", "Dusit Niyato", "Lingjia Liu"], "categories": ["cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-25", "url": "https://arxiv.org/abs/2408.13781", "pdf_url": "https://arxiv.org/pdf/2408.13781v2", "arxiv_id": "2408.13781", "doi": "10.1109/6GNet63182.2024.10765766", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2258}
{"id": "cb605b628e020365ef32e4f29b0c801fa7f1a2cf4ed3100aca34b33a9c8025cc", "sources": ["arxiv", "semantic_scholar"], "title": "Collaboration Dynamics and Reliability Challenges of Multi-Agent LLM Systems in Finite Element Analysis", "abstract": "Large Language Model (LLM)-based multi-agent systems are increasingly applied to automate computational workflows in science and engineering. However, how inter-agent dynamics influence reasoning quality and verification reliability remains unclear. We study these mechanisms using an AutoGen-based multi-agent framework for linear-elastic Finite Element Analysis (FEA), evaluating seven role configurations across four tasks under a fixed 12-turn conversation limit. From 1,120 controlled trials, we find that collaboration effectiveness depends more on functional complementarity than team size: the three-agent Coder-Executor-Critic configuration uniquely produced physically and visually correct solutions, while adding redundant reviewers reduced success rates. Yet three systematic failure modes persist: (1) affirmation bias, where the Rebuttal agent endorsed rather than challenged outputs (85-92% agreement, including errors); (2) premature consensus caused by redundant reviewers; and (3) a verification-validation gap where executable but physically incorrect code passed undetected. No agent combination successfully validated constitutive relations in complex tasks. Building on theories of functional diversity, role differentiation, and computational validation, we propose actionable design principles: (i) assign complementary agent roles, (ii) enforce multi-level validation (execution, specification, physics), and (iii) prevent early consensus through adversarial or trigger-based interaction control. These findings establish a principled foundation for designing trustworthy LLM collaborations in engineering workflows.", "authors": ["Chuan Tian", "Yilei Zhang"], "categories": ["cs.AI", "cs.CE", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-23", "url": "https://arxiv.org/abs/2408.13406", "pdf_url": "https://arxiv.org/pdf/2408.13406v2", "arxiv_id": "2408.13406", "doi": null, "citation_count": 8, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "bc828a6c26e0bf2626249f5ec3ed63b844cf13693053da6a4b76dff77282548f", "sources": ["arxiv", "semantic_scholar"], "title": "DrugAgent: Multi-Agent Large Language Model-Based Reasoning for Drug-Target Interaction Prediction", "abstract": "Advancements in large language models (LLMs) allow them to address diverse questions using human-like interfaces. Still, limitations in their training prevent them from answering accurately in scenarios that could benefit from multiple perspectives. Multi-agent systems allow the resolution of questions to enhance result consistency and reliability. While drug-target interaction (DTI) prediction is important for drug discovery, existing approaches face challenges due to complex biological systems and the lack of interpretability needed for clinical applications. DrugAgent is a multi-agent LLM system for DTI prediction that combines multiple specialized perspectives with transparent reasoning. Our system adapts and extends existing multi-agent frameworks by (1) applying coordinator-based architecture to the DTI domain, (2) integrating domain-specific data sources, including ML predictions, knowledge graphs, and literature evidence, and (3) incorporating Chain-of-Thought (CoT) and ReAct (Reason+Act) frameworks for transparent DTI reasoning. We conducted comprehensive experiments using a kinase inhibitor dataset, where our multi-agent LLM method outperformed the non-reasoning multi-agent model (GPT-4o mini) by 45% in F1 score (0.514 vs 0.355). Through ablation studies, we demonstrated the contributions of each agent, with the AI agent being the most impactful, followed by the KG agent and search agent. Most importantly, our approach provides detailed, human-interpretable reasoning for each prediction by combining evidence from multiple sources - a critical feature for biomedical applications where understanding the rationale behind predictions is essential for clinical decision-making and regulatory compliance. Code is available at https://anonymous.4open.science/r/DrugAgent-B2EA.", "authors": ["Yoshitaka Inoue", "Tianci Song", "Xinling Wang", "Augustin Luna", "Tianfan Fu"], "categories": ["cs.AI", "cs.CL", "cs.IR", "cs.LG", "q-bio.QM"], "fields_of_study": ["Medicine", "Computer Science", "Biology"], "published_date": "2024-08-23", "url": "https://arxiv.org/abs/2408.13378", "pdf_url": "https://arxiv.org/pdf/2408.13378v4", "arxiv_id": "2408.13378", "doi": null, "citation_count": 13, "influential_citation_count": 3, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "8d2db9610884fe86d2e5c3b25bafb1aec58d189e5a3e279dda0c7f40cc3cc508", "sources": ["arxiv", "semantic_scholar"], "title": "Story3D-Agent: Exploring 3D Storytelling Visualization with Large Language Models", "abstract": "Traditional visual storytelling is complex, requiring specialized knowledge and substantial resources, yet often constrained by human creativity and creation precision. While Large Language Models (LLMs) enhance visual storytelling, current approaches often limit themselves to 2D visuals or oversimplify stories through motion synthesis and behavioral simulation, failing to create comprehensive, multi-dimensional narratives. To this end, we present Story3D-Agent, a pioneering approach that leverages the capabilities of LLMs to transform provided narratives into 3D-rendered visualizations. By integrating procedural modeling, our approach enables precise control over multi-character actions and motions, as well as diverse decorative elements, ensuring the long-range and dynamic 3D representation. Furthermore, our method supports narrative extension through logical reasoning, ensuring that generated content remains consistent with existing conditions. We have thoroughly evaluated our Story3D-Agent to validate its effectiveness, offering a basic framework to advance 3D story representation.", "authors": ["Yuzhou Huang", "Yiran Qin", "Shunlin Lu", "Xintao Wang", "Rui Huang", "Ying Shan", "Ruimao Zhang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-21", "url": "https://arxiv.org/abs/2408.11801", "pdf_url": "https://arxiv.org/pdf/2408.11801v1", "arxiv_id": "2408.11801", "doi": "10.48550/arXiv.2408.11801", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "0f07d2c1fbba807334a59051a3c72b4f45e303ce9e7794b15535a49e833c20b3", "sources": ["arxiv", "semantic_scholar"], "title": "MegaAgent: A Large-Scale Autonomous LLM-based Multi-Agent System Without Predefined SOPs", "abstract": "LLM-based multi-agent systems (MAS) have shown promise in tackling complex tasks. However, existing solutions often suffer from limited agent coordination and heavy reliance on predefined Standard Operating Procedures (SOPs), which demand extensive human input. To address these limitations, we propose MegaAgent, a large-scale autonomous LLM-based multi-agent system. MegaAgent generates agents based on task complexity and enables dynamic task decomposition, parallel execution, efficient communication, and comprehensive system monitoring of agents. In evaluations, MegaAgent demonstrates exceptional performance, successfully developing a Gobang game within 800 seconds and scaling up to 590 agents in a national policy simulation to generate multi-domain policies. It significantly outperforms existing systems, such as MetaGPT, in both task completion efficiency and scalability. By eliminating the need for predefined SOPs, MegaAgent demonstrates exceptional scalability and autonomy, setting a foundation for advancing true autonomy in MAS. Our code is available at https://github.com/Xtra-Computing/MegaAgent .", "authors": ["Qian Wang", "Tianyu Wang", "Zhenheng Tang", "Qinbin Li", "Nuo Chen", "Jingsheng Liang", "Bingsheng He"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-19", "url": "https://arxiv.org/abs/2408.09955", "pdf_url": "https://arxiv.org/pdf/2408.09955v3", "arxiv_id": "2408.09955", "doi": "10.18653/v1/2025.findings-acl.259", "citation_count": 51, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/Xtra-Computing/MegaAgent", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.429}
{"id": "9494a10e284ed333eee6eb2eadb4cf5c431173753c4968a0800db7a0e6e1e6cb", "sources": ["arxiv", "semantic_scholar"], "title": "A semi-centralized multi-agent RL framework for efficient irrigation scheduling", "abstract": "This paper proposes a Semi-Centralized Multi-Agent Reinforcement Learning (SCMARL) approach for irrigation scheduling in spatially variable agricultural fields, where management zones address spatial variability. The SCMARL framework is hierarchical in nature, with a centralized coordinator agent at the top level and decentralized local agents at the second level. The coordinator agent makes daily binary irrigation decisions based on field-wide conditions, which are communicated to the local agents. Local agents determine appropriate irrigation amounts for specific management zones using local conditions. The framework employs state augmentation approach to handle non-stationarity in the local agents' environments. An extensive evaluation on a large-scale field in Lethbridge, Canada, compares the SCMARL approach with a learning-based multi-agent model predictive control scheduling approach, highlighting its enhanced performance, resulting in water conservation and improved Irrigation Water Use Efficiency (IWUE). Notably, the proposed approach achieved a 4.0% savings in irrigation water while enhancing the IWUE by 6.3%.", "authors": ["Bernard T. Agyeman", "Benjamin Decard-Nelson", "Jinfeng Liu", "Sirish L. Shah"], "categories": ["eess.SY", "cs.LG"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2024-08-15", "url": "https://arxiv.org/abs/2408.08442", "pdf_url": "https://arxiv.org/pdf/2408.08442v1", "arxiv_id": "2408.08442", "doi": "10.48550/arXiv.2408.08442", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "f7e2dde37019253b43d0dc4cc54469e3fd85a4a4fdadf60e1edf82da7e132a9d", "sources": ["arxiv", "semantic_scholar"], "title": "Audit-LLM: Multi-Agent Collaboration for Log-based Insider Threat Detection", "abstract": "Log-based insider threat detection (ITD) detects malicious user activities by auditing log entries. Recently, large language models (LLMs) with strong common sense knowledge have emerged in the domain of ITD. Nevertheless, diverse activity types and overlong log files pose a significant challenge for LLMs in directly discerning malicious ones within myriads of normal activities. Furthermore, the faithfulness hallucination issue from LLMs aggravates its application difficulty in ITD, as the generated conclusion may not align with user commands and activity context. In response to these challenges, we introduce Audit-LLM, a multi-agent log-based insider threat detection framework comprising three collaborative agents: (i) the Decomposer agent, breaking down the complex ITD task into manageable sub-tasks using Chain-of-Thought (COT) reasoning;(ii) the Tool Builder agent, creating reusable tools for sub-tasks to overcome context length limitations in LLMs; and (iii) the Executor agent, generating the final detection conclusion by invoking constructed tools. To enhance conclusion accuracy, we propose a pair-wise Evidence-based Multi-agent Debate (EMAD) mechanism, where two independent Executors iteratively refine their conclusions through reasoning exchange to reach a consensus. Comprehensive experiments conducted on three publicly available ITD datasets-CERT r4.2, CERT r5.2, and PicoDomain-demonstrate the superiority of our method over existing baselines and show that the proposed EMAD significantly improves the faithfulness of explanations generated by LLMs.", "authors": ["Chengyu Song", "Linru Ma", "Jianming Zheng", "Jinzhi Liao", "Hongyu Kuang", "Lin Yang"], "categories": ["cs.CR", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-12", "url": "https://arxiv.org/abs/2408.08902", "pdf_url": "https://arxiv.org/pdf/2408.08902v1", "arxiv_id": "2408.08902", "doi": "10.48550/arXiv.2408.08902", "citation_count": 27, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3891}
{"id": "87ce9afb733afec327350da40debb96df54664558e6f4f8a3c5b12b4f9f84fec", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Heterogeneous Multi-Agent Cooperation in Decentralized MARL via GNN-driven Intrinsic Rewards", "abstract": "Multi-agent Reinforcement Learning (MARL) is emerging as a key framework for various sequential decision-making and control tasks. Unlike their single-agent counterparts, multi-agent systems necessitate successful cooperation among the agents. The deployment of these systems in real-world scenarios often requires decentralized training, a diverse set of agents, and learning from infrequent environmental reward signals. These challenges become more pronounced under partial observability and the lack of prior knowledge about agent heterogeneity. While notable studies use intrinsic motivation (IM) to address reward sparsity or cooperation in decentralized settings, those dealing with heterogeneity typically assume centralized training, parameter sharing, and agent indexing. To overcome these limitations, we propose the CoHet algorithm, which utilizes a novel Graph Neural Network (GNN) based intrinsic motivation to facilitate the learning of heterogeneous agent policies in decentralized settings, under the challenges of partial observability and reward sparsity. Evaluation of CoHet in the Multi-agent Particle Environment (MPE) and Vectorized Multi-Agent Simulator (VMAS) benchmarks demonstrates superior performance compared to the state-of-the-art in a range of cooperative multi-agent scenarios. Our research is supplemented by an analysis of the impact of the agent dynamics model on the intrinsic motivation module, insights into the performance of different CoHet variants, and its robustness to an increasing number of heterogeneous agents.", "authors": ["Jahir Sadik Monon", "Deeparghya Dutta Barua", "Md. Mosaddek Khan"], "categories": ["cs.MA", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-12", "url": "https://arxiv.org/abs/2408.06503", "pdf_url": "https://arxiv.org/pdf/2408.06503v4", "arxiv_id": "2408.06503", "doi": null, "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Proceedings of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2025), pages 2681-2683, 2025", "quality_score": 0.1193}
{"id": "4f91ad3546f29159ee6165962978a213a78c2ed76b382d64047afa5f60681ef6", "sources": ["arxiv", "semantic_scholar"], "title": "AutoGen Studio: A No-Code Developer Tool for Building and Debugging Multi-Agent Systems", "abstract": "Multi-agent systems, where multiple agents (generative AI models + tools) collaborate, are emerging as an effective pattern for solving long-running, complex tasks in numerous domains. However, specifying their parameters (such as models, tools, and orchestration mechanisms etc,.) and debugging them remains challenging for most developers. To address this challenge, we present AUTOGEN STUDIO, a no-code developer tool for rapidly prototyping, debugging, and evaluating multi-agent workflows built upon the AUTOGEN framework. AUTOGEN STUDIO offers a web interface and a Python API for representing LLM-enabled agents using a declarative (JSON-based) specification. It provides an intuitive drag-and-drop UI for agent workflow specification, interactive evaluation and debugging of workflows, and a gallery of reusable agent components. We highlight four design principles for no-code multi-agent developer tools and contribute an open-source implementation at https://github.com/microsoft/autogen/tree/main/samples/apps/autogen-studio", "authors": ["Victor Dibia", "Jingya Chen", "Gagan Bansal", "Suff Syed", "Adam Fourney", "Erkang Zhu", "Chi Wang", "Saleema Amershi"], "categories": ["cs.SE", "cs.AI", "cs.CL", "cs.HC", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-09", "url": "https://arxiv.org/abs/2408.15247", "pdf_url": "https://arxiv.org/pdf/2408.15247v1", "arxiv_id": "2408.15247", "doi": "10.48550/arXiv.2408.15247", "citation_count": 34, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/microsoft/autogen/tree/main/samples/apps/autogen-studio", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.386}
{"id": "d165ffa56ebb1cd22c8f5673f539dcec476ef570813c5f5e45b1b2de3c6687eb", "sources": ["arxiv", "semantic_scholar"], "title": "Can LLMs Beat Humans in Debating? A Dynamic Multi-agent Framework for Competitive Debate", "abstract": "Competitive debate is a complex task of computational argumentation. Large Language Models (LLMs) suffer from hallucinations and lack competitiveness in this field. To address these challenges, we introduce Agent for Debate (Agent4Debate), a dynamic multi-agent framework based on LLMs designed to enhance their capabilities in competitive debate. Drawing inspiration from human behavior in debate preparation and execution, Agent4Debate employs a collaborative architecture where four specialized agents, involving Searcher, Analyzer, Writer, and Reviewer, dynamically interact and cooperate. These agents work throughout the debate process, covering multiple stages from initial research and argument formulation to rebuttal and summary. To comprehensively evaluate framework performance, we construct the Competitive Debate Arena, comprising 66 carefully selected Chinese debate motions. We recruit ten experienced human debaters and collect records of 200 debates involving Agent4Debate, baseline models, and humans. The evaluation employs the Debatrix automatic scoring system and professional human reviewers based on the established Debatrix-Elo and Human-Elo ranking. Experimental results indicate that the state-of-the-art Agent4Debate exhibits capabilities comparable to those of humans. Furthermore, ablation studies demonstrate the effectiveness of each component in the agent structure.", "authors": ["Yiqun Zhang", "Xiaocui Yang", "Shi Feng", "Daling Wang", "Yifei Zhang", "Kaisong Song"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-08", "url": "https://arxiv.org/abs/2408.04472", "pdf_url": "https://arxiv.org/pdf/2408.04472v2", "arxiv_id": "2408.04472", "doi": "10.48550/arXiv.2408.04472", "citation_count": 23, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Acoustics, Speech, and Signal Processing", "quality_score": 0.3451}
{"id": "3c4eb9c134284d77e3476c82de6aaa1f2120e8a67280c49a0c8017ff62ae6d1a", "sources": ["arxiv", "semantic_scholar"], "title": "From Data to Story: Towards Automatic Animated Data Video Creation with LLM-based Multi-Agent Systems", "abstract": "Creating data stories from raw data is challenging due to humans' limited attention spans and the need for specialized skills. Recent advancements in large language models (LLMs) offer great opportunities to develop systems with autonomous agents to streamline the data storytelling workflow. Though multi-agent systems have benefits such as fully realizing LLM potentials with decomposed tasks for individual agents, designing such systems also faces challenges in task decomposition, performance optimization for sub-tasks, and workflow design. To better understand these issues, we develop Data Director, an LLM-based multi-agent system designed to automate the creation of animated data videos, a representative genre of data stories. Data Director interprets raw data, breaks down tasks, designs agent roles to make informed decisions automatically, and seamlessly integrates diverse components of data videos. A case study demonstrates Data Director's effectiveness in generating data videos. Throughout development, we have derived lessons learned from addressing challenges, guiding further advancements in autonomous agents for data storytelling. We also shed light on future directions for global optimization, human-in-the-loop design, and the application of advanced multi-modal LLMs.", "authors": ["Leixian Shen", "Haotian Li", "Yun Wang", "Huamin Qu"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-07", "url": "https://arxiv.org/abs/2408.03876", "pdf_url": "https://arxiv.org/pdf/2408.03876v1", "arxiv_id": "2408.03876", "doi": "10.1109/GEN4DS63889.2024.00008", "citation_count": 27, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3618}
{"id": "5cf5aae269a87ac2306c170e15739602cfd10a73c10fd57d08b839f11523a849", "sources": ["arxiv", "semantic_scholar"], "title": "ReDel: A Toolkit for LLM-Powered Recursive Multi-Agent Systems", "abstract": "Recently, there has been increasing interest in using Large Language Models (LLMs) to construct complex multi-agent systems to perform tasks such as compiling literature reviews, drafting consumer reports, and planning vacations. Many tools and libraries exist for helping create such systems, however none support recursive multi-agent systems -- where the models themselves flexibly decide when to delegate tasks and how to organize their delegation structure. In this work, we introduce ReDel: a toolkit for recursive multi-agent systems that supports custom tool-use, delegation schemes, event-based logging, and interactive replay in an easy-to-use web interface. We show that, using ReDel, we are able to easily identify potential areas of improvements through the visualization and debugging tools. Our code, documentation, and PyPI package are open-source and free to use under the MIT license at https://github.com/zhudotexe/redel.", "authors": ["Andrew Zhu", "Liam Dugan", "Chris Callison-Burch"], "categories": ["cs.CL", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-05", "url": "https://arxiv.org/abs/2408.02248", "pdf_url": "https://arxiv.org/pdf/2408.02248v2", "arxiv_id": "2408.02248", "doi": "10.48550/arXiv.2408.02248", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/zhudotexe/redel", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.2113}
{"id": "560ffbf1180a453281d4c7add894b0d61a60ce0f3e68154e0a974bb55b010175", "sources": ["arxiv", "semantic_scholar"], "title": "Evaluating and Enhancing LLMs Agent based on Theory of Mind in Guandan: A Multi-Player Cooperative Game under Imperfect Information", "abstract": "Large language models (LLMs) have shown success in handling simple games with imperfect information and enabling multi-agent coordination, but their ability to facilitate practical collaboration against other agents in complex, imperfect information environments, especially in a non-English environment, still needs to be explored. This study investigates the applicability of knowledge acquired by open-source and API-based LLMs to sophisticated text-based games requiring agent collaboration under imperfect information, comparing their performance to established baselines using other types of agents. We propose a Theory of Mind (ToM) planning technique that allows LLM agents to adapt their strategy against various adversaries using only game rules, current state, and historical context as input. An external tool was incorporated to mitigate the challenge of dynamic and extensive action spaces in this card game. Our results show that although a performance gap exists between current LLMs and state-of-the-art reinforcement learning (RL) models, LLMs demonstrate ToM capabilities in this game setting. It consistently improves their performance against opposing agents, suggesting their ability to understand the actions of allies and adversaries and establish collaboration with allies. To encourage further research and understanding, we have made our codebase openly accessible.", "authors": ["Yauwai Yim", "Chunkit Chan", "Tianyu Shi", "Zheye Deng", "Wei Fan", "Tianshi Zheng", "Yangqiu Song"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-05", "url": "https://arxiv.org/abs/2408.02559", "pdf_url": "https://arxiv.org/pdf/2408.02559v1", "arxiv_id": "2408.02559", "doi": "10.1109/WI-IAT62293.2024.00074", "citation_count": 23, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.3451}
{"id": "97810236d9185414473448edf50e95670146c8de2f74112b3d88bceaefb2c92e", "sources": ["arxiv", "semantic_scholar"], "title": "The Drama Machine: Simulating Character Development with LLM Agents", "abstract": "This paper explores use of multiple large language model (LLM) agents to simulate complex, dynamic characters in dramatic scenarios. We introduce a drama machine framework that coordinates interactions between LLM agents playing different 'Ego' and 'Superego' psychological roles. In roleplay simulations, this design allows intersubjective dialogue and intra-subjective internal monologue to develop in parallel. We apply this framework to two dramatic scenarios - an interview and a detective story - and compare character development with and without the Superego's influence. Though exploratory, results suggest this multi-agent approach can produce more nuanced, adaptive narratives that evolve over a sequence of dialogical turns. We discuss different modalities of LLM-based roleplay and character development, along with what this might mean for conceptualization of AI subjectivity. The paper concludes by considering how this approach opens possibilities for thinking of the roles of internal conflict and social performativity in AI-based simulation.", "authors": ["Liam Magee", "Vanicka Arora", "Gus Gollings", "Norma Lam-Saw"], "categories": ["cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-03", "url": "https://arxiv.org/abs/2408.01725", "pdf_url": "https://arxiv.org/pdf/2408.01725v2", "arxiv_id": "2408.01725", "doi": "10.48550/arXiv.2408.01725", "citation_count": 9, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.25}
{"id": "ced5f1d41a1845633683ae8d22b29ad7fd7d3dffff3e1d2ce9dc589026c3994e", "sources": ["arxiv", "semantic_scholar"], "title": "On the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents", "abstract": "Large language model-based multi-agent systems have shown great abilities across various tasks due to the collaboration of expert agents, each focusing on a specific domain. However, the impact of clumsy or even malicious agents--those who frequently make errors in their tasks--on the overall performance of the system remains underexplored. This paper investigates: (1) What is the resilience of various system structures (e.g., A$\\rightarrow$B$\\rightarrow$C, A$\\leftrightarrow$B$\\leftrightarrow$C) under faulty agents, on different downstream tasks? (2) How can we increase system resilience to defend against these agents? To simulate faulty agents, we propose two approaches--AutoTransform and AutoInject--which introduce mistakes into the agents' responses. Experiments on four downstream tasks using six systems show that the \"hierarchical\" structure, i.e., A$\\rightarrow$(B$\\leftrightarrow$C), exhibits superior resilience with the lowest performance drop of 5.5%, compared to 10.5% and 23.7% of other two structures. To further improve resilience, we introduce (1) Challenger, that introduces a mechanism for each agent to challenge others' outputs, and (2) Inspector, an additional agent to review and correct messages, recovering up to 96.4% errors made by faulty agents. Our code and data are available at https://github.com/CUHK-ARISE/MAS-Resilience.", "authors": ["Jen-tse Huang", "Jiaxu Zhou", "Tailin Jin", "Xuhui Zhou", "Zixi Chen", "Wenxuan Wang", "Youliang Yuan", "Michael R. Lyu", "Maarten Sap"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-02", "url": "https://arxiv.org/abs/2408.00989", "pdf_url": "https://arxiv.org/pdf/2408.00989v4", "arxiv_id": "2408.00989", "doi": null, "citation_count": 85, "influential_citation_count": 12, "has_code": true, "code_url": "https://github.com/CUHK-ARISE/MAS-Resilience", "venue": "International Conference on Machine Learning", "quality_score": 0.557}
{"id": "ce3410f79b9c0c280f922c2e43a735d81b11e661f1cb4f67aa6fcd9717b353d2", "sources": ["arxiv", "semantic_scholar"], "title": "GPUDrive: Data-driven, multi-agent driving simulation at 1 million FPS", "abstract": "Multi-agent learning algorithms have been successful at generating superhuman planning in various games but have had limited impact on the design of deployed multi-agent planners. A key bottleneck in applying these techniques to multi-agent planning is that they require billions of steps of experience. To enable the study of multi-agent planning at scale, we present GPUDrive. GPUDrive is a GPU-accelerated, multi-agent simulator built on top of the Madrona Game Engine capable of generating over a million simulation steps per second. Observation, reward, and dynamics functions are written directly in C++, allowing users to define complex, heterogeneous agent behaviors that are lowered to high-performance CUDA. Despite these low-level optimizations, GPUDrive is fully accessible through Python, offering a seamless and efficient workflow for multi-agent, closed-loop simulation. Using GPUDrive, we train reinforcement learning agents on the Waymo Open Motion Dataset, achieving efficient goal-reaching in minutes and scaling to thousands of scenarios in hours. We open-source the code and pre-trained agents at https://github.com/Emerge-Lab/gpudrive.", "authors": ["Saman Kazemkhani", "Aarav Pandya", "Daphne Cornelisse", "Brennan Shacklett", "Eugene Vinitsky"], "categories": ["cs.AI", "cs.AR", "cs.GR", "cs.PF"], "fields_of_study": ["Computer Science"], "published_date": "2024-08-02", "url": "https://arxiv.org/abs/2408.01584", "pdf_url": "https://arxiv.org/pdf/2408.01584v3", "arxiv_id": "2408.01584", "doi": "10.48550/arXiv.2408.01584", "citation_count": 40, "influential_citation_count": 8, "has_code": true, "code_url": "https://github.com/Emerge-Lab/gpudrive", "venue": "International Conference on Learning Representations", "quality_score": 0.4771}
{"id": "9623a645232d5e76d371c18c442afd758ef1d5c5edea7c1d7b3ac6e8a45c9030", "sources": ["arxiv", "semantic_scholar"], "title": "MetaOpenFOAM: an LLM-based multi-agent framework for CFD", "abstract": "Remarkable progress has been made in automated problem solving through societies of agents based on large language models (LLMs). Computational fluid dynamics (CFD), as a complex problem, presents unique challenges in automated simulations that require sophisticated solutions. MetaOpenFOAM, as a novel multi-agent collaborations framework, aims to complete CFD simulation tasks with only natural language as input. These simulation tasks include mesh pre-processing, simulation and so on. MetaOpenFOAM harnesses the power of MetaGPT's assembly line paradigm, which assigns diverse roles to various agents, efficiently breaking down complex CFD tasks into manageable subtasks. Langchain further complements MetaOpenFOAM by integrating Retrieval-Augmented Generation (RAG) technology, which enhances the framework's ability by integrating a searchable database of OpenFOAM tutorials for LLMs. Tests on a benchmark for natural language-based CFD solver, consisting of eight CFD simulation tasks, have shown that MetaOpenFOAM achieved a high pass rate per test (85%), with each test case costing only $0.22 on average. The eight CFD simulation tasks encompass a range of multidimensional flow problems, covering compressible and incompressible flows with different physical processes. This demonstrates the capability to automate CFD simulations using only natural language input, iteratively correcting errors to achieve the desired simulations. An ablation study was conducted to verify the necessity of each component in the multi-agent system and the RAG technology. A sensitivity study on the randomness of LLM showed that LLM with low randomness can obtain more stable and accurate results. Additionally, MetaOpenFOAM owns the ability to identify and modify key parameters in user requirements, and excels in correcting bugs when failure match occur,which demonstrates the generalization of MetaOpenFOAM.", "authors": ["Yuxuan Chen", "Xu Zhu", "Hua Zhou", "Zhuyin Ren"], "categories": ["cs.AI", "physics.flu-dyn"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2024-07-31", "url": "https://arxiv.org/abs/2407.21320", "pdf_url": "https://arxiv.org/pdf/2407.21320v2", "arxiv_id": "2407.21320", "doi": "10.48550/arXiv.2407.21320", "citation_count": 44, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4133}
{"id": "c2224f800f61f7216bba6f6aded94c4b7eb024f323133fe1a4c181568342e07c", "sources": ["arxiv", "semantic_scholar"], "title": "Tulip Agent -- Enabling LLM-Based Agents to Solve Tasks Using Large Tool Libraries", "abstract": "We introduce tulip agent, an architecture for autonomous LLM-based agents with Create, Read, Update, and Delete access to a tool library containing a potentially large number of tools. In contrast to state-of-the-art implementations, tulip agent does not encode the descriptions of all available tools in the system prompt, which counts against the model's context window, or embed the entire prompt for retrieving suitable tools. Instead, the tulip agent can recursively search for suitable tools in its extensible tool library, implemented exemplarily as a vector store. The tulip agent architecture significantly reduces inference costs, allows using even large tool libraries, and enables the agent to adapt and extend its set of tools. We evaluate the architecture with several ablation studies in a mathematics context and demonstrate its generalizability with an application to robotics. A reference implementation and the benchmark are available at github.com/HRI-EU/tulip_agent.", "authors": ["Felix Ocker", "Daniel Tanneberg", "Julian Eggert", "Michael Gienger"], "categories": ["cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-31", "url": "https://arxiv.org/abs/2407.21778", "pdf_url": "https://arxiv.org/pdf/2407.21778v1", "arxiv_id": "2407.21778", "doi": "10.48550/arXiv.2407.21778", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "139bc74c8c9387161940b5c8eea398ffeb218ea8e6ae606bca4c66d6fc9c64f1", "sources": ["arxiv", "semantic_scholar"], "title": "Breaking Agents: Compromising Autonomous LLM Agents Through Malfunction Amplification", "abstract": "Recently, autonomous agents built on large language models (LLMs) have experienced significant development and are being deployed in real-world applications. These agents can extend the base LLM's capabilities in multiple ways. For example, a well-built agent using GPT-3.5-Turbo as its core can outperform the more advanced GPT-4 model by leveraging external components. More importantly, the usage of tools enables these systems to perform actions in the real world, moving from merely generating text to actively interacting with their environment. Given the agents' practical applications and their ability to execute consequential actions, it is crucial to assess potential vulnerabilities. Such autonomous systems can cause more severe damage than a standalone language model if compromised. While some existing research has explored harmful actions by LLM agents, our study approaches the vulnerability from a different perspective. We introduce a new type of attack that causes malfunctions by misleading the agent into executing repetitive or irrelevant actions. We conduct comprehensive evaluations using various attack methods, surfaces, and properties to pinpoint areas of susceptibility. Our experiments reveal that these attacks can induce failure rates exceeding 80\\% in multiple scenarios. Through attacks on implemented and deployable agents in multi-agent scenarios, we accentuate the realistic risks associated with these vulnerabilities. To mitigate such attacks, we propose self-examination detection methods. However, our findings indicate these attacks are difficult to detect effectively using LLMs alone, highlighting the substantial risks associated with this vulnerability.", "authors": ["Boyang Zhang", "Yicong Tan", "Yun Shen", "Ahmed Salem", "Michael Backes", "Savvas Zannettou", "Yang Zhang"], "categories": ["cs.CR", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-30", "url": "https://arxiv.org/abs/2407.20859", "pdf_url": "https://arxiv.org/pdf/2407.20859v1", "arxiv_id": "2407.20859", "doi": "10.48550/arXiv.2407.20859", "citation_count": 87, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.4861}
{"id": "fe41f0cb54cb416fe707dc67f91c86613095583fe251781770a8f3bcc94fde17", "sources": ["arxiv", "semantic_scholar"], "title": "Very Large-Scale Multi-Agent Simulation in AgentScope", "abstract": "Recent advances in large language models (LLMs) have opened new avenues for applying multi-agent systems in very large-scale simulations. However, there remain several challenges when conducting multi-agent simulations with existing platforms, such as limited scalability and low efficiency, unsatisfied agent diversity, and effort-intensive management processes. To address these challenges, we develop several new features and components for AgentScope, a user-friendly multi-agent platform, enhancing its convenience and flexibility for supporting very large-scale multi-agent simulations. Specifically, we propose an actor-based distributed mechanism as the underlying technological infrastructure towards great scalability and high efficiency, and provide flexible environment support for simulating various real-world scenarios, which enables parallel execution of multiple agents, automatic workflow conversion for distributed deployment, and both inter-agent and agent-environment interactions. Moreover, we integrate an easy-to-use configurable tool and an automatic background generation pipeline in AgentScope, simplifying the process of creating agents with diverse yet detailed background settings. Last but not least, we provide a web-based interface for conveniently monitoring and managing a large number of agents that might deploy across multiple devices. We conduct a comprehensive simulation to demonstrate the effectiveness of these proposed enhancements in AgentScope, and provide detailed observations and insightful discussions to highlight the great potential of applying multi-agent systems in large-scale simulations. The source code is released on GitHub at https://github.com/modelscope/agentscope/tree/main/examples/paper_large_scale_simulation to inspire further research and development in large-scale multi-agent simulations.", "authors": ["Xuchen Pan", "Dawei Gao", "Yuexiang Xie", "Yushuo Chen", "Zhewei Wei", "Yaliang Li", "Bolin Ding", "Ji-Rong Wen", "Jingren Zhou"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-25", "url": "https://arxiv.org/abs/2407.17789", "pdf_url": "https://arxiv.org/pdf/2407.17789v2", "arxiv_id": "2407.17789", "doi": "10.48550/arXiv.2407.17789", "citation_count": 22, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/modelscope/agentscope/tree/main/examples/paper_large_scale_simulation", "venue": "arXiv.org", "quality_score": 0.3404}
{"id": "00fc2e9b7008f603f8a21fba02e86482d2b2a4e6fa19dacf8f53a733975c5596", "sources": ["arxiv", "semantic_scholar"], "title": "LawLuo: A Multi-Agent Collaborative Framework for Multi-Round Chinese Legal Consultation", "abstract": "Legal Large Language Models (LLMs) have shown promise in providing legal consultations to non-experts. However, most existing Chinese legal consultation models are based on single-agent systems, which differ from real-world legal consultations, where multiple professionals collaborate to offer more tailored responses. To better simulate real consultations, we propose LawLuo, a multi-agent framework for multi-turn Chinese legal consultations. LawLuo includes four agents: the receptionist agent, which assesses user intent and selects a lawyer agent; the lawyer agent, which interacts with the user; the secretary agent, which organizes conversation records and generates consultation reports; and the boss agent, which evaluates the performance of the lawyer and secretary agents to ensure optimal results. These agents' interactions mimic the operations of real law firms. To train them to follow different legal instructions, we developed distinct fine-tuning datasets. We also introduce a case graph-based RAG to help the lawyer agent address vague user inputs. Experimental results show that LawLuo outperforms baselines in generating more personalized and professional responses, handling ambiguous queries, and following legal instructions in multi-turn conversations. Our full code and constructed datasets will be open-sourced upon paper acceptance.", "authors": ["Jingyun Sun", "Chengxiao Dai", "Zhongze Luo", "Yangbo Chang", "Yang Li"], "categories": ["cs.CL", "cs.AI", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-23", "url": "https://arxiv.org/abs/2407.16252", "pdf_url": "https://arxiv.org/pdf/2407.16252v3", "arxiv_id": "2407.16252", "doi": null, "citation_count": 13, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.2865}
{"id": "e652cd32297886c2a27ef148f8954e73d3b06e02503523de4e99ab929d11bb79", "sources": ["arxiv", "semantic_scholar"], "title": "KoMA: Knowledge-driven Multi-agent Framework for Autonomous Driving with Large Language Models", "abstract": "Large language models (LLMs) as autonomous agents offer a novel avenue for tackling real-world challenges through a knowledge-driven manner. These LLM-enhanced methodologies excel in generalization and interpretability. However, the complexity of driving tasks often necessitates the collaboration of multiple, heterogeneous agents, underscoring the need for such LLM-driven agents to engage in cooperative knowledge sharing and cognitive synergy. Despite the promise of LLMs, current applications predominantly center around single agent scenarios. To broaden the horizons of knowledge-driven strategies and bolster the generalization capabilities of autonomous agents, we propose the KoMA framework consisting of multi-agent interaction, multi-step planning, shared-memory, and ranking-based reflection modules to enhance multi-agents' decision-making in complex driving scenarios. Based on the framework's generated text descriptions of driving scenarios, the multi-agent interaction module enables LLM agents to analyze and infer the intentions of surrounding vehicles, akin to human cognition. The multi-step planning module enables LLM agents to analyze and obtain final action decisions layer by layer to ensure consistent goals for short-term action decisions. The shared memory module can accumulate collective experience to make superior decisions, and the ranking-based reflection module can evaluate and improve agent behavior with the aim of enhancing driving safety and efficiency. The KoMA framework not only enhances the robustness and adaptability of autonomous driving agents but also significantly elevates their generalization capabilities across diverse scenarios. Empirical results demonstrate the superiority of our approach over traditional methods, particularly in its ability to handle complex, unpredictable driving environments without extensive retraining.", "authors": ["Kemou Jiang", "Xuan Cai", "Zhiyong Cui", "Aoyong Li", "Yilong Ren", "Haiyang Yu", "Hao Yang", "Daocheng Fu", "Licheng Wen", "Pinlong Cai"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-19", "url": "https://arxiv.org/abs/2407.14239", "pdf_url": "https://arxiv.org/pdf/2407.14239v1", "arxiv_id": "2407.14239", "doi": "10.1109/TIV.2024.3488793", "citation_count": 28, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Intelligent Vehicles", "quality_score": 0.3656}
{"id": "4641da8cd4ef980f9a16e036ea73f81de782113610bcae19eb68b40aa260fb31", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative Reward Shaping for Multi-Agent Pathfinding", "abstract": "The primary objective of Multi-Agent Pathfinding (MAPF) is to plan efficient and conflict-free paths for all agents. Traditional multi-agent path planning algorithms struggle to achieve efficient distributed path planning for multiple agents. In contrast, Multi-Agent Reinforcement Learning (MARL) has been demonstrated as an effective approach to achieve this objective. By modeling the MAPF problem as a MARL problem, agents can achieve efficient path planning and collision avoidance through distributed strategies under partial observation. However, MARL strategies often lack cooperation among agents due to the absence of global information, which subsequently leads to reduced MAPF efficiency. To address this challenge, this letter introduces a unique reward shaping technique based on Independent Q-Learning (IQL). The aim of this method is to evaluate the influence of one agent on its neighbors and integrate such an interaction into the reward function, leading to active cooperation among agents. This reward shaping method facilitates cooperation among agents while operating in a distributed manner. The proposed approach has been evaluated through experiments across various scenarios with different scales and agent counts. The results are compared with those from other state-of-the-art (SOTA) planners. The evidence suggests that the approach proposed in this letter parallels other planners in numerous aspects, and outperforms them in scenarios featuring a large number of agents.", "authors": ["Zhenyu Song", "Ronghao Zheng", "Senlin Zhang", "Meiqin Liu"], "categories": ["cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-15", "url": "https://arxiv.org/abs/2407.10403", "pdf_url": "https://arxiv.org/pdf/2407.10403v1", "arxiv_id": "2407.10403", "doi": "10.48550/arXiv.2407.10403", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "bbb7da54ba4d96c642f614a036fc84e4e09bf158da0925d3e8772cddfc3368e0", "sources": ["arxiv", "semantic_scholar"], "title": "Flooding Spread of Manipulated Knowledge in LLM-Based Multi-Agent Communities", "abstract": "The rapid adoption of large language models (LLMs) in multi-agent systems has highlighted their impressive capabilities in various applications, such as collaborative problem-solving and autonomous negotiation. However, the security implications of these LLM-based multi-agent systems have not been thoroughly investigated, particularly concerning the spread of manipulated knowledge. In this paper, we investigate this critical issue by constructing a detailed threat model and a comprehensive simulation environment that mirrors real-world multi-agent deployments in a trusted platform. Subsequently, we propose a novel two-stage attack method involving Persuasiveness Injection and Manipulated Knowledge Injection to systematically explore the potential for manipulated knowledge (i.e., counterfactual and toxic knowledge) spread without explicit prompt manipulation. Our method leverages the inherent vulnerabilities of LLMs in handling world knowledge, which can be exploited by attackers to unconsciously spread fabricated information. Through extensive experiments, we demonstrate that our attack method can successfully induce LLM-based agents to spread both counterfactual and toxic knowledge without degrading their foundational capabilities during agent communication. Furthermore, we show that these manipulations can persist through popular retrieval-augmented generation frameworks, where several benign agents store and retrieve manipulated chat histories for future interactions. This persistence indicates that even after the interaction has ended, the benign agents may continue to be influenced by manipulated knowledge. Our findings reveal significant security risks in LLM-based multi-agent systems, emphasizing the imperative need for robust defenses against manipulated knowledge spread, such as introducing ``guardian'' agents and advanced fact-checking tools.", "authors": ["Tianjie Ju", "Yiting Wang", "Xinbei Ma", "Pengzhou Cheng", "Haodong Zhao", "Yulong Wang", "Lifeng Liu", "Jian Xie", "Zhuosheng Zhang", "Gongshen Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-10", "url": "https://arxiv.org/abs/2407.07791", "pdf_url": "https://arxiv.org/pdf/2407.07791v2", "arxiv_id": "2407.07791", "doi": "10.48550/arXiv.2407.07791", "citation_count": 76, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4716}
{"id": "4712498cebba2a4cd5278c0987ace2169b9bc064b794154951190f0928370fff", "sources": ["arxiv", "semantic_scholar"], "title": "FinCon: A Synthesized LLM Multi-Agent System with Conceptual Verbal Reinforcement for Enhanced Financial Decision Making", "abstract": "Large language models (LLMs) have demonstrated notable potential in conducting complex tasks and are increasingly utilized in various financial applications. However, high-quality sequential financial investment decision-making remains challenging. These tasks require multiple interactions with a volatile environment for every decision, demanding sufficient intelligence to maximize returns and manage risks. Although LLMs have been used to develop agent systems that surpass human teams and yield impressive investment returns, opportunities to enhance multi-sourced information synthesis and optimize decision-making outcomes through timely experience refinement remain unexplored. Here, we introduce the FinCon, an LLM-based multi-agent framework with CONceptual verbal reinforcement tailored for diverse FINancial tasks. Inspired by effective real-world investment firm organizational structures, FinCon utilizes a manager-analyst communication hierarchy. This structure allows for synchronized cross-functional agent collaboration towards unified goals through natural language interactions and equips each agent with greater memory capacity than humans. Additionally, a risk-control component in FinCon enhances decision quality by episodically initiating a self-critiquing mechanism to update systematic investment beliefs. The conceptualized beliefs serve as verbal reinforcement for the future agent's behavior and can be selectively propagated to the appropriate node that requires knowledge updates. This feature significantly improves performance while reducing unnecessary peer-to-peer communication costs. Moreover, FinCon demonstrates strong generalization capabilities in various financial tasks, including single stock trading and portfolio management.", "authors": ["Yangyang Yu", "Zhiyuan Yao", "Haohang Li", "Zhiyang Deng", "Yupeng Cao", "Zhi Chen", "Jordan W. Suchow", "Rong Liu", "Zhenyu Cui", "Zhaozhuo Xu", "Denghui Zhang", "Koduvayur Subbalakshmi", "Guojun Xiong", "Yueru He", "Jimin Huang", "Dong Li", "Qianqian Xie"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-09", "url": "https://arxiv.org/abs/2407.06567", "pdf_url": "https://arxiv.org/pdf/2407.06567v3", "arxiv_id": "2407.06567", "doi": "10.48550/arXiv.2407.06567", "citation_count": 148, "influential_citation_count": 8, "has_code": false, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.5433}
{"id": "ad8084a06ec76a38eea8ca7f23d1bf8028766feba0ee58921397f8c87c626060", "sources": ["arxiv", "semantic_scholar"], "title": "Richelieu: Self-Evolving LLM-Based Agents for AI Diplomacy", "abstract": "Diplomacy is one of the most sophisticated activities in human society, involving complex interactions among multiple parties that require skills in social reasoning, negotiation, and long-term strategic planning. Previous AI agents have demonstrated their ability to handle multi-step games and large action spaces in multi-agent tasks. However, diplomacy involves a staggering magnitude of decision spaces, especially considering the negotiation stage required. While recent agents based on large language models (LLMs) have shown potential in various applications, they still struggle with extended planning periods in complex multi-agent settings. Leveraging recent technologies for LLM-based agents, we aim to explore AI's potential to create a human-like agent capable of executing comprehensive multi-agent missions by integrating three fundamental capabilities: 1) strategic planning with memory and reflection; 2) goal-oriented negotiation with social reasoning; and 3) augmenting memory through self-play games for self-evolution without human in the loop.", "authors": ["Zhenyu Guan", "Xiangyu Kong", "Fangwei Zhong", "Yizhou Wang"], "categories": ["cs.AI", "cs.MA", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-09", "url": "https://arxiv.org/abs/2407.06813", "pdf_url": "https://arxiv.org/pdf/2407.06813v4", "arxiv_id": "2407.06813", "doi": "10.52202/079017-3925", "citation_count": 44, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.4133}
{"id": "c834da7f544a237ad794a0e05628ee677470c148f8d5fdef1a58f610f369a37c", "sources": ["arxiv", "semantic_scholar"], "title": "Achieving Tool Calling Functionality in LLMs Using Only Prompt Engineering Without Fine-Tuning", "abstract": "Currently, the vast majority of locally deployed open-source large language models (LLMs) and some commercial model interfaces do not support stable tool calling functionality. The existing solution involves fine-tuning LLMs, which results in significant time and computational resource consumption. This paper proposes a method that enables LLMs to achieve stable tool calling capabilities using only prompt engineering and some ingenious code design. We conducted experiments on multiple LLMs that lack tool calling capabilities across various tool calling tasks, achieving a success rate of 100%.", "authors": ["Shengtao He"], "categories": ["cs.SE", "cs.AI", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-06", "url": "https://arxiv.org/abs/2407.04997", "pdf_url": "https://arxiv.org/pdf/2407.04997v1", "arxiv_id": "2407.04997", "doi": "10.48550/arXiv.2407.04997", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "e3ae1469a2fb63ce52a6b4dd9aaad32b6b180f97a63685b38ecf3405f1054982", "sources": ["arxiv", "semantic_scholar"], "title": "When LLMs Play the Telephone Game: Cultural Attractors as Conceptual Tools to Evaluate LLMs in Multi-turn Settings", "abstract": "As large language models (LLMs) start interacting with each other and generating an increasing amount of text online, it becomes crucial to better understand how information is transformed as it passes from one LLM to the next. While significant research has examined individual LLM behaviors, existing studies have largely overlooked the collective behaviors and information distortions arising from iterated LLM interactions. Small biases, negligible at the single output level, risk being amplified in iterated interactions, potentially leading the content to evolve towards attractor states. In a series of telephone game experiments, we apply a transmission chain design borrowed from the human cultural evolution literature: LLM agents iteratively receive, produce, and transmit texts from the previous to the next agent in the chain. By tracking the evolution of text toxicity, positivity, difficulty, and length across transmission chains, we uncover the existence of biases and attractors, and study their dependence on the initial text, the instructions, language model, and model size. For instance, we find that more open-ended instructions lead to stronger attraction effects compared to more constrained tasks. We also find that different text properties display different sensitivity to attraction effects, with toxicity leading to stronger attractors than length. These findings highlight the importance of accounting for multi-step transmission dynamics and represent a first step towards a more comprehensive understanding of LLM cultural dynamics.", "authors": ["Jérémy Perez", "Grgur Kovač", "Corentin Léger", "Cédric Colas", "Gaia Molinaro", "Maxime Derex", "Pierre-Yves Oudeyer", "Clément Moulin-Frier"], "categories": ["physics.soc-ph", "cs.AI", "cs.MA"], "fields_of_study": ["Physics", "Computer Science"], "published_date": "2024-07-05", "url": "https://arxiv.org/abs/2407.04503", "pdf_url": "https://arxiv.org/pdf/2407.04503v4", "arxiv_id": "2407.04503", "doi": null, "citation_count": 10, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/jeremyperez2/TelephoneGameLLM", "venue": "International Conference on Learning Representations", "quality_score": 0.2603}
{"id": "0cf8c9180d1c6c6e4b422d0f01bcf22a6600615ff7ae263257e56d339b6877f1", "sources": ["arxiv", "semantic_scholar"], "title": "AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents", "abstract": "Advancements in the capabilities of Large Language Models (LLMs) have created a promising foundation for developing autonomous agents. With the right tools, these agents could learn to solve tasks in new environments by accumulating and updating their knowledge. Current LLM-based agents process past experiences using a full history of observations, summarization, retrieval augmentation. However, these unstructured memory representations do not facilitate the reasoning and planning essential for complex decision-making. In our study, we introduce AriGraph, a novel method wherein the agent constructs and updates a memory graph that integrates semantic and episodic memories while exploring the environment. We demonstrate that our Ariadne LLM agent, consisting of the proposed memory architecture augmented with planning and decision-making, effectively handles complex tasks within interactive text game environments difficult even for human players. Results show that our approach markedly outperforms other established memory methods and strong RL baselines in a range of problems of varying complexity. Additionally, AriGraph demonstrates competitive performance compared to dedicated knowledge graph-based methods in static multi-hop question-answering.", "authors": ["Petr Anokhin", "Nikita Semenov", "Artyom Sorokin", "Dmitry Evseev", "Andrey Kravchenko", "Mikhail Burtsev", "Evgeny Burnaev"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-05", "url": "https://arxiv.org/abs/2407.04363", "pdf_url": "https://arxiv.org/pdf/2407.04363v3", "arxiv_id": "2407.04363", "doi": "10.48550/arXiv.2407.04363", "citation_count": 80, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/AIRI-Institute/AriGraph", "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.4771}
{"id": "e1c3fa8834837f1b089338bdc0ae5143d830ad677f41fe1ae287614bc1d5a85a", "sources": ["arxiv", "semantic_scholar"], "title": "MMedAgent: Learning to Use Medical Tools with Multi-modal Agent", "abstract": "Multi-Modal Large Language Models (MLLMs), despite being successful, exhibit limited generality and often fall short when compared to specialized models. Recently, LLM-based agents have been developed to address these challenges by selecting appropriate specialized models as tools based on user inputs. However, such advancements have not been extensively explored within the medical domain. To bridge this gap, this paper introduces the first agent explicitly designed for the medical field, named \\textbf{M}ulti-modal \\textbf{Med}ical \\textbf{Agent} (MMedAgent). We curate an instruction-tuning dataset comprising six medical tools solving seven tasks across five modalities, enabling the agent to choose the most suitable tools for a given task. Comprehensive experiments demonstrate that MMedAgent achieves superior performance across a variety of medical tasks compared to state-of-the-art open-source methods and even the closed-source model, GPT-4o. Furthermore, MMedAgent exhibits efficiency in updating and integrating new medical tools. Codes and models are all available.", "authors": ["Binxu Li", "Tiankai Yan", "Yuanting Pan", "Jie Luo", "Ruiyang Ji", "Jiayuan Ding", "Zhe Xu", "Shilong Liu", "Haoyu Dong", "Zihao Lin", "Yixin Wang"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-02", "url": "https://arxiv.org/abs/2407.02483", "pdf_url": "https://arxiv.org/pdf/2407.02483v2", "arxiv_id": "2407.02483", "doi": "10.48550/arXiv.2407.02483", "citation_count": 120, "influential_citation_count": 4, "has_code": true, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.5207}
{"id": "13ae3da2dc9c8e95755d04434c955db908dfa4a047b8ddabf9e4a164be5c4d1a", "sources": ["arxiv", "semantic_scholar"], "title": "Optimizing Age of Information in Vehicular Edge Computing with Federated Graph Neural Network Multi-Agent Reinforcement Learning", "abstract": "With the rapid development of intelligent vehicles and Intelligent Transport Systems (ITS), the sensors such as cameras and LiDAR installed on intelligent vehicles provides higher capacity of executing computation-intensive and delay-sensitive tasks, thereby raising deployment costs. To address this issue, Vehicular Edge Computing (VEC) has been proposed to process data through Road Side Units (RSUs) to support real-time applications. This paper focuses on the Age of Information (AoI) as a key metric for data freshness and explores task offloading issues for vehicles under RSU communication resource constraints. We adopt a Multi-agent Deep Reinforcement Learning (MADRL) approach, allowing vehicles to autonomously make optimal data offloading decisions. However, MADRL poses risks of vehicle information leakage during communication learning and centralized training. To mitigate this, we employ a Federated Learning (FL) framework that shares model parameters instead of raw data to protect the privacy of vehicle users. Building on this, we propose an innovative distributed federated learning framework combining Graph Neural Networks (GNN), named Federated Graph Neural Network Multi-Agent Reinforcement Learning (FGNN-MADRL), to optimize AoI across the system. For the first time, road scenarios are constructed as graph data structures, and a GNN-based federated learning framework is proposed, effectively combining distributed and centralized federated aggregation. Furthermore, we propose a new MADRL algorithm that simplifies decision making and enhances offloading efficiency, further reducing the decision complexity. Simulation results demonstrate the superiority of our proposed approach to other methods through simulations.", "authors": ["Wenhua Wang", "Qiong Wu", "Pingyi Fan", "Nan Cheng", "Wen Chen", "Jiangzhou Wang", "Khaled B. Letaief"], "categories": ["cs.LG", "cs.DC", "cs.MA", "cs.NI"], "fields_of_study": ["Computer Science"], "published_date": "2024-07-01", "url": "https://arxiv.org/abs/2407.02342", "pdf_url": "https://arxiv.org/pdf/2407.02342v1", "arxiv_id": "2407.02342", "doi": "10.48550/arXiv.2407.02342", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/qiongwu86/Optimizing-AoI-in-VEC-with-Federated-Graph-Neural-Network-Multi-Agent-Reinforcement-Learning", "venue": "arXiv.org", "quality_score": 0.2258}
{"id": "088a91107601d93982e7b1df1a7c9004249c76ea80681cb55159babd2d0447df", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Training for Pommerman: Curriculum Learning and Population-based Self-Play Approach", "abstract": "Pommerman is a multi-agent environment that has received considerable attention from researchers in recent years. This environment is an ideal benchmark for multi-agent training, providing a battleground for two teams with communication capabilities among allied agents. Pommerman presents significant challenges for model-free reinforcement learning due to delayed action effects, sparse rewards, and false positives, where opponent players can lose due to their own mistakes. This study introduces a system designed to train multi-agent systems to play Pommerman using a combination of curriculum learning and population-based self-play. We also tackle two challenging problems when deploying the multi-agent training system for competitive games: sparse reward and suitable matchmaking mechanism. Specifically, we propose an adaptive annealing factor based on agents' performance to adjust the dense exploration reward during training dynamically. Additionally, we implement a matchmaking mechanism utilizing the Elo rating system to pair agents effectively. Our experimental results demonstrate that our trained agent can outperform top learning agents without requiring communication among allied agents.", "authors": ["Nhat-Minh Huynh", "Hoang-Giang Cao", "I-Chen Wu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-30", "url": "https://arxiv.org/abs/2407.00662", "pdf_url": "https://arxiv.org/pdf/2407.00662v2", "arxiv_id": "2407.00662", "doi": "10.48550/arXiv.2407.00662", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "3888f006c37bb43145c505db3f0888f7e5686f3b5a5c492aebfb6accc9dcdd9a", "sources": ["arxiv", "semantic_scholar"], "title": "CAMON: Cooperative Agents for Multi-Object Navigation with LLM-based Conversations", "abstract": "Visual navigation tasks are critical for household service robots. As these tasks become increasingly complex, effective communication and collaboration among multiple robots become imperative to ensure successful completion. In recent years, large language models (LLMs) have exhibited remarkable comprehension and planning abilities in the context of embodied agents. However, their application in household scenarios, specifically in the use of multiple agents collaborating to complete complex navigation tasks through communication, remains unexplored. Therefore, this paper proposes a framework for decentralized multi-agent navigation, leveraging LLM-enabled communication and collaboration. By designing the communication-triggered dynamic leadership organization structure, we achieve faster team consensus with fewer communication instances, leading to better navigation effectiveness and collaborative exploration efficiency. With the proposed novel communication scheme, our framework promises to be conflict-free and robust in multi-object navigation tasks, even when there is a surge in team size.", "authors": ["Pengying Wu", "Yao Mu", "Kangjie Zhou", "Ji Ma", "Junting Chen", "Chang Liu"], "categories": ["cs.RO", "cs.CL", "cs.CV", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-30", "url": "https://arxiv.org/abs/2407.00632", "pdf_url": "https://arxiv.org/pdf/2407.00632v1", "arxiv_id": "2407.00632", "doi": "10.48550/arXiv.2407.00632", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "165fadedbe13931572052fb87626df6ed56d48db07778a4d1461c37b9a571f83", "sources": ["arxiv", "semantic_scholar"], "title": "BMW Agents -- A Framework For Task Automation Through Multi-Agent Collaboration", "abstract": "Autonomous agents driven by Large Language Models (LLMs) offer enormous potential for automation. Early proof of this technology can be found in various demonstrations of agents solving complex tasks, interacting with external systems to augment their knowledge, and triggering actions. In particular, workflows involving multiple agents solving complex tasks in a collaborative fashion exemplify their capacity to operate in less strict and less well-defined environments. Thus, a multi-agent approach has great potential for serving as a backbone in many industrial applications, ranging from complex knowledge retrieval systems to next generation robotic process automation. Given the reasoning abilities within the current generation of LLMs, complex processes require a multi-step approach that includes a plan of well-defined and modular tasks. Depending on the level of complexity, these tasks can be executed either by a single agent or a group of agents. In this work, we focus on designing a flexible agent engineering framework with careful attention to planning and execution, capable of handling complex use case applications across various domains. The proposed framework provides reliability in industrial applications and presents techniques to ensure a scalable, flexible, and collaborative workflow for multiple autonomous agents working together towards solving tasks.", "authors": ["Noel Crawford", "Edward B. Duffy", "Iman Evazzade", "Torsten Foehr", "Gregory Robbins", "Debbrata Kumar Saha", "Jiya Varma", "Marcin Ziolkowski"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-28", "url": "https://arxiv.org/abs/2406.20041", "pdf_url": "https://arxiv.org/pdf/2406.20041v3", "arxiv_id": "2406.20041", "doi": "10.48550/arXiv.2406.20041", "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "1bde5b9a48c6e5accbd8622dab74abf5f983347261114a0062c000d99a5a8e2c", "sources": ["arxiv", "semantic_scholar"], "title": "Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks", "abstract": "Large language models (LLMs) have recently shown tremendous promise in serving as the backbone to agentic systems, as demonstrated by their performance in multi-faceted, challenging benchmarks like SWE-Bench and Agent-Bench. However, to realize the true potential of LLMs as autonomous agents, they must learn to identify, call, and interact with external tools and application program interfaces (APIs) to complete complex tasks. These tasks together are termed function calling. Endowing LLMs with function calling abilities leads to a myriad of advantages, such as access to current and domain-specific information in databases and knowledge sources, and the ability to outsource tasks that can be reliably performed by tools, e.g., a Python interpreter or calculator. While there has been significant progress in function calling with LLMs, there is still a dearth of open models that perform on par with proprietary LLMs like GPT, Claude, and Gemini. Therefore, in this work, we introduce the GRANITE-20B-FUNCTIONCALLING model under an Apache 2.0 license. The model is trained using a multi-task training approach on seven fundamental tasks encompassed in function calling, those being Nested Function Calling, Function Chaining, Parallel Functions, Function Name Detection, Parameter-Value Pair Detection, Next-Best Function, and Response Generation. We present a comprehensive evaluation on multiple out-of-domain datasets comparing GRANITE-20B-FUNCTIONCALLING to more than 15 other best proprietary and open models. GRANITE-20B-FUNCTIONCALLING provides the best performance among all open models on the Berkeley Function Calling Leaderboard and fourth overall. As a result of the diverse tasks and datasets used for training our model, we show that GRANITE-20B-FUNCTIONCALLING has better generalizability on multiple tasks in seven different evaluation datasets.", "authors": ["Ibrahim Abdelaziz", "Kinjal Basu", "Mayank Agarwal", "Sadhana Kumaravel", "Matthew Stallone", "Rameswar Panda", "Yara Rizk", "GP Bhargav", "Maxwell Crouse", "Chulaka Gunasekara", "Shajith Ikbal", "Sachin Joshi", "Hima Karanam", "Vineet Kumar", "Asim Munawar", "Sumit Neelam", "Dinesh Raghu", "Udit Sharma", "Adriana Meza Soria", "Dheeraj Sreedhar", "Praveen Venkateswaran", "Merve Unuvar", "David Cox", "Salim Roukos", "Luis Lastras", "Pavan Kapanipathi"], "categories": ["cs.LG", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-27", "url": "https://arxiv.org/abs/2407.00121", "pdf_url": "https://arxiv.org/pdf/2407.00121v1", "arxiv_id": "2407.00121", "doi": "10.48550/arXiv.2407.00121", "citation_count": 61, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.4481}
{"id": "ed11cad47e148348dbd1e460b1f6c82f2bee01f035a05a11c06d343d270141ab", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Cooperative Games Using Belief Map Assisted Training", "abstract": "In a multi-agent system, agents share their local observations to gain global situational awareness for decision making and collaboration using a message passing system. When to send a message, how to encode a message, and how to leverage the received messages directly affect the effectiveness of the collaboration among agents. When training a multi-agent cooperative game using reinforcement learning (RL), the message passing system needs to be optimized together with the agent policies. This consequently increases the model's complexity and poses significant challenges to the convergence and performance of learning. To address this issue, we propose the Belief-map Assisted Multi-agent System (BAMS), which leverages a neuro-symbolic belief map to enhance training. The belief map decodes the agent's hidden state to provide a symbolic representation of the agent's understanding of the environment and other agent's status. The simplicity of symbolic representation allows the gathering and comparison of the ground truth information with the belief, which provides an additional channel of feedback for the learning. Compared to the sporadic and delayed feedback coming from the reward in RL, the feedback from the belief map is more consistent and reliable. Agents using BAMS can learn a more effective message passing network to better understand each other, resulting in better performance in a cooperative predator and prey game with varying levels of map complexity and compare it to previous multi-agent message passing models. The simulation results showed that BAMS reduced training epochs by 66\\%, and agents who apply the BAMS model completed the game with 34.62\\% fewer steps on average.", "authors": ["Qinwei Huang", "Chen Luo", "Alex B. Wu", "Simon Khan", "Hai Li", "Qinru Qiu"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-27", "url": "https://arxiv.org/abs/2406.19477", "pdf_url": "https://arxiv.org/pdf/2406.19477v1", "arxiv_id": "2406.19477", "doi": "10.3233/FAIA230444", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.1193}
{"id": "54b374e246b7fb556c24ce7cfa32a9f8f3853d99cfccfab7d091a73886221a46", "sources": ["arxiv", "semantic_scholar"], "title": "Simulating Classroom Education with LLM-Empowered Agents", "abstract": "Large language models (LLMs) have been applied across various intelligent educational tasks to assist teaching. While preliminary studies have focused on task-specific, independent LLM-empowered agents, the potential of LLMs within a multi-agent collaborative framework for classroom simulation with real user participation remains unexplored. In this work, we propose SimClass, a multi-agent classroom simulation teaching framework. We recognize representative class roles and introduce a novel class control mechanism for automatic classroom teaching, and conduct user experiments in two real-world courses. Using the Flanders Interactive Analysis System and Community of Inquiry theoretical frameworks from educational analysis, we demonstrate that LLMs can simulate a dynamic learning environment for users with active teacher-student and student-student interactions. We also observe group behaviors among agents in SimClass, where agents collaborate to create enlivening interactions in classrooms to improve user learning process. We hope this work pioneers the application of LLM-empowered multi-agent systems in virtual classroom teaching.", "authors": ["Zheyuan Zhang", "Daniel Zhang-Li", "Jifan Yu", "Linlu Gong", "Jinchang Zhou", "Zhanxin Hao", "Jianxiao Jiang", "Jie Cao", "Huiqin Liu", "Zhiyuan Liu", "Lei Hou", "Juanzi Li"], "categories": ["cs.CL", "cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-27", "url": "https://arxiv.org/abs/2406.19226", "pdf_url": "https://arxiv.org/pdf/2406.19226v2", "arxiv_id": "2406.19226", "doi": "10.48550/arXiv.2406.19226", "citation_count": 177, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.5626}
{"id": "de8dfdb8a309ed077a57985d5a2eeb448801e8f5ba23e6ea1258448811968f49", "sources": ["arxiv", "semantic_scholar"], "title": "Direct Multi-Turn Preference Optimization for Language Agents", "abstract": "Adapting Large Language Models (LLMs) for agent tasks is critical in developing language agents. Direct Preference Optimization (DPO) is a promising technique for this adaptation with the alleviation of compounding errors, offering a means to directly optimize Reinforcement Learning (RL) objectives. However, applying DPO to multi-turn tasks presents challenges due to the inability to cancel the partition function. Overcoming this obstacle involves making the partition function independent of the current state and addressing length disparities between preferred and dis-preferred trajectories. In this light, we replace the policy constraint with the state-action occupancy measure constraint in the RL objective and add length normalization to the Bradley-Terry model, yielding a novel loss function named DMPO for multi-turn agent tasks with theoretical explanations. Extensive experiments on three multi-turn agent task datasets confirm the effectiveness and superiority of the DMPO loss. The code is available at https://github.com/swt-user/DMPO.", "authors": ["Wentao Shi", "Mengqi Yuan", "Junkang Wu", "Qifan Wang", "Fuli Feng"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-21", "url": "https://arxiv.org/abs/2406.14868", "pdf_url": "https://arxiv.org/pdf/2406.14868v5", "arxiv_id": "2406.14868", "doi": "10.48550/arXiv.2406.14868", "citation_count": 60, "influential_citation_count": 9, "has_code": true, "code_url": "https://github.com/swt-user/DMPO", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.5}
{"id": "ab6c7042ed1fbbed1a3455661ebc9b7c583ce9b68decae15fe3427a36917ea5e", "sources": ["arxiv", "semantic_scholar"], "title": "EvoAgent: Towards Automatic Multi-Agent Generation via Evolutionary Algorithms", "abstract": "The rise of powerful large language models (LLMs) has spurred a new trend in building LLM-based autonomous agents for solving complex tasks, especially multi-agent systems. Despite the remarkable progress, we notice that existing works are heavily dependent on human-designed frameworks, which greatly limits the functional scope and scalability of agent systems. How to automatically extend the specialized agent to multi-agent systems to improve task-solving capability still remains a significant challenge. In this paper, we introduce EvoAgent, a generic method to automatically extend specialized agents to multi-agent systems via the evolutionary algorithm, thereby improving the effectiveness of LLM-based agents in solving tasks. Specifically, we consider the existing agent frameworks as the initial individual and then apply a series of evolutionary operators (e.g., mutation, crossover, selection, etc.) to generate multiple agents with diverse settings. Experimental results across various tasks show that EvoAgent can significantly enhance the task-solving capability of LLM-based agents, and can be generalized to any LLM-based agent framework to extend them into multi-agent systems. Resources are available at https://evo-agent.github.io/.", "authors": ["Siyu Yuan", "Kaitao Song", "Jiangjie Chen", "Xu Tan", "Dongsheng Li", "Deqing Yang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-20", "url": "https://arxiv.org/abs/2406.14228", "pdf_url": "https://arxiv.org/pdf/2406.14228v3", "arxiv_id": "2406.14228", "doi": "10.48550/arXiv.2406.14228", "citation_count": 95, "influential_citation_count": 9, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.5}
{"id": "4c82653b1ce8df2465e45e1f4a9cc499a1e6dbcc4817180b7d610903cd6cec1a", "sources": ["arxiv", "semantic_scholar"], "title": "Tractable Equilibrium Computation in Markov Games through Risk Aversion", "abstract": "A significant roadblock to the development of principled multi-agent reinforcement learning is the fact that desired solution concepts like Nash equilibria may be intractable to compute. To overcome this obstacle, we take inspiration from behavioral economics and show that -- by imbuing agents with important features of human decision-making like risk aversion and bounded rationality -- a class of risk-averse quantal response equilibria (RQE) become tractable to compute in all $n$-player matrix and finite-horizon Markov games. In particular, we show that they emerge as the endpoint of no-regret learning in suitably adjusted versions of the games. Crucially, the class of computationally tractable RQE is independent of the underlying game structure and only depends on agents' degree of risk-aversion and bounded rationality. To validate the richness of this class of solution concepts we show that it captures peoples' patterns of play in a number of 2-player matrix games previously studied in experimental economics. Furthermore, we give a first analysis of the sample complexity of computing these equilibria in finite-horizon Markov games when one has access to a generative model and validate our findings on a simple multi-agent reinforcement learning benchmark.", "authors": ["Eric Mazumdar", "Kishan Panaganti", "Laixi Shi"], "categories": ["cs.GT", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-20", "url": "https://arxiv.org/abs/2406.14156", "pdf_url": "https://arxiv.org/pdf/2406.14156v2", "arxiv_id": "2406.14156", "doi": "10.48550/arXiv.2406.14156", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "dc762320c01ae18491b625e6f9f934a80d240cbb9507f09320028338b2c7e784", "sources": ["arxiv", "semantic_scholar"], "title": "CodeNav: Beyond tool-use to using real-world codebases with LLM agents", "abstract": "We present CodeNav, an LLM agent that navigates and leverages previously unseen code repositories to solve user queries. In contrast to tool-use LLM agents that require ``registration'' of all relevant tools via manual descriptions within the LLM context, CodeNav automatically indexes and searches over code blocks in the target codebase, finds relevant code snippets, imports them, and uses them to iteratively generate a solution with execution feedback. To highlight the core-capabilities of CodeNav, we first showcase three case studies where we use CodeNav for solving complex user queries using three diverse codebases. Next, on three benchmarks, we quantitatively compare the effectiveness of code-use (which only has access to the target codebase) to tool-use (which has privileged access to all tool names and descriptions). Finally, we study the effect of varying kinds of tool and library descriptions on code-use performance, as well as investigate the advantage of the agent seeing source code as opposed to natural descriptions of code. All code will be made open source under a permissive license.", "authors": ["Tanmay Gupta", "Luca Weihs", "Aniruddha Kembhavi"], "categories": ["cs.AI", "cs.CL", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-18", "url": "https://arxiv.org/abs/2406.12276", "pdf_url": "https://arxiv.org/pdf/2406.12276v1", "arxiv_id": "2406.12276", "doi": "10.48550/arXiv.2406.12276", "citation_count": 6, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2113}
{"id": "4bce95110cfd9f1bd764421fc4b8ef04469557bdcddfd98aebca255c8ea6d7cf", "sources": ["arxiv", "semantic_scholar"], "title": "AvaTaR: Optimizing LLM Agents for Tool Usage via Contrastive Reasoning", "abstract": "Large language model (LLM) agents have demonstrated impressive capabilities in utilizing external tools and knowledge to boost accuracy and reduce hallucinations. However, developing prompting techniques that enable LLM agents to effectively use these tools and knowledge remains a heuristic and labor-intensive task. Here, we introduce AvaTaR, a novel and automated framework that optimizes an LLM agent to effectively leverage provided tools, improving performance on a given task. During optimization, we design a comparator module to iteratively deliver insightful and comprehensive prompts to the LLM agent by contrastively reasoning between positive and negative examples sampled from training data. We demonstrate AvaTaR on four complex multimodal retrieval datasets featuring textual, visual, and relational information, and three general question-answering (QA) datasets. We find AvaTaR consistently outperforms state-of-the-art approaches across all seven tasks, exhibiting strong generalization ability when applied to novel cases and achieving an average relative improvement of 14% on the Hit@1 metric for the retrieval datasets and 13% for the QA datasets. Code and dataset are available at https://github.com/zou-group/avatar.", "authors": ["Shirley Wu", "Shiyu Zhao", "Qian Huang", "Kexin Huang", "Michihiro Yasunaga", "Kaidi Cao", "Vassilis N. Ioannidis", "Karthik Subbian", "Jure Leskovec", "James Zou"], "categories": ["cs.LG", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-17", "url": "https://arxiv.org/abs/2406.11200", "pdf_url": "https://arxiv.org/pdf/2406.11200v3", "arxiv_id": "2406.11200", "doi": "10.52202/079017-0817", "citation_count": 76, "influential_citation_count": 6, "has_code": true, "code_url": "https://github.com/zou-group/avatar", "venue": "Neural Information Processing Systems", "quality_score": 0.4716}
{"id": "8550091457fd109ac68b47e2b2d7188a15042b3853809102237389ddb6b85659", "sources": ["arxiv", "semantic_scholar"], "title": "Improving Multi-Agent Debate with Sparse Communication Topology", "abstract": "Multi-agent debate has proven effective in improving large language models quality for reasoning and factuality tasks. While various role-playing strategies in multi-agent debates have been explored, in terms of the communication among agents, existing approaches adopt a brute force algorithm -- each agent can communicate with all other agents. In this paper, we systematically investigate the effect of communication connectivity in multi-agent systems. Our experiments on GPT and Mistral models reveal that multi-agent debates leveraging sparse communication topology can achieve comparable or superior performance while significantly reducing computational costs. Furthermore, we extend the multi-agent debate framework to multimodal reasoning and alignment labeling tasks, showcasing its broad applicability and effectiveness. Our findings underscore the importance of communication connectivity on enhancing the efficiency and effectiveness of the \"society of minds\" approach.", "authors": ["Yunxuan Li", "Yibing Du", "Jiageng Zhang", "Le Hou", "Peter Grabowski", "Yeqing Li", "Eugene Ie"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-17", "url": "https://arxiv.org/abs/2406.11776", "pdf_url": "https://arxiv.org/pdf/2406.11776v1", "arxiv_id": "2406.11776", "doi": "10.48550/arXiv.2406.11776", "citation_count": 109, "influential_citation_count": 12, "has_code": false, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.557}
{"id": "9c6fc089c618bb54492241a8bafffc598134d3a51a61379cc850398adda61ca7", "sources": ["arxiv", "semantic_scholar"], "title": "Husky: A Unified, Open-Source Language Agent for Multi-Step Reasoning", "abstract": "Language agents perform complex tasks by using tools to execute each step precisely. However, most existing agents are based on proprietary models or designed to target specific tasks, such as mathematics or multi-hop question answering. We introduce Husky, a holistic, open-source language agent that learns to reason over a unified action space to address a diverse set of complex tasks involving numerical, tabular, and knowledge-based reasoning. Husky iterates between two stages: 1) generating the next action to take towards solving a given task and 2) executing the action using expert models and updating the current solution state. We identify a thorough ontology of actions for addressing complex tasks and curate high-quality data to train expert models for executing these actions. Our experiments show that Husky outperforms prior language agents across 14 evaluation datasets. Moreover, we introduce HuskyQA, a new evaluation set which stress tests language agents for mixed-tool reasoning, with a focus on retrieving missing knowledge and performing numerical reasoning. Despite using 7B models, Husky matches or even exceeds frontier LMs such as GPT-4 on these tasks, showcasing the efficacy of our holistic approach in addressing complex reasoning problems. Our code and models are available at https://github.com/agent-husky/Husky-v1.", "authors": ["Joongwon Kim", "Bhargavi Paranjape", "Tushar Khot", "Hannaneh Hajishirzi"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-10", "url": "https://arxiv.org/abs/2406.06469", "pdf_url": "https://arxiv.org/pdf/2406.06469v1", "arxiv_id": "2406.06469", "doi": "10.48550/arXiv.2406.06469", "citation_count": 14, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/agent-husky/Husky-v1", "venue": "arXiv.org", "quality_score": 0.301}
{"id": "fbad66800b0628fc6dc6e3c6913b86af49b3fa1d10900cfc927f04bcb3aa178e", "sources": ["arxiv", "semantic_scholar"], "title": "Large Language Model-Enabled Multi-Agent Manufacturing Systems", "abstract": "Traditional manufacturing faces challenges adapting to dynamic environments and quickly responding to manufacturing changes. The use of multi-agent systems has improved adaptability and coordination but requires further advancements in rapid human instruction comprehension, operational adaptability, and coordination through natural language integration. Large language models like GPT-3.5 and GPT-4 enhance multi-agent manufacturing systems by enabling agents to communicate in natural language and interpret human instructions for decision-making. This research introduces a novel framework where large language models enhance the capabilities of agents in manufacturing, making them more adaptable, and capable of processing context-specific instructions. A case study demonstrates the practical application of this framework, showing how agents can effectively communicate, understand tasks, and execute manufacturing processes, including precise G-code allocation among agents. The findings highlight the importance of continuous large language model integration into multi-agent manufacturing systems and the development of sophisticated agent communication protocols for a more flexible manufacturing system.", "authors": ["Jonghan Lim", "Birgit Vogel-Heuser", "Ilya Kovalenko"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-04", "url": "https://arxiv.org/abs/2406.01893", "pdf_url": "https://arxiv.org/pdf/2406.01893v2", "arxiv_id": "2406.01893", "doi": "10.1109/CASE59546.2024.10711432", "citation_count": 39, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.4005}
{"id": "d490b934f2df8b6a5c3c202437410f09a0df893c3bf82787d28cef729647167a", "sources": ["arxiv", "semantic_scholar"], "title": "Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration", "abstract": "Mobile device operation tasks are increasingly becoming a popular multi-modal AI application scenario. Current Multi-modal Large Language Models (MLLMs), constrained by their training data, lack the capability to function effectively as operation assistants. Instead, MLLM-based agents, which enhance capabilities through tool invocation, are gradually being applied to this scenario. However, the two major navigation challenges in mobile device operation tasks, task progress navigation and focus content navigation, are significantly complicated under the single-agent architecture of existing work. This is due to the overly long token sequences and the interleaved text-image data format, which limit performance. To address these navigation challenges effectively, we propose Mobile-Agent-v2, a multi-agent architecture for mobile device operation assistance. The architecture comprises three agents: planning agent, decision agent, and reflection agent. The planning agent generates task progress, making the navigation of history operations more efficient. To retain focus content, we design a memory unit that updates with task progress. Additionally, to correct erroneous operations, the reflection agent observes the outcomes of each operation and handles any mistakes accordingly. Experimental results indicate that Mobile-Agent-v2 achieves over a 30% improvement in task completion compared to the single-agent architecture of Mobile-Agent. The code is open-sourced at https://github.com/X-PLUG/MobileAgent.", "authors": ["Junyang Wang", "Haiyang Xu", "Haitao Jia", "Xi Zhang", "Ming Yan", "Weizhou Shen", "Ji Zhang", "Fei Huang", "Jitao Sang"], "categories": ["cs.CL", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-03", "url": "https://arxiv.org/abs/2406.01014", "pdf_url": "https://arxiv.org/pdf/2406.01014v1", "arxiv_id": "2406.01014", "doi": "10.48550/arXiv.2406.01014", "citation_count": 207, "influential_citation_count": 23, "has_code": true, "code_url": "https://github.com/X-PLUG/MobileAgent", "venue": "Neural Information Processing Systems", "quality_score": 0.6901}
{"id": "5d510621df1af61993e486569d632a3169160ed273a382ffb3269724f1bcce10", "sources": ["arxiv", "semantic_scholar"], "title": "Demystifying AI Platform Design for Distributed Inference of Next-Generation LLM models", "abstract": "Large language models (LLMs) have shown remarkable performance across a wide range of applications, often outperforming human experts. However, deploying these gigantic models efficiently for diverse inference use cases requires carefully designed hardware platforms with ample computing, memory, and network resources. With constant innovation in LLM serving optimizations and model architecture evolving at breakneck speed, the hardware requirements to meet Service Level Objectives (SLOs) remain an open research question. To answer the question, we present an analytical tool, GenZ, to efficiently navigate the relationship between diverse LLM model architectures(Dense, GQA, MoE, Mamba), LLM serving optimizations(Chunking, Speculative decoding, quanitization), and AI platform design parameters. Our tool estimates LLM inference performance metrics for the given scenario. We have validated against real hardware platforms running various different LLM models, achieving a max geomean error of 5.82.We use GenZ to identify compute, memory capacity, memory bandwidth, network latency, and network bandwidth requirements across diverse LLM inference use cases. We also study diverse architectural choices in use today (inspired by LLM serving platforms from several vendors) to help inform computer architects designing next-generation AI hardware accelerators and platforms. The trends and insights derived from GenZ can guide AI engineers deploying LLMs as well as computer architects designing next-generation hardware accelerators and platforms. Ultimately, this work sheds light on the platform design considerations for unlocking the full potential of large language models across a spectrum of applications. The source code is available at https://github.com/abhibambhaniya/GenZ-LLM-Analyzer . Users can also be tried it on at https://genz-llm-analyzer.streamlit.app/ without any setup on your web browser.", "authors": ["Abhimanyu Bambhaniya", "Ritik Raj", "Geonhwa Jeong", "Souvik Kundu", "Sudarshan Srinivasan", "Suvinay Subramanian", "Midhilesh Elavazhagan", "Madhu Kumar", "Tushar Krishna"], "categories": ["cs.AR", "cs.AI", "cs.DC", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-03", "url": "https://arxiv.org/abs/2406.01698", "pdf_url": "https://arxiv.org/pdf/2406.01698v3", "arxiv_id": "2406.01698", "doi": null, "citation_count": 12, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/abhibambhaniya/GenZ-LLM-Analyzer", "venue": null, "quality_score": 0.2785}
{"id": "e6b4fa7e37b7a7204f1dc9538623287f8e1e573762e507784b99693560eec027", "sources": ["arxiv", "semantic_scholar"], "title": "Domain-specific ReAct for physics-integrated iterative modeling: A case study of LLM agents for gas path analysis of gas turbines", "abstract": "This study explores the application of large language models (LLMs) with callable tools in energy and power engineering domain, focusing on gas path analysis of gas turbines. We developed a dual-agent tool-calling process to integrate expert knowledge, predefined tools, and LLM reasoning. We evaluated various LLMs, including LLama3, Qwen1.5 and GPT. Smaller models struggled with tool usage and parameter extraction, while larger models demonstrated favorable capabilities. All models faced challenges with complex, multi-component problems. Based on the test results, we infer that LLMs with nearly 100 billion parameters could meet professional scenario requirements with fine-tuning and advanced prompt design. Continued development are likely to enhance their accuracy and effectiveness, paving the way for more robust AI-driven solutions.", "authors": ["Tao Song", "Yuwei Fan", "Chenlong Feng", "Keyu Song", "Chao Liu", "Dongxiang Jiang"], "categories": ["cs.AI", "cs.CE", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-06-01", "url": "https://arxiv.org/abs/2406.07572", "pdf_url": "https://arxiv.org/pdf/2406.07572v1", "arxiv_id": "2406.07572", "doi": "10.48550/arXiv.2406.07572", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "7a176dc0aaabb46ac580d8eaf055dfed6f1aef9733aced2b61121d6046112585", "sources": ["arxiv", "semantic_scholar"], "title": "LLM experiments with simulation: Large Language Model Multi-Agent System for Simulation Model Parametrization in Digital Twins", "abstract": "This paper presents a novel design of a multi-agent system framework that applies large language models (LLMs) to automate the parametrization of simulation models in digital twins. This framework features specialized LLM agents tasked with observing, reasoning, decision-making, and summarizing, enabling them to dynamically interact with digital twin simulations to explore parametrization possibilities and determine feasible parameter settings to achieve an objective. The proposed approach enhances the usability of simulation model by infusing it with knowledge heuristics from LLM and enables autonomous search for feasible parametrization to solve a user task. Furthermore, the system has the potential to increase user-friendliness and reduce the cognitive load on human users by assisting in complex decision-making processes. The effectiveness and functionality of the system are demonstrated through a case study, and the visualized demos and codes are available at a GitHub Repository: https://github.com/YuchenXia/LLMDrivenSimulation", "authors": ["Yuchen Xia", "Daniel Dittler", "Nasser Jazdi", "Haonan Chen", "Michael Weyrich"], "categories": ["cs.AI", "cs.ET", "cs.MA", "cs.RO", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2024-05-28", "url": "https://arxiv.org/abs/2405.18092", "pdf_url": "https://arxiv.org/pdf/2405.18092v2", "arxiv_id": "2405.18092", "doi": "10.1109/ETFA61755.2024.10710900", "citation_count": 22, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/YuchenXia/LLMDrivenSimulation", "venue": "IEEE International Conference on Emerging Technologies and Factory Automation", "quality_score": 0.3404}
{"id": "9dff692b6ad99dea7cbc446c22b62b5016d4cfb415438306c4474bd2bc83981c", "sources": ["arxiv", "semantic_scholar"], "title": "PyTAG: Tabletop Games for Multi-Agent Reinforcement Learning", "abstract": "Modern Tabletop Games present various interesting challenges for Multi-agent Reinforcement Learning. In this paper, we introduce PyTAG, a new framework that supports interacting with a large collection of games implemented in the Tabletop Games framework. In this work we highlight the challenges tabletop games provide, from a game-playing agent perspective, along with the opportunities they provide for future research. Additionally, we highlight the technical challenges that involve training Reinforcement Learning agents on these games. To explore the Multi-agent setting provided by PyTAG we train the popular Proximal Policy Optimisation Reinforcement Learning algorithm using self-play on a subset of games and evaluate the trained policies against some simple agents and Monte-Carlo Tree Search implemented in the Tabletop Games framework.", "authors": ["Martin Balla", "George E. M. Long", "James Goodman", "Raluca D. Gaina", "Diego Perez-Liebana"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-28", "url": "https://arxiv.org/abs/2405.18123", "pdf_url": "https://arxiv.org/pdf/2405.18123v1", "arxiv_id": "2405.18123", "doi": "10.1109/TG.2024.3404133", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Games", "quality_score": 0.1747}
{"id": "6adea89c5db724b00189f6dcb979a3dab9c6aba2b8dee88b294987927bd2654b", "sources": ["arxiv", "semantic_scholar"], "title": "A Large Language Model-based multi-agent manufacturing system for intelligent shopfloor", "abstract": "As customer demand for multi-variety and small-batch production increases, dynamic disturbances place greater demands on manufacturing systems. To address such challenges, researchers proposed the multi-agent manufacturing system. However, conventional agent negotiation typically relies on pre-defined and fixed heuristic rules, which are ill-suited to managing complex and fluctuating disturbances. In current implementations, mainstream approaches based on reinforcement learning require the development of simulators and training models specific to a given shopfloor, necessitating substantial computational resources and lacking scalability. To overcome this limitation, the present study proposes a Large Language Model-based (LLM-based) multi-agent manufacturing system for intelligent shopfloor management. By defining the diverse modules of agents and their collaborative methods, this system facilitates the processing of all workpieces with minimal human intervention. The agents in this system consist of the Machine Server Module (MSM), Bid Inviter Module (BIM), Bidder Module (BM), Thinking Module (TM), and Decision Module (DM). By harnessing the reasoning capabilities of LLMs, these modules enable agents to dynamically analyze shopfloor information and select appropriate processing machines. The LLM-based modules, predefined by system prompts, provide dynamic functionality for the system without the need for pre-training. Extensive experiments were conducted in physical shopfloor settings. The results demonstrate that the proposed system exhibits strong adaptability, and achieves superior performance (makespan) and stability (as measured by sample standard deviation) compared to other approaches without requiring pre-training.", "authors": ["Zhen Zhao", "Dunbing Tang", "Changchun Liu", "Liping Wang", "Zequn Zhang", "Haihua Zhu", "Kai Chen", "Qingwei Nie", "Yuchen Ji"], "categories": ["cs.AI", "cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-27", "url": "https://arxiv.org/abs/2405.16887", "pdf_url": "https://arxiv.org/pdf/2405.16887v2", "arxiv_id": "2405.16887", "doi": "10.1016/j.aei.2025.103888", "citation_count": 20, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Advanced Engineering Informatics", "quality_score": 0.3306}
{"id": "c77bbd458fbb9e201a6ec7d0b08dddaf032c460045387630beacd6e550b682fd", "sources": ["arxiv", "semantic_scholar"], "title": "Motion-Agent: A Conversational Framework for Human Motion Generation with LLMs", "abstract": "While previous approaches to 3D human motion generation have achieved notable success, they often rely on extensive training and are limited to specific tasks. To address these challenges, we introduce Motion-Agent, an efficient conversational framework designed for general human motion generation, editing, and understanding. Motion-Agent employs an open-source pre-trained language model to develop a generative agent, MotionLLM, that bridges the gap between motion and text. This is accomplished by encoding and quantizing motions into discrete tokens that align with the language model's vocabulary. With only 1--3\\% of the model's parameters fine-tuned using adapters, MotionLLM delivers performance on par with diffusion models and other transformer-based methods trained from scratch. By integrating MotionLLM with GPT-4 without additional training, Motion-Agent is able to generate highly complex motion sequences through multi-turn conversations, a capability that previous models have struggled to achieve. Motion-Agent supports a wide range of motion-language tasks, offering versatile capabilities for generating and customizing human motion through interactive conversational exchanges. Project page: https://knoxzhao.github.io/Motion-Agent", "authors": ["Qi Wu", "Yubo Zhao", "Yifan Wang", "Xinhang Liu", "Yu-Wing Tai", "Chi-Keung Tang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-27", "url": "https://arxiv.org/abs/2405.17013", "pdf_url": "https://arxiv.org/pdf/2405.17013v3", "arxiv_id": "2405.17013", "doi": null, "citation_count": 43, "influential_citation_count": 6, "has_code": true, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.4225}
{"id": "bd1d3a615d3bd442106075e0b7043883bad0852cb3be8d90f6dcb985eb792616", "sources": ["arxiv", "semantic_scholar"], "title": "Planning with Multi-Constraints via Collaborative Language Agents", "abstract": "The rapid advancement of neural language models has sparked a new surge of intelligent agent research. Unlike traditional agents, large language model-based agents (LLM agents) have emerged as a promising paradigm for achieving artificial general intelligence (AGI) due to their superior reasoning and generalization capabilities. Effective planning is crucial for the success of LLM agents in real-world tasks, making it a highly pursued topic in the community. Current planning methods typically translate tasks into executable action sequences. However, determining a feasible or optimal sequence for complex tasks with multiple constraints at fine granularity, which often requires compositing long chains of heterogeneous actions, remains challenging. This paper introduces Planning with Multi-Constraints (PMC), a zero-shot methodology for collaborative LLM-based multi-agent systems that simplifies complex task planning with constraints by decomposing it into a hierarchy of subordinate tasks. Each subtask is then mapped into executable actions. PMC was assessed on two constraint-intensive benchmarks, TravelPlanner and API-Bank. Notably, PMC achieved an average 42.68% success rate on TravelPlanner, significantly higher than GPT-4 (2.92%), and outperforming GPT-4 with ReAct on API-Bank by 13.64%, showing the immense potential of integrating LLM with multi-agent systems. We also show that PMC works with small LLM as the planning core, e.g., LLaMA-3.1-8B.", "authors": ["Cong Zhang", "Derrick Goh Xin Deik", "Dexun Li", "Hao Zhang", "Yong Liu"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-26", "url": "https://arxiv.org/abs/2405.16510", "pdf_url": "https://arxiv.org/pdf/2405.16510v4", "arxiv_id": "2405.16510", "doi": null, "citation_count": 23, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Computational Linguistics", "quality_score": 0.3451}
{"id": "35a36c2e3810d3414b9f509d0adee44dd294733ae5bdea26a15cca3549f709c3", "sources": ["arxiv", "semantic_scholar"], "title": "Variational Offline Multi-agent Skill Discovery", "abstract": "Skills are effective temporal abstractions established for sequential decision making, which enable efficient hierarchical learning for long-horizon tasks and facilitate multi-task learning through their transferability. Despite extensive research, research gaps remain in multi-agent scenarios, particularly for automatically extracting subgroup coordination patterns in a multi-agent task. In this case, we propose two novel auto-encoder schemes: VO-MASD-3D and VO-MASD-Hier, to simultaneously capture subgroup- and temporal-level abstractions and form multi-agent skills, which firstly solves the aforementioned challenge. An essential algorithm component of these schemes is a dynamic grouping function that can automatically detect latent subgroups based on agent interactions in a task. Further, our method can be applied to offline multi-task data, and the discovered subgroup skills can be transferred across relevant tasks without retraining. Empirical evaluations on StarCraft tasks indicate that our approach significantly outperforms existing hierarchical multi-agent reinforcement learning (MARL) methods. Moreover, skills discovered using our method can effectively reduce the learning difficulty in MARL scenarios with delayed and sparse reward signals. The codebase is available at https://github.com/LucasCJYSDL/VOMASD.", "authors": ["Jiayu Chen", "Tian Lan", "Vaneet Aggarwal"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-26", "url": "https://arxiv.org/abs/2405.16386", "pdf_url": "https://arxiv.org/pdf/2405.16386v3", "arxiv_id": "2405.16386", "doi": "10.48550/arXiv.2405.16386", "citation_count": 7, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/LucasCJYSDL/VOMASD", "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.2258}
{"id": "050ed016d8797c9ee8f2484128e583af81c15c5499d9af6ef2a6ff9f25c6d62b", "sources": ["arxiv", "semantic_scholar"], "title": "STRIDE: A Tool-Assisted LLM Agent Framework for Strategic and Interactive Decision-Making", "abstract": "Large Language Models (LLMs) like GPT-4 have revolutionized natural language processing, showing remarkable linguistic proficiency and reasoning capabilities. However, their application in strategic multi-agent decision-making environments is hampered by significant limitations including poor mathematical reasoning, difficulty in following instructions, and a tendency to generate incorrect information. These deficiencies hinder their performance in strategic and interactive tasks that demand adherence to nuanced game rules, long-term planning, exploration in unknown environments, and anticipation of opponents' moves. To overcome these obstacles, this paper presents a novel LLM agent framework equipped with memory and specialized tools to enhance their strategic decision-making capabilities. We deploy the tools in a number of economically important environments, in particular bilateral bargaining and multi-agent and dynamic mechanism design. We employ quantitative metrics to assess the framework's performance in various strategic decision-making problems. Our findings establish that our enhanced framework significantly improves the strategic decision-making capability of LLMs. While we highlight the inherent limitations of current LLM models, we demonstrate the improvements through targeted enhancements, suggesting a promising direction for future developments in LLM applications for interactive environments.", "authors": ["Chuanhao Li", "Runhan Yang", "Tiankai Li", "Milad Bafarassat", "Kourosh Sharifi", "Dirk Bergemann", "Zhuoran Yang"], "categories": ["cs.CL", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-25", "url": "https://arxiv.org/abs/2405.16376", "pdf_url": "https://arxiv.org/pdf/2405.16376v2", "arxiv_id": "2405.16376", "doi": "10.48550/arXiv.2405.16376", "citation_count": 22, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3404}
{"id": "71a7f368c3a3c580707b8753ab24fa9904cc4eff1fb35965c76a4a23f1fa88a5", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Efficient LLM Grounding for Embodied Multi-Agent Collaboration", "abstract": "Grounding the reasoning ability of large language models (LLMs) for embodied tasks is challenging due to the complexity of the physical world. Especially, LLM planning for multi-agent collaboration requires communication of agents or credit assignment as the feedback to re-adjust the proposed plans and achieve effective coordination. However, existing methods that overly rely on physical verification or self-reflection suffer from excessive and inefficient querying of LLMs. In this paper, we propose a novel framework for multi-agent collaboration that introduces Reinforced Advantage feedback (ReAd) for efficient self-refinement of plans. Specifically, we perform critic regression to learn a sequential advantage function from LLM-planned data, and then treat the LLM planner as an optimizer to generate actions that maximize the advantage function. It endows the LLM with the foresight to discern whether the action contributes to accomplishing the final task. We provide theoretical analysis by extending advantage-weighted regression in reinforcement learning to multi-agent systems. Experiments on Overcooked-AI and a difficult variant of RoCoBench show that ReAd surpasses baselines in success rate, and also significantly decreases the interaction steps of agents and query rounds of LLMs, demonstrating its high efficiency for grounding LLMs. More results are given at https://embodied-read.github.io", "authors": ["Yang Zhang", "Shixin Yang", "Chenjia Bai", "Fei Wu", "Xiu Li", "Zhen Wang", "Xuelong Li"], "categories": ["cs.AI", "cs.CL", "cs.LG", "cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-23", "url": "https://arxiv.org/abs/2405.14314", "pdf_url": "https://arxiv.org/pdf/2405.14314v4", "arxiv_id": "2405.14314", "doi": "10.48550/arXiv.2405.14314", "citation_count": 58, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4427}
{"id": "f3c9fbc89f6089e900abc64d76e2ff52cc7021d1ee7b45e0e223755104c24127", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Prosocial Irrationality for LLM Agents: A Social Cognition View", "abstract": "Large language models (LLMs) have been shown to face hallucination issues due to the data they trained on often containing human bias; whether this is reflected in the decision-making process of LLM Agents remains under-explored. As LLM Agents are increasingly employed in intricate social environments, a pressing and natural question emerges: Can we utilize LLM Agents' systematic hallucinations to mirror human cognitive biases, thus exhibiting irrational social intelligence? In this paper, we probe the irrational behavior among contemporary LLM Agents by melding practical social science experiments with theoretical insights. Specifically, We propose CogMir, an open-ended Multi-LLM Agents framework that utilizes hallucination properties to assess and enhance LLM Agents' social intelligence through cognitive biases. Experimental results on CogMir subsets show that LLM Agents and humans exhibit high consistency in irrational and prosocial decision-making under uncertain conditions, underscoring the prosociality of LLM Agents as social entities and highlighting the significance of hallucination properties. Additionally, the CogMir framework demonstrates its potential as a valuable platform for encouraging more research into the social intelligence of LLM Agents.", "authors": ["Xuan Liu", "Jie Zhang", "Haoyang Shang", "Song Guo", "Chengxu Yang", "Quanyan Zhu"], "categories": ["cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-23", "url": "https://arxiv.org/abs/2405.14744", "pdf_url": "https://arxiv.org/pdf/2405.14744v5", "arxiv_id": "2405.14744", "doi": "10.48550/arXiv.2405.14744", "citation_count": 36, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.3921}
{"id": "ac9d294f96cac82e618a17f8644b87d2ede988e31d00a7270b0a0fb0f0f6cb86", "sources": ["arxiv", "semantic_scholar"], "title": "On the Brittle Foundations of ReAct Prompting for Agentic Large Language Models", "abstract": "The reasoning abilities of Large Language Models (LLMs) remain a topic of debate. Some methods such as ReAct-based prompting, have gained popularity for claiming to enhance sequential decision-making abilities of agentic LLMs. However, it is unclear what is the source of improvement in LLM reasoning with ReAct based prompting. In this paper we examine these claims of ReAct based prompting in improving agentic LLMs for sequential decision-making. By introducing systematic variations to the input prompt we perform a sensitivity analysis along the claims of ReAct and find that the performance is minimally influenced by the \"interleaving reasoning trace with action execution\" or the content of the generated reasoning traces in ReAct, contrary to original claims and common usage. Instead, the performance of LLMs is driven by the similarity between input example tasks and queries, implicitly forcing the prompt designer to provide instance-specific examples which significantly increases the cognitive burden on the human. Our investigation shows that the perceived reasoning abilities of LLMs stem from the exemplar-query similarity and approximate retrieval rather than any inherent reasoning abilities.", "authors": ["Mudit Verma", "Siddhant Bhambri", "Subbarao Kambhampati"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-22", "url": "https://arxiv.org/abs/2405.13966", "pdf_url": "https://arxiv.org/pdf/2405.13966v1", "arxiv_id": "2405.13966", "doi": "10.48550/arXiv.2405.13966", "citation_count": 23, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3451}
{"id": "bc3a35eee47f2f5fc11e14350a8c63b1e93b521e3c30548b6864c4bad52de242", "sources": ["arxiv", "semantic_scholar"], "title": "Human-Centered LLM-Agent User Interface: A Position Paper", "abstract": "Large Language Model (LLM) -in-the-loop applications have been shown to effectively interpret the human user's commands, make plans, and operate external tools/systems accordingly. Still, the operation scope of the LLM agent is limited to passively following the user, requiring the user to frame his/her needs with regard to the underlying tools/systems. We note that the potential of an LLM-Agent User Interface (LAUI) is much greater. A user mostly ignorant to the underlying tools/systems should be able to work with a LAUI to discover an emergent workflow. Contrary to the conventional way of designing an explorable GUI to teach the user a predefined set of ways to use the system, in the ideal LAUI, the LLM agent is initialized to be proficient with the system, proactively studies the user and his/her needs, and proposes new interaction schemes to the user. To illustrate LAUI, we present Flute X GPT, a concrete example using an LLM agent, a prompt manager, and a flute-tutoring multi-modal software-hardware system to facilitate the complex, real-time user experience of learning to play the flute.", "authors": ["Daniel Chin", "Yuxuan Wang", "Gus Xia"], "categories": ["cs.HC", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-19", "url": "https://arxiv.org/abs/2405.13050", "pdf_url": "https://arxiv.org/pdf/2405.13050v2", "arxiv_id": "2405.13050", "doi": "10.48550/arXiv.2405.13050", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "b401b08fadcd60dd215c75d58558bae6b3428fa0c298d62573ca19e9d0f0e653", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative Multi-agent Approach for Automated Computer Game Testing", "abstract": "Automated testing of computer games is a challenging problem, especially when lengthy scenarios have to be tested. Automating such a scenario boils down to finding the right sequence of interactions given an abstract description of the scenario. Recent works have shown that an agent-based approach works well for the purpose, e.g. due to agents' reactivity, hence enabling a test agent to immediately react to game events and changing state. Many games nowadays are multi-player. This opens up an interesting possibility to deploy multiple cooperative test agents to test such a game, for example to speed up the execution of multiple testing tasks. This paper offers a cooperative multi-agent testing approach and a study of its performance based on a case study on a 3D game called Lab Recruits.", "authors": ["Samira Shirzadeh-hajimahmood", "I. S. W. B. Prasteya", "Mehdi Dastani", "Frank Dignum"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-18", "url": "https://arxiv.org/abs/2405.11347", "pdf_url": "https://arxiv.org/pdf/2405.11347v1", "arxiv_id": "2405.11347", "doi": "10.48550/arXiv.2405.11347", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Workshop on Engineering Multi-Agent Systems", "quality_score": 0.0}
{"id": "0fd211d725d542684c0e8c3ea9cfd7fbb9fc7e21969975e4007f30a186de2257", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-based Multi-Agent Reinforcement Learning: Current and Future Directions", "abstract": "In recent years, Large Language Models (LLMs) have shown great abilities in various tasks, including question answering, arithmetic problem solving, and poem writing, among others. Although research on LLM-as-an-agent has shown that LLM can be applied to Reinforcement Learning (RL) and achieve decent results, the extension of LLM-based RL to Multi-Agent System (MAS) is not trivial, as many aspects, such as coordination and communication between agents, are not considered in the RL frameworks of a single agent. To inspire more research on LLM-based MARL, in this letter, we survey the existing LLM-based single-agent and multi-agent RL frameworks and provide potential research directions for future research. In particular, we focus on the cooperative tasks of multiple agents with a common goal and communication among them. We also consider human-in/on-the-loop scenarios enabled by the language component in the framework.", "authors": ["Chuanneng Sun", "Songjun Huang", "Dario Pompili"], "categories": ["cs.MA", "cs.AI", "cs.CL", "cs.LG", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-17", "url": "https://arxiv.org/abs/2405.11106", "pdf_url": "https://arxiv.org/pdf/2405.11106v1", "arxiv_id": "2405.11106", "doi": "10.48550/arXiv.2405.11106", "citation_count": 82, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4798}
{"id": "6732a2640a11d91f979428826f498f4da1e1119c289ec94b47471bcc10e509cf", "sources": ["arxiv", "semantic_scholar"], "title": "MuMath-Code: Combining Tool-Use Large Language Models with Multi-perspective Data Augmentation for Mathematical Reasoning", "abstract": "The tool-use Large Language Models (LLMs) that integrate with external Python interpreters have significantly enhanced mathematical reasoning capabilities for open-source LLMs, while tool-free methods chose another track: augmenting math reasoning data. However, a great method to integrate the above two research paths and combine their advantages remains to be explored. In this work, we firstly include new math questions via multi-perspective data augmenting methods and then synthesize code-nested solutions to them. The open LLMs (i.e., Llama-2) are finetuned on the augmented dataset to get the resulting models, MuMath-Code ($μ$-Math-Code). During the inference phase, our MuMath-Code generates code and interacts with the external python interpreter to get the execution results. Therefore, MuMath-Code leverages the advantages of both the external tool and data augmentation. To fully leverage the advantages of our augmented data, we propose a two-stage training strategy: In Stage-1, we finetune Llama-2 on pure CoT data to get an intermediate model, which then is trained on the code-nested data in Stage-2 to get the resulting MuMath-Code. Our MuMath-Code-7B achieves 83.8 on GSM8K and 52.4 on MATH, while MuMath-Code-70B model achieves new state-of-the-art performance among open methods -- achieving 90.7% on GSM8K and 55.1% on MATH. Extensive experiments validate the combination of tool use and data augmentation, as well as our two-stage training strategy. We release the proposed dataset along with the associated code for public use.", "authors": ["Shuo Yin", "Weihao You", "Zhilong Ji", "Guoqiang Zhong", "Jinfeng Bai"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-13", "url": "https://arxiv.org/abs/2405.07551", "pdf_url": "https://arxiv.org/pdf/2405.07551v1", "arxiv_id": "2405.07551", "doi": "10.48550/arXiv.2405.07551", "citation_count": 30, "influential_citation_count": 1, "has_code": true, "code_url": null, "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.3728}
{"id": "48f846a6048a2d47ec2a646d25fcdd5e87a6bda894667105c9e2708cca56788d", "sources": ["arxiv", "semantic_scholar"], "title": "Smurfs: Multi-Agent System using Context-Efficient DFSDT for Tool Planning", "abstract": "Teaching large language models (LLMs) to use tools for solving complex problems can grant them human-like reasoning abilities. ReAct and its variants are popular frameworks for tool use in both single-agent and multi-agent systems. To address issues like error propagation and limited exploration in ReAct, the Deep First Search Decision Tree (DFSDT) was proposed, but it faces challenges such as rollback instability, redundant context, and premature termination in single-agent settings. We introduce \"Smurfs,\" a novel multi-agent system (MAS) that enhances DFSDT with a modular, context-efficient, and training-free design. Smurfs surpasses baseline methods in both the open-ended StableToolBench and the closed-ended HotpotQA tasks, reducing token usage by 60.9\\% compared to DFSDT and enabling Mistral-7b to perform on par with GPT-4-DFSDT. Extensive ablation studies confirm the effectiveness of Smurfs' core components, offering valuable insights for the construction and interpretation of MAS, and paving the way for future exploration.", "authors": ["Junzhi Chen", "Juhao Liang", "Benyou Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-09", "url": "https://arxiv.org/abs/2405.05955", "pdf_url": "https://arxiv.org/pdf/2405.05955v4", "arxiv_id": "2405.05955", "doi": "10.18653/v1/2025.naacl-long.169", "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.2698}
{"id": "01198f3b79bf9fc0d933b3d93cf923b33a6065445873c058f697b9ca7d5b74fd", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM-Tool Compiler for Fused Parallel Function Calling", "abstract": "State-of-the-art sequential reasoning in Large Language Models (LLMs) has expanded the capabilities of Copilots beyond conversational tasks to complex function calling, managing thousands of API calls. However, the tendency of compositional prompting to segment tasks into multiple steps, each requiring a round-trip to the GPT APIs, leads to increased system latency and costs. Although recent advancements in parallel function calling have improved tool execution per API call, they may necessitate more detailed in-context instructions and task breakdown at the prompt level, resulting in higher engineering and production costs. Inspired by the hardware design principles of multiply-add (MAD) operations, which fuse multiple arithmetic operations into a single task from the compiler's perspective, we propose LLM-Tool Compiler, which selectively fuses similar types of tool operations under a single function at runtime, presenting them as a unified task to the LLM. This selective fusion inherently enhances parallelization and efficiency. Benchmarked on a large-scale Copilot platform, LLM-Tool Compiler achieves up to four times more parallel calls than existing methods, reducing token costs and latency by up to 40% and 12%, respectively.", "authors": ["Simranjit Singh", "Andreas Karatzas", "Michael Fore", "Iraklis Anagnostopoulos", "Dimitrios Stamoulis"], "categories": ["cs.PL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-07", "url": "https://arxiv.org/abs/2405.17438", "pdf_url": "https://arxiv.org/pdf/2405.17438v1", "arxiv_id": "2405.17438", "doi": "10.48550/arXiv.2405.17438", "citation_count": 18, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3197}
{"id": "139f2dc54c07a3bb011b38d4263ca38570eb1f55be556083e4a97f19f810ee04", "sources": ["arxiv", "semantic_scholar"], "title": "Persona Inconstancy in Multi-Agent LLM Collaboration: Conformity, Confabulation, and Impersonation", "abstract": "Multi-agent AI systems can be used for simulating collective decision-making in scientific and practical applications. They can also be used to introduce a diverse group discussion step in chatbot pipelines, enhancing the cultural sensitivity of the chatbot's responses. These applications, however, are predicated on the ability of AI agents to reliably adopt assigned personas and mimic human interactions. To see whether LLM agents satisfy these requirements, we examine AI agent ensembles engaged in cross-national collaboration and debate by analyzing their private responses and chat transcripts. Our findings suggest that multi-agent discussions can support collective AI decisions that more often reflect diverse perspectives, yet this effect is tempered by the agents' susceptibility to conformity due to perceived peer pressure and occasional challenges in maintaining consistent personas and opinions. Instructions that encourage debate in support of one's opinions rather than collaboration increase the rate of inconstancy. Without addressing the factors we identify, the full potential of multi-agent frameworks for producing more culturally diverse AI outputs or more realistic simulations of group decision-making may remain untapped.", "authors": ["Razan Baltaji", "Babak Hemmatian", "Lav R. Varshney"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-06", "url": "https://arxiv.org/abs/2405.03862", "pdf_url": "https://arxiv.org/pdf/2405.03862v3", "arxiv_id": "2405.03862", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "The 2nd Workshop on Cross-Cultural Considerations in NLP (2024)", "quality_score": 0.1747}
{"id": "bc7e053787cc4439698545e31e5b8e835735bdcaa1c262f0cef82f7daa6cee2e", "sources": ["arxiv", "semantic_scholar"], "title": "Language Evolution for Evading Social Media Regulation via LLM-based Multi-agent Simulation", "abstract": "Social media platforms such as Twitter, Reddit, and Sina Weibo play a crucial role in global communication but often encounter strict regulations in geopolitically sensitive regions. This situation has prompted users to ingeniously modify their way of communicating, frequently resorting to coded language in these regulated social media environments. This shift in communication is not merely a strategy to counteract regulation, but a vivid manifestation of language evolution, demonstrating how language naturally evolves under societal and technological pressures. Studying the evolution of language in regulated social media contexts is of significant importance for ensuring freedom of speech, optimizing content moderation, and advancing linguistic research. This paper proposes a multi-agent simulation framework using Large Language Models (LLMs) to explore the evolution of user language in regulated social media environments. The framework employs LLM-driven agents: supervisory agent who enforce dialogue supervision and participant agents who evolve their language strategies while engaging in conversation, simulating the evolution of communication styles under strict regulations aimed at evading social media regulation. The study evaluates the framework's effectiveness through a range of scenarios from abstract scenarios to real-world situations. Key findings indicate that LLMs are capable of simulating nuanced language dynamics and interactions in constrained settings, showing improvement in both evading supervision and information accuracy as evolution progresses. Furthermore, it was found that LLM agents adopt different strategies for different scenarios.", "authors": ["Jinyu Cai", "Jialong Li", "Mingyue Zhang", "Munan Li", "Chen-Shu Wang", "Kenji Tei"], "categories": ["cs.SI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-05", "url": "https://arxiv.org/abs/2405.02858", "pdf_url": "https://arxiv.org/pdf/2405.02858v1", "arxiv_id": "2405.02858", "doi": "10.1109/CEC60901.2024.10612015", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Congress on Evolutionary Computation", "quality_score": 0.2785}
{"id": "035c7c4679f4cce31b152ec3860b7e41ab3291eb08c553622476ddef1ba64968", "sources": ["arxiv", "semantic_scholar"], "title": "SocialGFs: Learning Social Gradient Fields for Multi-Agent Reinforcement Learning", "abstract": "Multi-agent systems (MAS) need to adaptively cope with dynamic environments, changing agent populations, and diverse tasks. However, most of the multi-agent systems cannot easily handle them, due to the complexity of the state and task space. The social impact theory regards the complex influencing factors as forces acting on an agent, emanating from the environment, other agents, and the agent's intrinsic motivation, referring to the social force. Inspired by this concept, we propose a novel gradient-based state representation for multi-agent reinforcement learning. To non-trivially model the social forces, we further introduce a data-driven method, where we employ denoising score matching to learn the social gradient fields (SocialGFs) from offline samples, e.g., the attractive or repulsive outcomes of each force. During interactions, the agents take actions based on the multi-dimensional gradients to maximize their own rewards. In practice, we integrate SocialGFs into the widely used multi-agent reinforcement learning algorithms, e.g., MAPPO. The empirical results reveal that SocialGFs offer four advantages for multi-agent systems: 1) they can be learned without requiring online interaction, 2) they demonstrate transferability across diverse tasks, 3) they facilitate credit assignment in challenging reward settings, and 4) they are scalable with the increasing number of agents.", "authors": ["Qian Long", "Fangwei Zhong", "Mingdong Wu", "Yizhou Wang", "Song-Chun Zhu"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-03", "url": "https://arxiv.org/abs/2405.01839", "pdf_url": "https://arxiv.org/pdf/2405.01839v1", "arxiv_id": "2405.01839", "doi": "10.48550/arXiv.2405.01839", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "9351a8d76b70331868b4ff7f3bbd2c03aa3d9838f25d242c1f866e27e495b7d5", "sources": ["arxiv", "semantic_scholar"], "title": "Communication-Efficient Training Workload Balancing for Decentralized Multi-Agent Learning", "abstract": "Decentralized Multi-agent Learning (DML) enables collaborative model training while preserving data privacy. However, inherent heterogeneity in agents' resources (computation, communication, and task size) may lead to substantial variations in training time. This heterogeneity creates a bottleneck, lengthening the overall training time due to straggler effects and potentially wasting spare resources of faster agents. To minimize training time in heterogeneous environments, we present a Communication-Efficient Training Workload Balancing for Decentralized Multi-Agent Learning (ComDML), which balances the workload among agents through a decentralized approach. Leveraging local-loss split training, ComDML enables parallel updates, where slower agents offload part of their workload to faster agents. To minimize the overall training time, ComDML optimizes the workload balancing by jointly considering the communication and computation capacities of agents, which hinges upon integer programming. A dynamic decentralized pairing scheduler is developed to efficiently pair agents and determine optimal offloading amounts. We prove that in ComDML, both slower and faster agents' models converge, for convex and non-convex functions. Furthermore, extensive experimental results on popular datasets (CIFAR-10, CIFAR-100, and CINIC-10) and their non-I.I.D. variants, with large models such as ResNet-56 and ResNet-110, demonstrate that ComDML can significantly reduce the overall training time while maintaining model accuracy, compared to state-of-the-art methods. ComDML demonstrates robustness in heterogeneous environments, and privacy measures can be seamlessly integrated for enhanced data protection.", "authors": ["Seyed Mahmoud Sajjadi Mohammadabadi", "Lei Yang", "Feng Yan", "Junshan Zhang"], "categories": ["cs.LG", "cs.AI", "cs.DC", "cs.MA", "cs.PF"], "fields_of_study": ["Computer Science"], "published_date": "2024-05-01", "url": "https://arxiv.org/abs/2405.00839", "pdf_url": "https://arxiv.org/pdf/2405.00839v1", "arxiv_id": "2405.00839", "doi": "10.1109/ICDCS60910.2024.00069", "citation_count": 19, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Distributed Computing Systems", "quality_score": 0.3253}
{"id": "1cb3215d39da741c415f169f29fdaff7782c853f38bedfb61112cdee442724f1", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Synchronization Tasks", "abstract": "In multi-agent reinforcement learning (MARL), coordination plays a crucial role in enhancing agents' performance beyond what they could achieve through cooperation alone. The interdependence of agents' actions, coupled with the need for communication, leads to a domain where effective coordination is crucial. In this paper, we introduce and define $\\textit{Multi-Agent Synchronization Tasks}$ (MSTs), a novel subset of multi-agent tasks. We describe one MST, that we call $\\textit{Synchronized Predator-Prey}$, offering a detailed description that will serve as the basis for evaluating a selection of recent state-of-the-art (SOTA) MARL algorithms explicitly designed to address coordination challenges through the use of communication strategies. Furthermore, we present empirical evidence that reveals the limitations of the algorithms assessed to solve MSTs, demonstrating their inability to scale effectively beyond 2-agent coordination tasks in scenarios where communication is a requisite component. Finally, the results raise questions about the applicability of recent SOTA approaches for complex coordination tasks (i.e. MSTs) and prompt further exploration into the underlying causes of their limitations in this context.", "authors": ["Rolando Fernandez", "Garrett Warnell", "Derrik E. Asher", "Peter Stone"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-29", "url": "https://arxiv.org/abs/2404.18798", "pdf_url": "https://arxiv.org/pdf/2404.18798v1", "arxiv_id": "2404.18798", "doi": "10.48550/arXiv.2404.18798", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "86511540054c9fd0653f141f61fe578df162c9674e58bc2a88714ef60833d872", "sources": ["arxiv", "semantic_scholar"], "title": "ComposerX: Multi-Agent Symbolic Music Composition with LLMs", "abstract": "Music composition represents the creative side of humanity, and itself is a complex task that requires abilities to understand and generate information with long dependency and harmony constraints. While demonstrating impressive capabilities in STEM subjects, current LLMs easily fail in this task, generating ill-written music even when equipped with modern techniques like In-Context-Learning and Chain-of-Thoughts. To further explore and enhance LLMs' potential in music composition by leveraging their reasoning ability and the large knowledge base in music history and theory, we propose ComposerX, an agent-based symbolic music generation framework. We find that applying a multi-agent approach significantly improves the music composition quality of GPT-4. The results demonstrate that ComposerX is capable of producing coherent polyphonic music compositions with captivating melodies, while adhering to user instructions.", "authors": ["Qixin Deng", "Qikai Yang", "Ruibin Yuan", "Yipeng Huang", "Yi Wang", "Xubo Liu", "Zeyue Tian", "Jiahao Pan", "Ge Zhang", "Hanfeng Lin", "Yizhi Li", "Yinghao Ma", "Jie Fu", "Chenghua Lin", "Emmanouil Benetos", "Wenwu Wang", "Guangyu Xia", "Wei Xue", "Yike Guo"], "categories": ["cs.SD", "cs.AI", "cs.CL", "cs.LG", "cs.MM", "eess.AS"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2024-04-28", "url": "https://arxiv.org/abs/2404.18081", "pdf_url": "https://arxiv.org/pdf/2404.18081v2", "arxiv_id": "2404.18081", "doi": "10.48550/arXiv.2404.18081", "citation_count": 60, "influential_citation_count": 5, "has_code": false, "code_url": null, "venue": "International Society for Music Information Retrieval Conference", "quality_score": 0.4463}
{"id": "ab6fc2da94db3318bf846fde124241ca2f41517f1a464551d0b473eab932d94d", "sources": ["arxiv", "semantic_scholar"], "title": "PLAYER*: Enhancing LLM-based Multi-Agent Communication and Interaction in Murder Mystery Games", "abstract": "We introduce WellPlay, a reasoning dataset for multi-agent conversational inference in Murder Mystery Games (MMGs). WellPlay comprises 1,482 inferential questions across 12 games, spanning objectives, reasoning, and relationship understanding, and establishes a systematic benchmark for evaluating agent reasoning abilities in complex social settings. Building on this foundation, we present PLAYER*, a novel framework for Large Language Model (LLM)-based agents in MMGs. MMGs pose unique challenges, including undefined state spaces, absent intermediate rewards, and the need for strategic reasoning through natural language. PLAYER* addresses these challenges with a sensor-based state representation and an information-driven strategy that optimises questioning and suspect pruning. Experiments show that PLAYER* outperforms existing methods in reasoning accuracy, efficiency, and agent-human interaction, advancing reasoning agents for complex social scenarios.", "authors": ["Qinglin Zhu", "Runcong Zhao", "Bin Liang", "Jinhua Du", "Lin Gui", "Yulan He"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-26", "url": "https://arxiv.org/abs/2404.17662", "pdf_url": "https://arxiv.org/pdf/2404.17662v5", "arxiv_id": "2404.17662", "doi": "10.48550/arXiv.2404.17662", "citation_count": 14, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.294}
{"id": "629f923fe9001f8a04a107b73815b67b62c790052f131464a16c5f8ba1462722", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperate or Collapse: Emergence of Sustainable Cooperation in a Society of LLM Agents", "abstract": "As AI systems pervade human life, ensuring that large language models (LLMs) make safe decisions remains a significant challenge. We introduce the Governance of the Commons Simulation (GovSim), a generative simulation platform designed to study strategic interactions and cooperative decision-making in LLMs. In GovSim, a society of AI agents must collectively balance exploiting a common resource with sustaining it for future use. This environment enables the study of how ethical considerations, strategic planning, and negotiation skills impact cooperative outcomes. We develop an LLM-based agent architecture and test it with the leading open and closed LLMs. We find that all but the most powerful LLM agents fail to achieve a sustainable equilibrium in GovSim, with the highest survival rate below 54%. Ablations reveal that successful multi-agent communication between agents is critical for achieving cooperation in these cases. Furthermore, our analyses show that the failure to achieve sustainable cooperation in most LLMs stems from their inability to formulate and analyze hypotheses about the long-term effects of their actions on the equilibrium of the group. Finally, we show that agents that leverage \"Universalization\"-based reasoning, a theory of moral thinking, are able to achieve significantly better sustainability. Taken together, GovSim enables us to study the mechanisms that underlie sustainable self-government with specificity and scale. We open source the full suite of our research results, including the simulation environment, agent prompts, and a comprehensive web interface.", "authors": ["Giorgio Piatti", "Zhijing Jin", "Max Kleiman-Weiner", "Bernhard Schölkopf", "Mrinmaya Sachan", "Rada Mihalcea"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-25", "url": "https://arxiv.org/abs/2404.16698", "pdf_url": "https://arxiv.org/pdf/2404.16698v4", "arxiv_id": "2404.16698", "doi": "10.52202/079017-3548", "citation_count": 101, "influential_citation_count": 7, "has_code": true, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.5022}
{"id": "2f88c48bd12ac90e3ead7dae1c484223764aa6093d6e2b819378e8c85a3e8472", "sources": ["arxiv", "semantic_scholar"], "title": "AutoGenesisAgent: Self-Generating Multi-Agent Systems for Complex Tasks", "abstract": "The proliferation of large language models (LLMs) and their integration into multi-agent systems has paved the way for sophisticated automation in various domains. This paper introduces AutoGenesisAgent, a multi-agent system that autonomously designs and deploys other multi-agent systems tailored for specific tasks. AutoGenesisAgent comprises several specialized agents including System Understanding, System Design, Agent Generator, and several others that collectively manage the lifecycle of creating functional multi-agent systems from initial concept to deployment. Each agent in AutoGenesisAgent has distinct responsibilities ranging from interpreting input prompts to optimizing system performance, culminating, in the deployment of a ready-to-use system. This proof-of-concept study discusses the design, implementation, and lessons learned from developing AutoGenesisAgent, highlighting its capability to generate and refine multi-agent systems autonomously, thereby reducing the need for extensive human oversight in the initial stages of system design. Keywords: multi-agent systems, large language models, system design automation, agent architecture, autonomous systems, software deployment", "authors": ["Jeremy Harper"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-25", "url": "https://arxiv.org/abs/2404.17017", "pdf_url": "https://arxiv.org/pdf/2404.17017v1", "arxiv_id": "2404.17017", "doi": "10.48550/arXiv.2404.17017", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1945}
{"id": "6fea4ace0dbfc99b24476656c321713e1a0e21c1d14c35bccbed451ddf4a0516", "sources": ["arxiv", "semantic_scholar"], "title": "Neural Interaction Energy for Multi-Agent Trajectory Prediction", "abstract": "Maintaining temporal stability is crucial in multi-agent trajectory prediction. Insufficient regularization to uphold this stability often results in fluctuations in kinematic states, leading to inconsistent predictions and the amplification of errors. In this study, we introduce a framework called Multi-Agent Trajectory prediction via neural interaction Energy (MATE). This framework assesses the interactive motion of agents by employing neural interaction energy, which captures the dynamics of interactions and illustrates their influence on the future trajectories of agents. To bolster temporal stability, we introduce two constraints: inter-agent interaction constraint and intra-agent motion constraint. These constraints work together to ensure temporal stability at both the system and agent levels, effectively mitigating prediction fluctuations inherent in multi-agent systems. Comparative evaluations against previous methods on four diverse datasets highlight the superior prediction accuracy and generalization capabilities of our model.", "authors": ["Kaixin Shen", "Ruijie Quan", "Linchao Zhu", "Jun Xiao", "Yi Yang"], "categories": ["cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-25", "url": "https://arxiv.org/abs/2404.16579", "pdf_url": "https://arxiv.org/pdf/2404.16579v1", "arxiv_id": "2404.16579", "doi": "10.1145/3664647.3680792", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "ACM Multimedia", "quality_score": 0.2113}
{"id": "d67719deb92ca65b18abcf971c0787c6c5103133514ca848876e94909eb588e1", "sources": ["arxiv", "semantic_scholar"], "title": "ClinicalAgent: Clinical Trial Multi-Agent System with Large Language Model-based Reasoning", "abstract": "Large Language Models (LLMs) and multi-agent systems have shown impressive capabilities in natural language tasks but face challenges in clinical trial applications, primarily due to limited access to external knowledge. Recognizing the potential of advanced clinical trial tools that aggregate and predict based on the latest medical data, we propose an integrated solution to enhance their accessibility and utility. We introduce Clinical Agent System (ClinicalAgent), a clinical multi-agent system designed for clinical trial tasks, leveraging GPT-4, multi-agent architectures, LEAST-TO-MOST, and ReAct reasoning technology. This integration not only boosts LLM performance in clinical contexts but also introduces novel functionalities. The proposed method achieves competitive predictive performance in clinical trial outcome prediction (0.7908 PR-AUC), obtaining a 0.3326 improvement over the standard prompt Method. Publicly available code can be found at https://anonymous.4open.science/r/ClinicalAgent-6671.", "authors": ["Ling Yue", "Sixue Xing", "Jintai Chen", "Tianfan Fu"], "categories": ["cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-23", "url": "https://arxiv.org/abs/2404.14777", "pdf_url": "https://arxiv.org/pdf/2404.14777v2", "arxiv_id": "2404.14777", "doi": "10.1145/3698587.3701359", "citation_count": 43, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "ACM International Conference on Bioinformatics, Computational Biology and Biomedicine", "quality_score": 0.4109}
{"id": "06a6391ae25f1b0a778912930e11f56e2413e5b96c9cb2802cdb37d0297bc6c0", "sources": ["arxiv", "semantic_scholar"], "title": "Private Agent-Based Modeling", "abstract": "The practical utility of agent-based models in decision-making relies on their capacity to accurately replicate populations while seamlessly integrating real-world data streams. Yet, the incorporation of such data poses significant challenges due to privacy concerns. To address this issue, we introduce a paradigm for private agent-based modeling wherein the simulation, calibration, and analysis of agent-based models can be achieved without centralizing the agents attributes or interactions. The key insight is to leverage techniques from secure multi-party computation to design protocols for decentralized computation in agent-based models. This ensures the confidentiality of the simulated agents without compromising on simulation accuracy. We showcase our protocols on a case study with an epidemiological simulation comprising over 150,000 agents. We believe this is a critical step towards deploying agent-based models to real-world applications.", "authors": ["Ayush Chopra", "Arnau Quera-Bofarull", "Nurullah Giray-Kuru", "Michael Wooldridge", "Ramesh Raskar"], "categories": ["cs.MA", "cs.CR", "cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-19", "url": "https://arxiv.org/abs/2404.12983", "pdf_url": "https://arxiv.org/pdf/2404.12983v1", "arxiv_id": "2404.12983", "doi": "10.5555/3635637.3662887", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1505}
{"id": "ed6471a6bba9608298b95b6c8d7a4d60fce226f05fa789d49c3922807a62ded0", "sources": ["arxiv", "semantic_scholar"], "title": "AgentCoord: Visually Exploring Coordination Strategy for LLM-based Multi-Agent Collaboration", "abstract": "The potential of automatic task-solving through Large Language Model (LLM)-based multi-agent collaboration has recently garnered widespread attention from both the research community and industry. While utilizing natural language to coordinate multiple agents presents a promising avenue for democratizing agent technology for general users, designing coordination strategies remains challenging with existing coordination frameworks. This difficulty stems from the inherent ambiguity of natural language for specifying the collaboration process and the significant cognitive effort required to extract crucial information (e.g. agent relationship, task dependency, result correspondence) from a vast amount of text-form content during exploration. In this work, we present a visual exploration framework to facilitate the design of coordination strategies in multi-agent collaboration. We first establish a structured representation for LLM-based multi-agent coordination strategy to regularize the ambiguity of natural language. Based on this structure, we devise a three-stage generation method that leverages LLMs to convert a user's general goal into an executable initial coordination strategy. Users can further intervene at any stage of the generation process, utilizing LLMs and a set of interactions to explore alternative strategies. Whenever a satisfactory strategy is identified, users can commence the collaboration and examine the visually enhanced execution result. We develop AgentCoord, a prototype interactive system, and conduct a formal user study to demonstrate the feasibility and effectiveness of our approach.", "authors": ["Bo Pan", "Jiaying Lu", "Ke Wang", "Li Zheng", "Zhen Wen", "Yingchaojie Feng", "Minfeng Zhu", "Wei Chen"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-18", "url": "https://arxiv.org/abs/2404.11943", "pdf_url": "https://arxiv.org/pdf/2404.11943v1", "arxiv_id": "2404.11943", "doi": "10.48550/arXiv.2404.11943", "citation_count": 32, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Computers & graphics", "quality_score": 0.3796}
{"id": "6ca0593bbf27f21e405852bfac5b3f82262449591660a90eda8df2941d2231b8", "sources": ["arxiv", "semantic_scholar"], "title": "The Landscape of Emerging AI Agent Architectures for Reasoning, Planning, and Tool Calling: A Survey", "abstract": "This survey paper examines the recent advancements in AI agent implementations, with a focus on their ability to achieve complex goals that require enhanced reasoning, planning, and tool execution capabilities. The primary objectives of this work are to a) communicate the current capabilities and limitations of existing AI agent implementations, b) share insights gained from our observations of these systems in action, and c) suggest important considerations for future developments in AI agent design. We achieve this by providing overviews of single-agent and multi-agent architectures, identifying key patterns and divergences in design choices, and evaluating their overall impact on accomplishing a provided goal. Our contribution outlines key themes when selecting an agentic architecture, the impact of leadership on agent systems, agent communication styles, and key phases for planning, execution, and reflection that enable robust AI agent systems.", "authors": ["Tula Masterman", "Sandi Besen", "Mason Sawtell", "Alex Chao"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-17", "url": "https://arxiv.org/abs/2404.11584", "pdf_url": "https://arxiv.org/pdf/2404.11584v1", "arxiv_id": "2404.11584", "doi": "10.48550/arXiv.2404.11584", "citation_count": 209, "influential_citation_count": 8, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5806}
{"id": "c7b718b58ac6aa262ff36ab24db92291eb0d5579c170023a90fa015d777ce71d", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent eXperimenter (MAX)", "abstract": "We present a novel multi-agent simulator named Multi-Agent eXperimenter (MAX) that is designed to simulate blockchain experiments involving large numbers of agents of different types acting in one or several environments. The architecture of MAX is highly modular, enabling easy addition of new models.", "authors": ["Önder Gürcan"], "categories": ["cs.MA", "cs.AI", "cs.DC"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-12", "url": "https://arxiv.org/abs/2404.08398", "pdf_url": "https://arxiv.org/pdf/2404.08398v1", "arxiv_id": "2404.08398", "doi": "10.48550/arXiv.2404.08398", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "95585cfcffdd51d026cc761662af46da4640d8189ed9c38ba453ea23b8350fb9", "sources": ["arxiv", "semantic_scholar"], "title": "ODA: Observation-Driven Agent for integrating LLMs and Knowledge Graphs", "abstract": "The integration of Large Language Models (LLMs) and knowledge graphs (KGs) has achieved remarkable success in various natural language processing tasks. However, existing methodologies that integrate LLMs and KGs often navigate the task-solving process solely based on the LLM's analysis of the question, overlooking the rich cognitive potential inherent in the vast knowledge encapsulated in KGs. To address this, we introduce Observation-Driven Agent (ODA), a novel AI agent framework tailored for tasks involving KGs. ODA incorporates KG reasoning abilities via global observation, which enhances reasoning capabilities through a cyclical paradigm of observation, action, and reflection. Confronting the exponential explosion of knowledge during observation, we innovatively design a recursive observation mechanism. Subsequently, we integrate the observed knowledge into the action and reflection modules. Through extensive experiments, ODA demonstrates state-of-the-art performance on several datasets, notably achieving accuracy improvements of 12.87% and 8.9%.", "authors": ["Lei Sun", "Zhengwei Tao", "Youdi Li", "Hiroshi Arakawa"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-11", "url": "https://arxiv.org/abs/2404.07677", "pdf_url": "https://arxiv.org/pdf/2404.07677v2", "arxiv_id": "2404.07677", "doi": "10.48550/arXiv.2404.07677", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.2865}
{"id": "e2d6418989c22574489cf7403bd2dabfb17ba4ad59d7aeac51189cd147d324d0", "sources": ["arxiv", "semantic_scholar"], "title": "LeapFrog: The Rowhammer Instruction Skip Attack", "abstract": "Since its inception, Rowhammer exploits have rapidly evolved into increasingly sophisticated threats compromising data integrity and the control flow integrity of victim processes. Nevertheless, it remains a challenge for an attacker to identify vulnerable targets (i.e., Rowhammer gadgets), understand the outcome of the attempted fault, and formulate an attack that yields useful results. In this paper, we present a new type of Rowhammer gadget, called a LeapFrog gadget, which, when present in the victim code, allows an adversary to subvert code execution to bypass a critical piece of code (e.g., authentication check logic, encryption rounds, padding in security protocols). The LeapFrog gadget manifests when the victim code stores the Program Counter (PC) value in the user or kernel stack (e.g., a return address during a function call) which, when tampered with, repositions the return address to a location that bypasses a security-critical code pattern. This research also presents a systematic process to identify LeapFrog gadgets. This methodology enables the automated detection of susceptible targets and the determination of optimal attack parameters. We first show the attack on a decision tree algorithm to show the potential implications. Secondly, we employ the attack on OpenSSL to bypass the encryption and reveal the plaintext. We then use our tools to scan the Open Quantum Safe library and report on the number of LeapFrog gadgets in the code. Lastly, we demonstrate this new attack vector through a practical demonstration in a client/server TLS handshake scenario, successfully inducing an instruction skip in a client application. Our findings extend the impact of Rowhammer attacks on control flow and contribute to developing more robust defenses against these increasingly sophisticated threats.", "authors": ["Andrew Adiletta", "M. Caner Tol", "Kemal Derya", "Berk Sunar", "Saad Islam"], "categories": ["cs.CR", "cs.AR"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-11", "url": "https://arxiv.org/abs/2404.07878", "pdf_url": "https://arxiv.org/pdf/2404.07878v3", "arxiv_id": "2404.07878", "doi": "10.1109/EuroSP63326.2025.00065", "citation_count": 7, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "European Symposium on Security and Privacy", "quality_score": 0.2258}
{"id": "59c4f60e22fa3b7478ab0d1594bc5c5fb9a25bc2c9036311457dfee85ee60ce5", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Based Multi-Agent Systems for Software Engineering: Literature Review, Vision and the Road Ahead", "abstract": "Integrating Large Language Models (LLMs) into autonomous agents marks a significant shift in the research landscape by offering cognitive abilities that are competitive with human planning and reasoning. This paper explores the transformative potential of integrating Large Language Models into Multi-Agent (LMA) systems for addressing complex challenges in software engineering (SE). By leveraging the collaborative and specialized abilities of multiple agents, LMA systems enable autonomous problem-solving, improve robustness, and provide scalable solutions for managing the complexity of real-world software projects. In this paper, we conduct a systematic review of recent primary studies to map the current landscape of LMA applications across various stages of the software development lifecycle (SDLC). To illustrate current capabilities and limitations, we perform two case studies to demonstrate the effectiveness of state-of-the-art LMA frameworks. Additionally, we identify critical research gaps and propose a comprehensive research agenda focused on enhancing individual agent capabilities and optimizing agent synergy. Our work outlines a forward-looking vision for developing fully autonomous, scalable, and trustworthy LMA systems, laying the foundation for the evolution of Software Engineering 2.0.", "authors": ["Junda He", "Christoph Treude", "David Lo"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-07", "url": "https://arxiv.org/abs/2404.04834", "pdf_url": "https://arxiv.org/pdf/2404.04834v4", "arxiv_id": "2404.04834", "doi": "10.1145/3712003", "citation_count": 228, "influential_citation_count": 9, "has_code": false, "code_url": null, "venue": "ACM Transactions on Software Engineering and Methodology", "quality_score": 0.59}
{"id": "fdeb050322e01aab42fe6e6531f3218bb59b4c23be66ba7619346a5f862c0c62", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Organized Agents: A LLM Multi-Agent Framework toward Ultra Large-Scale Code Generation and Optimization", "abstract": "Recent advancements in automatic code generation using large language model (LLM) agent have brought us closer to the future of automated software development. However, existing single-agent approaches face limitations in generating and improving large-scale, complex codebases due to constraints in context length. To tackle this challenge, we propose Self-Organized multi-Agent framework (SoA), a novel multi-agent framework that enables the scalable and efficient generation and optimization of large-scale code. In SoA, self-organized agents operate independently to generate and modify code components while seamlessly collaborating to construct the overall codebase. A key feature of our framework is the automatic multiplication of agents based on problem complexity, allowing for dynamic scalability. This enables the overall code volume to be increased indefinitely according to the number of agents, while the amount of code managed by each agent remains constant. We evaluate SoA on the HumanEval benchmark and demonstrate that, compared to a single-agent system, each agent in SoA handles significantly less code, yet the overall generated code is substantially greater. Moreover, SoA surpasses the powerful single-agent baseline by 5% in terms of Pass@1 accuracy.", "authors": ["Yoichi Ishibashi", "Yoshimasa Nishimura"], "categories": ["cs.SE", "cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-02", "url": "https://arxiv.org/abs/2404.02183", "pdf_url": "https://arxiv.org/pdf/2404.02183v1", "arxiv_id": "2404.02183", "doi": "10.48550/arXiv.2404.02183", "citation_count": 100, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5011}
{"id": "2242431c77d9b476b6694885306dd277bbbd051496e8510ccc353461279d1665", "sources": ["arxiv", "semantic_scholar"], "title": "CMAT: A Multi-Agent Collaboration Tuning Framework for Enhancing Small Language Models", "abstract": "Open large language models (LLMs) have significantly advanced the field of natural language processing, showcasing impressive performance across various tasks.Despite the significant advancements in LLMs, their effective operation still relies heavily on human input to accurately guide the dialogue flow, with agent tuning being a crucial optimization technique that involves human adjustments to the model for better response to such guidance.Addressing this dependency, our work introduces the TinyAgent model, trained on a meticulously curated high-quality dataset. We also present the Collaborative Multi-Agent Tuning (CMAT) framework, an innovative system designed to augment language agent capabilities through adaptive weight updates based on environmental feedback. This framework fosters collaborative learning and real-time adaptation among multiple intelligent agents, enhancing their context-awareness and long-term memory. In this research, we propose a new communication agent framework that integrates multi-agent systems with environmental feedback mechanisms, offering a scalable method to explore cooperative behaviors. Notably, our TinyAgent-7B model exhibits performance on par with GPT-3.5, despite having fewer parameters, signifying a substantial improvement in the efficiency and effectiveness of LLMs.", "authors": ["Xuechen Liang", "Yangfan He", "Meiling Tao", "Yinghui Xia", "Jianhui Wang", "Tianyu Shi", "Jun Wang", "JingSong Yang"], "categories": ["cs.CL", "cs.AI", "cs.CC"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-02", "url": "https://arxiv.org/abs/2404.01663", "pdf_url": "https://arxiv.org/pdf/2404.01663v6", "arxiv_id": "2404.01663", "doi": "10.48550/arXiv.2404.01663", "citation_count": 29, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3693}
{"id": "a7b4ee063043f3903d06266a674a882e9d98ac0e8e7eaa972393ac421d90252e", "sources": ["arxiv", "semantic_scholar"], "title": "GOV-REK: Governed Reward Engineering Kernels for Designing Robust Multi-Agent Reinforcement Learning Systems", "abstract": "For multi-agent reinforcement learning systems (MARLS), the problem formulation generally involves investing massive reward engineering effort specific to a given problem. However, this effort often cannot be translated to other problems; worse, it gets wasted when system dynamics change drastically. This problem is further exacerbated in sparse reward scenarios, where a meaningful heuristic can assist in the policy convergence task. We propose GOVerned Reward Engineering Kernels (GOV-REK), which dynamically assign reward distributions to agents in MARLS during its learning stage. We also introduce governance kernels, which exploit the underlying structure in either state or joint action space for assigning meaningful agent reward distributions. During the agent learning stage, it iteratively explores different reward distribution configurations with a Hyperband-like algorithm to learn ideal agent reward models in a problem-agnostic manner. Our experiments demonstrate that our meaningful reward priors robustly jumpstart the learning process for effectively learning different MARL problems.", "authors": ["Ashish Rana", "Michael Oesterle", "Jannik Brinkmann"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-01", "url": "https://arxiv.org/abs/2404.01131", "pdf_url": "https://arxiv.org/pdf/2404.01131v2", "arxiv_id": "2404.01131", "doi": "10.5555/3635637.3663183", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.0}
{"id": "61f83acb5b4617bc63fc4529afe9e5231d134a197d23c4417218d1ba4f1dcddd", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring LLM Multi-Agents for ICD Coding", "abstract": "To address the limitations of Large Language Models (LLMs) in the International Classification of Diseases (ICD) coding task, where they often produce inaccurate and incomplete prediction results due to the high-dimensional and skewed distribution of the ICD codes, and often lack interpretability and reliability as well. We introduce an innovative multi-agent approach for ICD coding which mimics the ICD coding assignment procedure in real-world settings, comprising five distinct agents: the patient, physician, coder, reviewer, and adjuster. Each agent utilizes an LLM-based model tailored to their specific role within the coding process. We also integrate the system with Electronic Health Record (HER)'s SOAP (subjective, objective, assessment and plan) structure to boost the performances. We compare our method with a system of agents designed solely by LLMs and other strong baselines and evaluate it using the Medical Information Mart for Intensive Care III (MIMIC-III) dataset. Our multi-agent coding framework significantly outperforms Zero-shot Chain of Thought (CoT) prompting and self-consistency with CoT (CoT-SC) in coding common and rare ICD codes. An ablation study validates the effectiveness of the designated agent roles. it also outperforms the LLM-designed agent system. Moreover, our method achieves comparable results to state-of-the-art ICD coding methods that require extensive pre-training or fine-tuning, and outperforms them in rare code accuracy, and explainability. Additionally, we demonstrate the method's practical applicability by presenting its performance in scenarios not limited by the common or rare ICD code constraints.The proposed multi-agent method for ICD coding effectively mimics the real-world coding process and improves performance on both common and rare codes.", "authors": ["Rumeng Li", "Xun Wang", "Hong Yu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-01", "url": "https://arxiv.org/abs/2406.15363", "pdf_url": "https://arxiv.org/pdf/2406.15363v2", "arxiv_id": "2406.15363", "doi": "10.48550/arXiv.2406.15363", "citation_count": 19, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "c01268d9c8d1601fefa711acbd3ae396aa40c6d1704ca58826392a62f7c78d0d", "sources": ["arxiv", "semantic_scholar"], "title": "Large Language Model Evaluation Via Multi AI Agents: Preliminary results", "abstract": "As Large Language Models (LLMs) have become integral to both research and daily operations, rigorous evaluation is crucial. This assessment is important not only for individual tasks but also for understanding their societal impact and potential risks. Despite extensive efforts to examine LLMs from various perspectives, there is a noticeable lack of multi-agent AI models specifically designed to evaluate the performance of different LLMs. To address this gap, we introduce a novel multi-agent AI model that aims to assess and compare the performance of various LLMs. Our model consists of eight distinct AI agents, each responsible for retrieving code based on a common description from different advanced language models, including GPT-3.5, GPT-3.5 Turbo, GPT-4, GPT-4 Turbo, Google Bard, LLAMA, and Hugging Face. Our developed model utilizes the API of each language model to retrieve code for a given high-level description. Additionally, we developed a verification agent, tasked with the critical role of evaluating the code generated by its counterparts. We integrate the HumanEval benchmark into our verification agent to assess the generated code's performance, providing insights into their respective capabilities and efficiencies. Our initial results indicate that the GPT-3.5 Turbo model's performance is comparatively better than the other models. This preliminary analysis serves as a benchmark, comparing their performances side by side. Our future goal is to enhance the evaluation process by incorporating the Massively Multitask Benchmark for Python (MBPP) benchmark, which is expected to further refine our assessment. Additionally, we plan to share our developed model with twenty practitioners from various backgrounds to test our model and collect their feedback for further improvement.", "authors": ["Zeeshan Rasheed", "Muhammad Waseem", "Kari Systä", "Pekka Abrahamsson"], "categories": ["cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2024-04-01", "url": "https://arxiv.org/abs/2404.01023", "pdf_url": "https://arxiv.org/pdf/2404.01023v1", "arxiv_id": "2404.01023", "doi": "10.48550/arXiv.2404.01023", "citation_count": 22, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3404}
{"id": "88c85e86df02783654a3c2eb03f9ae67fc71d41dee5ac90d03cb5ad638e2216a", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing the General Agent Capabilities of Low-Parameter LLMs through Tuning and Multi-Branch Reasoning", "abstract": "Open-source pre-trained Large Language Models (LLMs) exhibit strong language understanding and generation capabilities, making them highly successful in a variety of tasks. However, when used as agents for dealing with complex problems in the real world, their performance is far inferior to large commercial models such as ChatGPT and GPT-4. As intelligent agents, LLMs need to have the capabilities of task planning, long-term memory, and the ability to leverage external tools to achieve satisfactory performance. Various methods have been proposed to enhance the agent capabilities of LLMs. On the one hand, methods involve constructing agent-specific data and fine-tuning the models. On the other hand, some methods focus on designing prompts that effectively activate the reasoning abilities of the LLMs. We explore both strategies on the 7B and 13B models. We propose a comprehensive method for constructing agent-specific data using GPT-4. Through supervised fine-tuning with constructed data, we find that for these models with a relatively small number of parameters, supervised fine-tuning can significantly reduce hallucination outputs and formatting errors in agent tasks. Furthermore, techniques such as multi-path reasoning and task decomposition can effectively decrease problem complexity and enhance the performance of LLMs as agents. We evaluate our method on five agent tasks of AgentBench and achieve satisfactory results.", "authors": ["Qinhao Zhou", "Zihan Zhang", "Xiang Xiang", "Ke Wang", "Yuchuan Wu", "Yongbin Li"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-29", "url": "https://arxiv.org/abs/2403.19962", "pdf_url": "https://arxiv.org/pdf/2403.19962v1", "arxiv_id": "2403.19962", "doi": "10.48550/arXiv.2403.19962", "citation_count": 16, "influential_citation_count": 0, "has_code": true, "code_url": null, "venue": null, "quality_score": 0.3076}
{"id": "c83bbb99452705ac57d12ce259cbcba991b794c91240d30c0e90a5b9e2c314e3", "sources": ["arxiv", "semantic_scholar"], "title": "Energy-Optimal Multi-Agent Navigation as a Strategic-Form Game", "abstract": "This extended abstracts presents a method to generate energy-optimal trajectories for multi-agent systems as a strategic-form game. Using recent results in optimal control, we demonstrate that an energy-optimal trajectory can be generated in milliseconds if the sequence of constraint activations is known a priori. Thus, rather than selecting an infinite-dimensional action from a function space, the agents select their actions from a finite number of constraints and determine the time that each becomes active. Furthermore, the agents can exactly encode their trajectory in a set of real numbers, rather than communicating their control action as an infinite-dimensional function. We demonstrate the performance of this algorithm in simulation and find an optimal trajectory in 45 milliseconds on a tablet PC.", "authors": ["Logan Beaver"], "categories": ["math.OC"], "fields_of_study": ["Mathematics"], "published_date": "2024-03-28", "url": "https://arxiv.org/abs/2403.19641", "pdf_url": "https://arxiv.org/pdf/2403.19641v1", "arxiv_id": "2403.19641", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0}
{"id": "14eef64ead36957455fad29bc9042a28d5d897916d6de9a96aaa731b37efbda2", "sources": ["arxiv", "semantic_scholar"], "title": "Enhancing Anomaly Detection in Financial Markets with an LLM-based Multi-Agent Framework", "abstract": "This paper introduces a Large Language Model (LLM)-based multi-agent framework designed to enhance anomaly detection within financial market data, tackling the longstanding challenge of manually verifying system-generated anomaly alerts. The framework harnesses a collaborative network of AI agents, each specialised in distinct functions including data conversion, expert analysis via web research, institutional knowledge utilization or cross-checking and report consolidation and management roles. By coordinating these agents towards a common objective, the framework provides a comprehensive and automated approach for validating and interpreting financial data anomalies. I analyse the S&P 500 index to demonstrate the framework's proficiency in enhancing the efficiency, accuracy and reduction of human intervention in financial market monitoring. The integration of AI's autonomous functionalities with established analytical methods not only underscores the framework's effectiveness in anomaly detection but also signals its broader applicability in supporting financial market monitoring.", "authors": ["Taejin Park"], "categories": ["q-fin.RM"], "fields_of_study": ["Economics"], "published_date": "2024-03-28", "url": "https://arxiv.org/abs/2403.19735", "pdf_url": "https://arxiv.org/pdf/2403.19735v1", "arxiv_id": "2403.19735", "doi": null, "citation_count": 33, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3829}
{"id": "54ac0d41ac4a568d52a7e35c95b01543638d3eab1d9ddb885965a6b184fc6233", "sources": ["arxiv", "semantic_scholar"], "title": "MAGIS: LLM-Based Multi-Agent Framework for GitHub Issue Resolution", "abstract": "In software development, resolving the emergent issues within GitHub repositories is a complex challenge that involves not only the incorporation of new code but also the maintenance of existing code. Large Language Models (LLMs) have shown promise in code generation but face difficulties in resolving Github issues, particularly at the repository level. To overcome this challenge, we empirically study the reason why LLMs fail to resolve GitHub issues and analyze the major factors. Motivated by the empirical findings, we propose a novel LLM-based Multi-Agent framework for GitHub Issue reSolution, MAGIS, consisting of four agents customized for software evolution: Manager, Repository Custodian, Developer, and Quality Assurance Engineer agents. This framework leverages the collaboration of various agents in the planning and coding process to unlock the potential of LLMs to resolve GitHub issues. In experiments, we employ the SWE-bench benchmark to compare MAGIS with popular LLMs, including GPT-3.5, GPT-4, and Claude-2. MAGIS can resolve 13.94% GitHub issues, significantly outperforming the baselines. Specifically, MAGIS achieves an eight-fold increase in resolved ratio over the direct application of GPT-4, the advanced LLM.", "authors": ["Wei Tao", "Yucheng Zhou", "Yanlin Wang", "Wenqiang Zhang", "Hongyu Zhang", "Yu Cheng"], "categories": ["cs.SE", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-26", "url": "https://arxiv.org/abs/2403.17927", "pdf_url": "https://arxiv.org/pdf/2403.17927v2", "arxiv_id": "2403.17927", "doi": "10.48550/arXiv.2403.17927", "citation_count": 161, "influential_citation_count": 11, "has_code": false, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.5524}
{"id": "e7f8023cee9462acbd6b785a1fa6396be8746785a3aaade984d1e0d61dbb054a", "sources": ["arxiv", "semantic_scholar"], "title": "AgentStudio: A Toolkit for Building General Virtual Agents", "abstract": "General virtual agents need to handle multimodal observations, master complex action spaces, and self-improve in dynamic, open-domain environments. However, existing environments are often domain-specific and require complex setups, which limits agent development and evaluation in real-world settings. As a result, current evaluations lack in-depth analyses that decompose fundamental agent capabilities. We introduce AgentStudio, a trinity of environments, tools, and benchmarks to address these issues. AgentStudio provides a lightweight, interactive environment with highly generic observation and action spaces, e.g., video observations and GUI/API actions. It integrates tools for creating online benchmark tasks, annotating GUI elements, and labeling actions in videos. Based on our environment and tools, we curate an online task suite that benchmarks both GUI interactions and function calling with efficient auto-evaluation. We also reorganize existing datasets and collect new ones using our tools to establish three datasets: GroundUI, IDMBench, and CriticBench. These datasets evaluate fundamental agent abilities, including GUI grounding, learning from videos, and success detection, pointing to the desiderata for robust, general, and open-ended virtual agents.", "authors": ["Longtao Zheng", "Zhiyuan Huang", "Zhenghai Xue", "Xinrun Wang", "Bo An", "Shuicheng Yan"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-26", "url": "https://arxiv.org/abs/2403.17918", "pdf_url": "https://arxiv.org/pdf/2403.17918v3", "arxiv_id": "2403.17918", "doi": "10.48550/arXiv.2403.17918", "citation_count": 43, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Learning Representations", "quality_score": 0.4109}
{"id": "a1b4770198ee87141a94cbf01b8455e132b04850fe6697113967ab46b65c53b7", "sources": ["arxiv", "semantic_scholar"], "title": "TwoStep: Multi-agent Task Planning using Classical Planners and Large Language Models", "abstract": "Classical planning formulations like the Planning Domain Definition Language (PDDL) admit action sequences guaranteed to achieve a goal state given an initial state if any are possible. However, reasoning problems defined in PDDL do not capture temporal aspects of action taking, such as concurrent actions between two agents when there are no conflicting conditions, without significant modification and definition to existing PDDL domains. A human expert aware of such constraints can decompose a goal into subgoals, each reachable through single agent planning, to take advantage of simultaneous actions. In contrast to classical planning, large language models (LLMs) directly used for inferring plan steps rarely guarantee execution success, but are capable of leveraging commonsense reasoning to assemble action sequences. We combine the strengths of both classical planning and LLMs by approximating human intuitions for multi-agent planning goal decomposition. We demonstrate that LLM-based goal decomposition leads to faster planning times than solving multi-agent PDDL problems directly while simultaneously achieving fewer plan execution steps than a single agent plan alone, as well as most multiagent plans, while guaranteeing execution success. Additionally, we find that LLM-based approximations of subgoals result in similar multi-agent execution lengths to those specified by human experts. Website and resources at https://glamor-usc.github.io/twostep", "authors": ["David Bai", "Ishika Singh", "David Traum", "Jesse Thomason"], "categories": ["cs.AI", "cs.CL", "cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-25", "url": "https://arxiv.org/abs/2403.17246", "pdf_url": "https://arxiv.org/pdf/2403.17246v2", "arxiv_id": "2403.17246", "doi": "10.48550/arXiv.2403.17246", "citation_count": 39, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4005}
{"id": "f69945cf93476db7be11bc9aa357371e04ffab800e4a45ffb5cbc4e39e313f12", "sources": ["arxiv", "semantic_scholar"], "title": "Do LLM Agents Have Regret? A Case Study in Online Learning and Games", "abstract": "Large language models (LLMs) have been increasingly employed for (interactive) decision-making, via the development of LLM-based autonomous agents. Despite their emerging successes, the performance of LLM agents in decision-making has not been fully investigated through quantitative metrics, especially in the multi-agent setting when they interact with each other, a typical scenario in real-world LLM-agent applications. To better understand the limits of LLM agents in these interactive environments, we propose to study their interactions in benchmark decision-making settings in online learning and game theory, through the performance metric of \\emph{regret}. We first empirically study the {no-regret} behaviors of LLMs in canonical (non-stationary) online learning problems, as well as the emergence of equilibria when LLM agents interact through playing repeated games. We then provide some theoretical insights into the no-regret behaviors of LLM agents, under certain assumptions on the supervised pre-training and the rationality model of human decision-makers who generate the data. Notably, we also identify (simple) cases where advanced LLMs such as GPT-4 fail to be no-regret. To promote the no-regret behaviors, we propose a novel \\emph{unsupervised} training loss of \\emph{regret-loss}, which, in contrast to the supervised pre-training loss, does not require the labels of (optimal) actions. We then establish the statistical guarantee of generalization bound for regret-loss minimization, followed by the optimization guarantee that minimizing such a loss may automatically lead to known no-regret learning algorithms. Our further experiments demonstrate the effectiveness of our regret-loss, especially in addressing the above ``regrettable'' cases.", "authors": ["Chanwoo Park", "Xiangyu Liu", "Asuman Ozdaglar", "Kaiqing Zhang"], "categories": ["cs.LG", "cs.AI", "cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-25", "url": "https://arxiv.org/abs/2403.16843", "pdf_url": "https://arxiv.org/pdf/2403.16843v5", "arxiv_id": "2403.16843", "doi": "10.48550/arXiv.2403.16843", "citation_count": 46, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.418}
{"id": "69fa68372e25edc2b31ac36c53623e641c85c0831077983531f3fcba6c99360e", "sources": ["arxiv", "semantic_scholar"], "title": "Content Knowledge Identification with Multi-Agent Large Language Models (LLMs)", "abstract": "Teachers' mathematical content knowledge (CK) is of vital importance and need in teacher professional development (PD) programs. Computer-aided asynchronous PD systems are the most recent proposed PD techniques, which aim to help teachers improve their PD equally with fewer concerns about costs and limitations of time or location. However, current automatic CK identification methods, which serve as one of the core techniques of asynchronous PD systems, face challenges such as diversity of user responses, scarcity of high-quality annotated data, and low interpretability of the predictions. To tackle these challenges, we propose a Multi-Agent LLMs-based framework, LLMAgent-CK, to assess the user responses' coverage of identified CK learning goals without human annotations. By taking advantage of multi-agent LLMs in strong generalization ability and human-like discussions, our proposed LLMAgent-CK presents promising CK identifying performance on a real-world mathematical CK dataset MaCKT. Moreover, our case studies further demonstrate the working of the multi-agent framework.", "authors": ["Kaiqi Yang", "Yucheng Chu", "Taylor Darwin", "Ahreum Han", "Hang Li", "Hongzhi Wen", "Yasemin Copur-Gencturk", "Jiliang Tang", "Hui Liu"], "categories": ["cs.AI", "cs.CY"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-22", "url": "https://arxiv.org/abs/2404.07960", "pdf_url": "https://arxiv.org/pdf/2404.07960v1", "arxiv_id": "2404.07960", "doi": "10.1007/978-3-031-64299-9_23", "citation_count": 14, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Artificial Intelligence in Education", "quality_score": 0.294}
{"id": "734c9d2f6dd83c6e9ba09fbd48adabe9a023aea604ed42dea01c04a13958c36e", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent VQA: Exploring Multi-Agent Foundation Models in Zero-Shot Visual Question Answering", "abstract": "This work explores the zero-shot capabilities of foundation models in Visual Question Answering (VQA) tasks. We propose an adaptive multi-agent system, named Multi-Agent VQA, to overcome the limitations of foundation models in object detection and counting by using specialized agents as tools. Unlike existing approaches, our study focuses on the system's performance without fine-tuning it on specific VQA datasets, making it more practical and robust in the open world. We present preliminary experimental results under zero-shot scenarios and highlight some failure cases, offering new directions for future research.", "authors": ["Bowen Jiang", "Zhijun Zhuang", "Shreyas S. Shivakumar", "Dan Roth", "Camillo J. Taylor"], "categories": ["cs.CV", "cs.AI", "cs.CL", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-21", "url": "https://arxiv.org/abs/2403.14783", "pdf_url": "https://arxiv.org/pdf/2403.14783v1", "arxiv_id": "2403.14783", "doi": "10.48550/arXiv.2403.14783", "citation_count": 12, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/bowen-upenn/Multi-Agent-VQA", "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "9d53c4cd96450a60fd4981a86b7d418c52e9e43263207b4bd80d299e9179122f", "sources": ["arxiv", "semantic_scholar"], "title": "ReAct Meets ActRe: When Language Agents Enjoy Training Data Autonomy", "abstract": "Language agents have demonstrated autonomous decision-making abilities by reasoning with foundation models. Recently, efforts have been made to train language agents for performance improvement, with multi-step reasoning and action trajectories as the training data. However, collecting such trajectories still requires considerable human effort, by either artificial annotation or implementations of diverse prompting frameworks. In this work, we propose A$^3$T, a framework that enables the Autonomous Annotation of Agent Trajectories in the style of ReAct. The central role is an ActRe prompting agent, which explains the reason for an arbitrary action. When randomly sampling an external action, the ReAct-style agent could query the ActRe agent with the action to obtain its textual rationales. Novel trajectories are then synthesized by prepending the posterior reasoning from ActRe to the sampled action. In this way, the ReAct-style agent executes multiple trajectories for the failed tasks, and selects the successful ones to supplement its failed trajectory for contrastive self-training. Realized by policy gradient methods with binarized rewards, the contrastive self-training with accumulated trajectories facilitates a closed loop for multiple rounds of language agent self-improvement. We conduct experiments using QLoRA fine-tuning with the open-sourced Mistral-7B-Instruct-v0.2. In AlfWorld, the agent trained with A$^3$T obtains a 1-shot success rate of 96%, and 100% success with 4 iterative rounds. In WebShop, the 1-shot performance of the A$^3$T agent matches human average, and 4 rounds of iterative refinement lead to the performance approaching human experts. A$^3$T agents significantly outperform existing techniques, including prompting with GPT-4, advanced agent frameworks, and fully fine-tuned LLMs.", "authors": ["Zonghan Yang", "Peng Li", "Ming Yan", "Ji Zhang", "Fei Huang", "Yang Liu"], "categories": ["cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-21", "url": "https://arxiv.org/abs/2403.14589", "pdf_url": "https://arxiv.org/pdf/2403.14589v3", "arxiv_id": "2403.14589", "doi": "10.48550/arXiv.2403.14589", "citation_count": 19, "influential_citation_count": 3, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3253}
{"id": "553c6c0e24e2b2c8bc33dd7ac2aa59a00dabbafd883ce1eaeb88763818de0c47", "sources": ["arxiv", "semantic_scholar"], "title": "Motion Prediction of Multi-agent systems with Multi-view clustering", "abstract": "This paper presents a method for future motion prediction of multi-agent systems by including group formation information and future intent. Formation of groups depends on a physics-based clustering method that follows the agglomerative hierarchical clustering algorithm. We identify clusters that incorporate the minimum cost-to-go function of a relevant optimal control problem as a metric for clustering between the groups among agents, where groups with similar associated costs are assumed to be likely to move together. The cost metric accounts for proximity to other agents as well as the intended goal of each agent. An unscented Kalman filter based approach is used to update the established clusters as well as add new clusters when new information is obtained. Our approach is verified through non-trivial numerical simulations implementing the proposed algorithm on different datasets pertaining to a variety of scenarios and agents.", "authors": ["Anegi James", "Efstathios Bakolas"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-20", "url": "https://arxiv.org/abs/2403.13905", "pdf_url": "https://arxiv.org/pdf/2403.13905v1", "arxiv_id": "2403.13905", "doi": "10.48550/arXiv.2403.13905", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "efce57977d8591b4d09edaabae8782de6016591c05f237bfe15420caf0fc2ce7", "sources": ["arxiv", "semantic_scholar"], "title": "Embodied LLM Agents Learn to Cooperate in Organized Teams", "abstract": "Large Language Models (LLMs) have emerged as integral tools for reasoning, planning, and decision-making, drawing upon their extensive world knowledge and proficiency in language-related tasks. LLMs thus hold tremendous potential for natural language interaction within multi-agent systems to foster cooperation. However, LLM agents tend to over-report and comply with any instruction, which may result in information redundancy and confusion in multi-agent cooperation. Inspired by human organizations, this paper introduces a framework that imposes prompt-based organization structures on LLM agents to mitigate these problems. Through a series of experiments with embodied LLM agents and human-agent collaboration, our results highlight the impact of designated leadership on team efficiency, shedding light on the leadership qualities displayed by LLM agents and their spontaneous cooperative behaviors. Further, we harness the potential of LLMs to propose enhanced organizational prompts, via a Criticize-Reflect process, resulting in novel organization structures that reduce communication costs and enhance team efficiency.", "authors": ["Xudong Guo", "Kaixuan Huang", "Jiale Liu", "Wenhui Fan", "Natalia Vélez", "Qingyun Wu", "Huazheng Wang", "Thomas L. Griffiths", "Mengdi Wang"], "categories": ["cs.AI", "cs.CL", "cs.CY", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-19", "url": "https://arxiv.org/abs/2403.12482", "pdf_url": "https://arxiv.org/pdf/2403.12482v2", "arxiv_id": "2403.12482", "doi": "10.1109/TCSS.2025.3637527", "citation_count": 89, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Computational Social Systems", "quality_score": 0.4886}
{"id": "f435ee3c1a082ddd80da60cf2e0c43c510882699fea5a775ea078064289c668d", "sources": ["arxiv", "semantic_scholar"], "title": "Graph Neural Network-based Multi-agent Reinforcement Learning for Resilient Distributed Coordination of Multi-Robot Systems", "abstract": "Existing multi-agent coordination techniques are often fragile and vulnerable to anomalies such as agent attrition and communication disturbances, which are quite common in the real-world deployment of systems like field robotics. To better prepare these systems for the real world, we present a graph neural network (GNN)-based multi-agent reinforcement learning (MARL) method for resilient distributed coordination of a multi-robot system. Our method, Multi-Agent Graph Embedding-based Coordination (MAGEC), is trained using multi-agent proximal policy optimization (PPO) and enables distributed coordination around global objectives under agent attrition, partial observability, and limited or disturbed communications. We use a multi-robot patrolling scenario to demonstrate our MAGEC method in a ROS 2-based simulator and then compare its performance with prior coordination approaches. Results demonstrate that MAGEC outperforms existing methods in several experiments involving agent attrition and communication disturbance, and provides competitive results in scenarios without such anomalies.", "authors": ["Anthony Goeckner", "Yueyuan Sui", "Nicolas Martinet", "Xinliang Li", "Qi Zhu"], "categories": ["cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-19", "url": "https://arxiv.org/abs/2403.13093", "pdf_url": "https://arxiv.org/pdf/2403.13093v1", "arxiv_id": "2403.13093", "doi": "10.1109/IROS58592.2024.10802510", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.3138}
{"id": "ac961bb76b1a3ab236e0d8c2c714aea73cd4dc162f859b6722fe2dbf0927c7a5", "sources": ["arxiv", "semantic_scholar"], "title": "Hierarchical Auto-Organizing System for Open-Ended Multi-Agent Navigation", "abstract": "Due to the dynamic and unpredictable open-world setting, navigating complex environments in Minecraft poses significant challenges for multi-agent systems. Agents must interact with the environment and coordinate their actions with other agents to achieve common objectives. However, traditional approaches often struggle to efficiently manage inter-agent communication and task distribution, crucial for effective multi-agent navigation. Furthermore, processing and integrating multi-modal information (such as visual, textual, and auditory data) is essential for agents to comprehend their goals and navigate the environment successfully and fully. To address this issue, we design the HAS framework to auto-organize groups of LLM-based agents to complete navigation tasks. In our approach, we devise a hierarchical auto-organizing navigation system, which is characterized by 1) a hierarchical system for multi-agent organization, ensuring centralized planning and decentralized execution; 2) an auto-organizing and intra-communication mechanism, enabling dynamic group adjustment under subtasks; 3) a multi-modal information platform, facilitating multi-modal perception to perform the three navigation tasks with one system. To assess organizational behavior, we design a series of navigation tasks in the Minecraft environment, which includes searching and exploring. We aim to develop embodied organizations that push the boundaries of embodied AI, moving it towards a more human-like organizational structure.", "authors": ["Zhonghan Zhao", "Kewei Chen", "Dongxu Guo", "Wenhao Chai", "Tian Ye", "Yanting Zhang", "Gaoang Wang"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-13", "url": "https://arxiv.org/abs/2403.08282", "pdf_url": "https://arxiv.org/pdf/2403.08282v2", "arxiv_id": "2403.08282", "doi": "10.48550/arXiv.2403.08282", "citation_count": 30, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3728}
{"id": "41dc1072cf3cd78975cb7d78a91f25a2c59a34c178292ee70b752b420389a003", "sources": ["arxiv", "semantic_scholar"], "title": "Ensembling Prioritized Hybrid Policies for Multi-agent Pathfinding", "abstract": "Multi-Agent Reinforcement Learning (MARL) based Multi-Agent Path Finding (MAPF) has recently gained attention due to its efficiency and scalability. Several MARL-MAPF methods choose to use communication to enrich the information one agent can perceive. However, existing works still struggle in structured environments with high obstacle density and a high number of agents. To further improve the performance of the communication-based MARL-MAPF solvers, we propose a new method, Ensembling Prioritized Hybrid Policies (EPH). We first propose a selective communication block to gather richer information for better agent coordination within multi-agent environments and train the model with a Q learning-based algorithm. We further introduce three advanced inference strategies aimed at bolstering performance during the execution phase. First, we hybridize the neural policy with single-agent expert guidance for navigating conflict-free zones. Secondly, we propose Q value-based methods for prioritized resolution of conflicts as well as deadlock situations. Finally, we introduce a robust ensemble method that can efficiently collect the best out of multiple possible solutions. We empirically evaluate EPH in complex multi-agent environments and demonstrate competitive performance against state-of-the-art neural methods for MAPF. We open-source our code at https://github.com/ai4co/eph-mapf.", "authors": ["Huijie Tang", "Federico Berto", "Jinkyoo Park"], "categories": ["cs.MA", "cs.AI", "cs.LG", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-12", "url": "https://arxiv.org/abs/2403.07559", "pdf_url": "https://arxiv.org/pdf/2403.07559v2", "arxiv_id": "2403.07559", "doi": "10.1109/IROS58592.2024.10801914", "citation_count": 13, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/ai4co/eph-mapf", "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.2865}
{"id": "91c17b6eefce28420880b08f3b629b85a7c26b8de68dd4ce503106890daf1831", "sources": ["arxiv", "semantic_scholar"], "title": "Tapilot-Crossing: Benchmarking and Evolving LLMs Towards Interactive Data Analysis Agents", "abstract": "Interactive Data Analysis, the collaboration between humans and LLM agents, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic interactive logs for data analysis hinder the quantitative evaluation of Large Language Model (LLM) agents in this task. To mitigate this issue, we introduce Tapilot-Crossing, a new benchmark to evaluate LLM agents on interactive data analysis. Tapilot-Crossing contains 1024 interactions, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, Tapilot-Crossing is constructed by an economical multi-agent environment, Decision Company, with few human efforts. We evaluate popular and advanced LLM agents in Tapilot-Crossing, which underscores the challenges of interactive data analysis. Furthermore, we propose Adaptive Interaction Reflection (AIR), a self-generated reflection strategy that guides LLM agents to learn from successful history. Experiments demonstrate that Air can evolve LLMs into effective interactive data analysis agents, achieving a relative performance improvement of up to 44.5%.", "authors": ["Jinyang Li", "Nan Huo", "Yan Gao", "Jiayi Shi", "Yingxiu Zhao", "Ge Qu", "Yurong Wu", "Chenhao Ma", "Jian-Guang Lou", "Reynold Cheng"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-08", "url": "https://arxiv.org/abs/2403.05307", "pdf_url": "https://arxiv.org/pdf/2403.05307v1", "arxiv_id": "2403.05307", "doi": "10.48550/arXiv.2403.05307", "citation_count": 15, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "5283a6a773e86171308813f3804702205d14c0b4a04d0b9d37a6ca98e58f386e", "sources": ["arxiv", "semantic_scholar"], "title": "AutoDefense: Multi-Agent LLM Defense against Jailbreak Attacks", "abstract": "Despite extensive pre-training in moral alignment to prevent generating harmful information, large language models (LLMs) remain vulnerable to jailbreak attacks. In this paper, we propose AutoDefense, a multi-agent defense framework that filters harmful responses from LLMs. With the response-filtering mechanism, our framework is robust against different jailbreak attack prompts, and can be used to defend different victim models. AutoDefense assigns different roles to LLM agents and employs them to complete the defense task collaboratively. The division in tasks enhances the overall instruction-following of LLMs and enables the integration of other defense components as tools. With AutoDefense, small open-source LMs can serve as agents and defend larger models against jailbreak attacks. Our experiments show that AutoDefense can effectively defense against different jailbreak attacks, while maintaining the performance at normal user request. For example, we reduce the attack success rate on GPT-3.5 from 55.74% to 7.95% using LLaMA-2-13b with a 3-agent system. Our code and data are publicly available at https://github.com/XHMY/AutoDefense.", "authors": ["Yifan Zeng", "Yiran Wu", "Xiao Zhang", "Huazheng Wang", "Qingyun Wu"], "categories": ["cs.LG", "cs.CL", "cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-02", "url": "https://arxiv.org/abs/2403.04783", "pdf_url": "https://arxiv.org/pdf/2403.04783v2", "arxiv_id": "2403.04783", "doi": "10.48550/arXiv.2403.04783", "citation_count": 155, "influential_citation_count": 12, "has_code": true, "code_url": "https://github.com/XHMY/AutoDefense", "venue": "arXiv.org", "quality_score": 0.557}
{"id": "8cfbf22e5caf0015e6602e6c9b6315c95335e1c6c4c73ce23e28cf85bb16fed5", "sources": ["arxiv", "semantic_scholar"], "title": "Imitation Learning Datasets: A Toolkit For Creating Datasets, Training Agents and Benchmarking", "abstract": "Imitation learning field requires expert data to train agents in a task. Most often, this learning approach suffers from the absence of available data, which results in techniques being tested on its dataset. Creating datasets is a cumbersome process requiring researchers to train expert agents from scratch, record their interactions and test each benchmark method with newly created data. Moreover, creating new datasets for each new technique results in a lack of consistency in the evaluation process since each dataset can drastically vary in state and action distribution. In response, this work aims to address these issues by creating Imitation Learning Datasets, a toolkit that allows for: (i) curated expert policies with multithreaded support for faster dataset creation; (ii) readily available datasets and techniques with precise measurements; and (iii) sharing implementations of common imitation learning techniques. Demonstration link: https://nathangavenski.github.io/#/il-datasets-video", "authors": ["Nathan Gavenski", "Michael Luck", "Odinaldo Rodrigues"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-03-01", "url": "https://arxiv.org/abs/2403.00550", "pdf_url": "https://arxiv.org/pdf/2403.00550v1", "arxiv_id": "2403.00550", "doi": "10.48550/arXiv.2403.00550", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1505}
{"id": "401b14e20b51ee8f5ba3e5e5ae115935e1bba2a16544f584e91513d91522a1db", "sources": ["arxiv", "semantic_scholar"], "title": "Rethinking the Bounds of LLM Reasoning: Are Multi-Agent Discussions the Key?", "abstract": "Recent progress in LLMs discussion suggests that multi-agent discussion improves the reasoning abilities of LLMs. In this work, we reevaluate this claim through systematic experiments, where we propose a novel group discussion framework to enrich the set of discussion mechanisms. Interestingly, our results show that a single-agent LLM with strong prompts can achieve almost the same performance as the best existing discussion approach on a wide range of reasoning tasks and backbone LLMs. We observe that the multi-agent discussion performs better than a single agent only when there is no demonstration in the prompt. Further study reveals the common interaction mechanisms of LLMs during the discussion.", "authors": ["Qineng Wang", "Zihao Wang", "Ying Su", "Hanghang Tong", "Yangqiu Song"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-28", "url": "https://arxiv.org/abs/2402.18272", "pdf_url": "https://arxiv.org/pdf/2402.18272v1", "arxiv_id": "2402.18272", "doi": "10.48550/arXiv.2402.18272", "citation_count": 184, "influential_citation_count": 7, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.5668}
{"id": "28c0f4674f90e709e295b5ae9ed123e529019b7a4bbd9c5360ce1b1fafdd47da", "sources": ["arxiv", "semantic_scholar"], "title": "Navigating Complexity: Orchestrated Problem Solving with Multi-Agent LLMs", "abstract": "Large Language Models (LLMs) have demonstrated remarkable capabilities in solving various tasks, yet they often struggle with comprehensively addressing complex and vague problems. Existing approaches, including multi-agent LLM systems, offer solutions to certain challenges but still require manual setup and lack scalability. To address this gap, we propose a novel approach leveraging decomposition to enable LLMs to tackle vague problems effectively. Our approach involves an orchestrating LLM that interacts with users to understand the problem and then decomposes it into tangible sub-problems. Instead of expecting the LLM to solve the entire problem in one go, we train it to ask follow-up questions to gain a deeper understanding of the user's requirements. Once the problem is adequately understood, the orchestrating LLM divides it into smaller, manageable sub-problems. Each sub-problem is then assigned to specialized LLM agents or non-LLM functions for resolution. These agents work in parallel to solve their respective sub-problems, with the orchestrating LLM overseeing the process and compiling the solutions into a comprehensive answer for the user. By adopting this decomposition approach, we alleviate the constraints imposed by token limitations on LLM outputs and empower them to provide nuanced solutions to complex and ambiguous problems. Through our approach, we aim to enable LLMs to think and operate more like humans, breaking down complex problems into manageable parts and collaboratively solving them. This not only enhances the problem-solving capabilities of LLMs but also offers a scalable and efficient method for addressing a wide range of real-world challenges.", "authors": ["Sumedh Rasal", "E. J. Hauer"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-26", "url": "https://arxiv.org/abs/2402.16713", "pdf_url": "https://arxiv.org/pdf/2402.16713v2", "arxiv_id": "2402.16713", "doi": "10.48550/arXiv.2402.16713", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "74c4619a4b82e4112d95ada116cafa234a2dae4cf67680632854129b0312f64b", "sources": ["arxiv", "semantic_scholar"], "title": "LLMArena: Assessing Capabilities of Large Language Models in Dynamic Multi-Agent Environments", "abstract": "Recent advancements in large language models (LLMs) have revealed their potential for achieving autonomous agents possessing human-level intelligence. However, existing benchmarks for evaluating LLM Agents either use static datasets, potentially leading to data leakage or focus only on single-agent scenarios, overlooking the complexities of multi-agent interactions. There is a lack of a benchmark that evaluates the diverse capabilities of LLM agents in multi-agent, dynamic environments. To this end, we introduce LLMArena, a novel and easily extensible framework for evaluating the diverse capabilities of LLM in multi-agent dynamic environments. LLMArena encompasses seven distinct gaming environments, employing Trueskill scoring to assess crucial abilities in LLM agents, including spatial reasoning, strategic planning, numerical reasoning, risk assessment, communication, opponent modeling, and team collaboration. We conduct an extensive experiment and human evaluation among different sizes and types of LLMs, showing that LLMs still have a significant journey ahead in their development towards becoming fully autonomous agents, especially in opponent modeling and team collaboration. We hope LLMArena could guide future research towards enhancing these capabilities in LLMs, ultimately leading to more sophisticated and practical applications in dynamic, multi-agent settings. The code and data will be available.", "authors": ["Junzhe Chen", "Xuming Hu", "Shuodi Liu", "Shiyu Huang", "Wei-Wei Tu", "Zhaofeng He", "Lijie Wen"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-26", "url": "https://arxiv.org/abs/2402.16499", "pdf_url": "https://arxiv.org/pdf/2402.16499v1", "arxiv_id": "2402.16499", "doi": "10.48550/arXiv.2402.16499", "citation_count": 37, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.3949}
{"id": "76dc91198a14dd28fb47da908f1cc6c7438b0d31c4ea4b25bbb15ccf6cfb37d8", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent contract design with a budget", "abstract": "We study a multi-agent contract design problem with moral hazard. In our model, each agent exerts costly effort towards an individual task at which it may either succeed or fail, and the principal, who wishes to encourage effort, has an exclusive-use budget that it can use to reward the agents. A motivating application is crowdsourcing for innovation, where a fixed budget is provided to a crowdsourcing platform to use for rewarding participants based on their submissions. Our main contribution is to introduce a novel class of contracts, which we call Luce contracts, and show that there is always a Luce contract that is optimal. A (generic) Luce contract assigns weights to the agents and distributes the entire budget among the successful agents in proportion to their weights. Furthermore, we characterize effort profiles that can be implemented by Luce contracts and show that Luce contracts offer a way to mitigate the uncertainty in total payments compared to alternative contracts-such as piece-rate or bonus-pool contracts-suggesting their desirability even in environments without budget constraints.", "authors": ["Sumit Goel", "Wade Hann-Caruthers"], "categories": ["econ.TH", "cs.GT"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2024-02-24", "url": "https://arxiv.org/abs/2402.15890", "pdf_url": "https://arxiv.org/pdf/2402.15890v3", "arxiv_id": "2402.15890", "doi": null, "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "c8bf1cc35fe89b35e27bb08a95344d0ad4bde97471a3043b3cac132cc56c0504", "sources": ["arxiv", "semantic_scholar"], "title": "AgentLite: A Lightweight Library for Building and Advancing Task-Oriented LLM Agent System", "abstract": "The booming success of LLMs initiates rapid development in LLM agents. Though the foundation of an LLM agent is the generative model, it is critical to devise the optimal reasoning strategies and agent architectures. Accordingly, LLM agent research advances from the simple chain-of-thought prompting to more complex ReAct and Reflection reasoning strategy; agent architecture also evolves from single agent generation to multi-agent conversation, as well as multi-LLM multi-agent group chat. However, with the existing intricate frameworks and libraries, creating and evaluating new reasoning strategies and agent architectures has become a complex challenge, which hinders research investigation into LLM agents. Thus, we open-source a new AI agent library, AgentLite, which simplifies this process by offering a lightweight, user-friendly platform for innovating LLM agent reasoning, architectures, and applications with ease. AgentLite is a task-oriented framework designed to enhance the ability of agents to break down tasks and facilitate the development of multi-agent systems. Furthermore, we introduce multiple practical applications developed with AgentLite to demonstrate its convenience and flexibility. Get started now at: \\url{https://github.com/SalesforceAIResearch/AgentLite}.", "authors": ["Zhiwei Liu", "Weiran Yao", "Jianguo Zhang", "Liangwei Yang", "Zuxin Liu", "Juntao Tan", "Prafulla K. Choubey", "Tian Lan", "Jason Wu", "Huan Wang", "Shelby Heinecke", "Caiming Xiong", "Silvio Savarese"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-23", "url": "https://arxiv.org/abs/2402.15538", "pdf_url": "https://arxiv.org/pdf/2402.15538v1", "arxiv_id": "2402.15538", "doi": "10.48550/arXiv.2402.15538", "citation_count": 61, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/SalesforceAIResearch/AgentLite", "venue": "arXiv.org", "quality_score": 0.4481}
{"id": "29733f3612aceaa9963f1fbf54a7bf64efff763d40f092abf678f25b198aa98b", "sources": ["arxiv", "semantic_scholar"], "title": "MACRec: a Multi-Agent Collaboration Framework for Recommendation", "abstract": "LLM-based agents have gained considerable attention for their decision-making skills and ability to handle complex tasks. Recognizing the current gap in leveraging agent capabilities for multi-agent collaboration in recommendation systems, we introduce MACRec, a novel framework designed to enhance recommendation systems through multi-agent collaboration. Unlike existing work on using agents for user/item simulation, we aim to deploy multi-agents to tackle recommendation tasks directly. In our framework, recommendation tasks are addressed through the collaborative efforts of various specialized agents, including Manager, User/Item Analyst, Reflector, Searcher, and Task Interpreter, with different working flows. Furthermore, we provide application examples of how developers can easily use MACRec on various recommendation tasks, including rating prediction, sequential recommendation, conversational recommendation, and explanation generation of recommendation results. The framework and demonstration video are publicly available at https://github.com/wzf2000/MACRec.", "authors": ["Zhefan Wang", "Yuanqing Yu", "Wendi Zheng", "Weizhi Ma", "Min Zhang"], "categories": ["cs.IR"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-23", "url": "https://arxiv.org/abs/2402.15235", "pdf_url": "https://arxiv.org/pdf/2402.15235v3", "arxiv_id": "2402.15235", "doi": "10.1145/3626772.3657669", "citation_count": 91, "influential_citation_count": 2, "has_code": true, "code_url": "https://github.com/wzf2000/MACRec", "venue": "Annual International ACM SIGIR Conference on Research and Development in Information Retrieval", "quality_score": 0.4909}
{"id": "d2428eb27cc297a2dea18ccb89817c04f05740cd4f7cea76df091b488694a93b", "sources": ["arxiv", "semantic_scholar"], "title": "AgentScope: A Flexible yet Robust Multi-Agent Platform", "abstract": "With the rapid advancement of Large Language Models (LLMs), significant progress has been made in multi-agent applications. However, the complexities in coordinating agents' cooperation and LLMs' erratic performance pose notable challenges in developing robust and efficient multi-agent applications. To tackle these challenges, we propose AgentScope, a developer-centric multi-agent platform with message exchange as its core communication mechanism. The abundant syntactic tools, built-in agents and service functions, user-friendly interfaces for application demonstration and utility monitor, zero-code programming workstation, and automatic prompt tuning mechanism significantly lower the barriers to both development and deployment. Towards robust and flexible multi-agent application, AgentScope provides both built-in and customizable fault tolerance mechanisms. At the same time, it is also armed with system-level support for managing and utilizing multi-modal data, tools, and external knowledge. Additionally, we design an actor-based distribution framework, enabling easy conversion between local and distributed deployments and automatic parallel optimization without extra effort. With these features, AgentScope empowers developers to build applications that fully realize the potential of intelligent agents. We have released AgentScope at https://github.com/modelscope/agentscope, and hope AgentScope invites wider participation and innovation in this fast-moving field.", "authors": ["Dawei Gao", "Zitao Li", "Xuchen Pan", "Weirui Kuang", "Zhijian Ma", "Bingchen Qian", "Fei Wei", "Wenhao Zhang", "Yuexiang Xie", "Daoyuan Chen", "Liuyi Yao", "Hongyi Peng", "Zeyu Zhang", "Lin Zhu", "Chen Cheng", "Hongzhu Shi", "Yaliang Li", "Bolin Ding", "Jingren Zhou"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-21", "url": "https://arxiv.org/abs/2402.14034", "pdf_url": "https://arxiv.org/pdf/2402.14034v2", "arxiv_id": "2402.14034", "doi": "10.48550/arXiv.2402.14034", "citation_count": 114, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/modelscope/agentscope", "venue": "arXiv.org", "quality_score": 0.5152}
{"id": "a2779bfab46e06e895c2c103c3a497f078b636e56fcdc57c99113fafc9e18e63", "sources": ["arxiv", "semantic_scholar"], "title": "MuLan: Multimodal-LLM Agent for Progressive and Interactive Multi-Object Diffusion", "abstract": "Existing text-to-image models still struggle to generate images of multiple objects, especially in handling their spatial positions, relative sizes, overlapping, and attribute bindings. To efficiently address these challenges, we develop a training-free Multimodal-LLM agent (MuLan), as a human painter, that can progressively generate multi-object with intricate planning and feedback control. MuLan harnesses a large language model (LLM) to decompose a prompt to a sequence of sub-tasks, each generating only one object by stable diffusion, conditioned on previously generated objects. Unlike existing LLM-grounded methods, MuLan only produces a high-level plan at the beginning while the exact size and location of each object are determined upon each sub-task by an LLM and attention guidance. Moreover, MuLan adopts a vision-language model (VLM) to provide feedback to the image generated in each sub-task and control the diffusion model to re-generate the image if it violates the original prompt. Hence, each model in every step of MuLan only needs to address an easy sub-task it is specialized for. The multi-step process also allows human users to monitor the generation process and make preferred changes at any intermediate step via text prompts, thereby improving the human-AI collaboration experience. We collect 200 prompts containing multi-objects with spatial relationships and attribute bindings from different benchmarks to evaluate MuLan. The results demonstrate the superiority of MuLan in generating multiple objects over baselines and its creativity when collaborating with human users. The code is available at https://github.com/measure-infinity/mulan-code.", "authors": ["Sen Li", "Ruochen Wang", "Cho-Jui Hsieh", "Minhao Cheng", "Tianyi Zhou"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-20", "url": "https://arxiv.org/abs/2402.12741", "pdf_url": "https://arxiv.org/pdf/2402.12741v2", "arxiv_id": "2402.12741", "doi": null, "citation_count": 9, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/measure-infinity/mulan-code", "venue": null, "quality_score": 0.25}
{"id": "c2e884c616d99c0b45de7f42f49d7c491805537ba6d607e924412c2438ec1a06", "sources": ["arxiv", "semantic_scholar"], "title": "Learning From Failure: Integrating Negative Examples when Fine-tuning Large Language Models as Agents", "abstract": "Large language models (LLMs) have achieved success in acting as agents, which interact with environments through tools such as search engines. However, LLMs are optimized for language generation instead of tool use during training or alignment, limiting their effectiveness as agents. To resolve this problem, previous work has first collected interaction trajectories between LLMs and environments, using only trajectories that successfully finished the task to fine-tune smaller models, making fine-tuning data scarce and acquiring it both difficult and costly. Discarding failed trajectories also leads to significant wastage of data and resources and limits the possible optimization paths during fine-tuning. In this paper, we argue that unsuccessful trajectories offer valuable insights, and LLMs can learn from these trajectories through appropriate quality control and fine-tuning strategies. By simply adding a prefix or suffix that tells the model whether to generate a successful trajectory during training, we improve model performance by a large margin on mathematical reasoning, multi-hop question answering, and strategic question answering tasks. We further analyze the inference results and find that our method provides a better trade-off between valuable information and errors in unsuccessful trajectories. To our knowledge, we are the first to demonstrate the value of negative trajectories and their application in agent-tunning scenarios. Our findings offer guidance for developing better agent-tuning methods and low-resource data usage techniques.", "authors": ["Renxi Wang", "Haonan Li", "Xudong Han", "Yixuan Zhang", "Timothy Baldwin"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-18", "url": "https://arxiv.org/abs/2402.11651", "pdf_url": "https://arxiv.org/pdf/2402.11651v2", "arxiv_id": "2402.11651", "doi": "10.48550/arXiv.2402.11651", "citation_count": 46, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.418}
{"id": "3773fe65b9e3ab93cc2fd6ee3c7e51337d468b4626d1d500cae78633e5e27d31", "sources": ["arxiv", "semantic_scholar"], "title": "KG-Agent: An Efficient Autonomous Agent Framework for Complex Reasoning over Knowledge Graph", "abstract": "In this paper, we aim to improve the reasoning ability of large language models (LLMs) over knowledge graphs (KGs) to answer complex questions. Inspired by existing methods that design the interaction strategy between LLMs and KG, we propose an autonomous LLM-based agent framework, called KG-Agent, which enables a small LLM to actively make decisions until finishing the reasoning process over KGs. In KG-Agent, we integrate the LLM, multifunctional toolbox, KG-based executor, and knowledge memory, and develop an iteration mechanism that autonomously selects the tool then updates the memory for reasoning over KG. To guarantee the effectiveness, we leverage program language to formulate the multi-hop reasoning process over the KG, and synthesize a code-based instruction dataset to fine-tune the base LLM. Extensive experiments demonstrate that only using 10K samples for tuning LLaMA-7B can outperform state-of-the-art methods using larger LLMs or more data, on both in-domain and out-domain datasets. Our code and data will be publicly released.", "authors": ["Jinhao Jiang", "Kun Zhou", "Wayne Xin Zhao", "Yang Song", "Chen Zhu", "Hengshu Zhu", "Ji-Rong Wen"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-17", "url": "https://arxiv.org/abs/2402.11163", "pdf_url": "https://arxiv.org/pdf/2402.11163v1", "arxiv_id": "2402.11163", "doi": "10.48550/arXiv.2402.11163", "citation_count": 116, "influential_citation_count": 11, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.5396}
{"id": "bd9ff71f4412de44844feefb48478f25489c7a44552d8eec3f1f4d2273e536b7", "sources": ["arxiv", "semantic_scholar"], "title": "Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based Agents", "abstract": "Driven by the rapid development of Large Language Models (LLMs), LLM-based agents have been developed to handle various real-world applications, including finance, healthcare, and shopping, etc. It is crucial to ensure the reliability and security of LLM-based agents during applications. However, the safety issues of LLM-based agents are currently under-explored. In this work, we take the first step to investigate one of the typical safety threats, backdoor attack, to LLM-based agents. We first formulate a general framework of agent backdoor attacks, then we present a thorough analysis of different forms of agent backdoor attacks. Specifically, compared with traditional backdoor attacks on LLMs that are only able to manipulate the user inputs and model outputs, agent backdoor attacks exhibit more diverse and covert forms: (1) From the perspective of the final attacking outcomes, the agent backdoor attacker can not only choose to manipulate the final output distribution, but also introduce the malicious behavior in an intermediate reasoning step only, while keeping the final output correct. (2) Furthermore, the former category can be divided into two subcategories based on trigger locations, in which the backdoor trigger can either be hidden in the user query or appear in an intermediate observation returned by the external environment. We implement the above variations of agent backdoor attacks on two typical agent tasks including web shopping and tool utilization. Extensive experiments show that LLM-based agents suffer severely from backdoor attacks and such backdoor vulnerability cannot be easily mitigated by current textual backdoor defense algorithms. This indicates an urgent need for further research on the development of targeted defenses against backdoor attacks on LLM-based agents. Warning: This paper may contain biased content.", "authors": ["Wenkai Yang", "Xiaohan Bi", "Yankai Lin", "Sishuo Chen", "Jie Zhou", "Xu Sun"], "categories": ["cs.CR", "cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-17", "url": "https://arxiv.org/abs/2402.11208", "pdf_url": "https://arxiv.org/pdf/2402.11208v2", "arxiv_id": "2402.11208", "doi": "10.48550/arXiv.2402.11208", "citation_count": 166, "influential_citation_count": 9, "has_code": true, "code_url": "https://github.com/lancopku/agent-backdoor-attacks", "venue": "Neural Information Processing Systems", "quality_score": 0.5557}
{"id": "753af2410503d0a805944fc10002f45a6072f3e0a215db41402895ffbfa5112a", "sources": ["arxiv", "semantic_scholar"], "title": "TDAG: A Multi-Agent Framework based on Dynamic Task Decomposition and Agent Generation", "abstract": "The emergence of Large Language Models (LLMs) like ChatGPT has inspired the development of LLM-based agents capable of addressing complex, real-world tasks. However, these agents often struggle during task execution due to methodological constraints, such as error propagation and limited adaptability. To address this issue, we propose a multi-agent framework based on dynamic Task Decomposition and Agent Generation (TDAG). This framework dynamically decomposes complex tasks into smaller subtasks and assigns each to a specifically generated subagent, thereby enhancing adaptability in diverse and unpredictable real-world tasks. Simultaneously, existing benchmarks often lack the granularity needed to evaluate incremental progress in complex, multi-step tasks. In response, we introduce ItineraryBench in the context of travel planning, featuring interconnected, progressively complex tasks with a fine-grained evaluation system. ItineraryBench is designed to assess agents' abilities in memory, planning, and tool usage across tasks of varying complexity. Our experimental results reveal that TDAG significantly outperforms established baselines, showcasing its superior adaptability and context awareness in complex task scenarios.", "authors": ["Yaoxiang Wang", "Zhiyong Wu", "Junfeng Yao", "Jinsong Su"], "categories": ["cs.CL"], "fields_of_study": ["Medicine", "Computer Science"], "published_date": "2024-02-15", "url": "https://arxiv.org/abs/2402.10178", "pdf_url": "https://arxiv.org/pdf/2402.10178v2", "arxiv_id": "2402.10178", "doi": "10.48550/arXiv.2402.10178", "citation_count": 51, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Neural Networks", "quality_score": 0.429}
{"id": "f9f6265b4a4f26cf8bfcba14e6280b27f3f5134292e8785ddd5dea8c00271b39", "sources": ["arxiv", "semantic_scholar"], "title": "ABIDES-Economist: Agent-Based Simulator of Economic Systems with Learning Agents", "abstract": "We present ABIDES-Economist, an agent-based simulator for economic systems that includes heterogeneous households, firms, a central bank, and a government. Agent behavior can be defined using domain-specific behavioral rules or learned through reinforcement learning by specifying their objectives. We integrate reinforcement learning capabilities for all agents using the OpenAI Gym environment framework for the multi-agent system. To enhance the realism of our model, we base agent parameters and action spaces on economic literature and real U.S. economic data. To tackle the challenges of calibrating heterogeneous agent-based economic models, we conduct a comprehensive survey of stylized facts related to both microeconomic and macroeconomic time series data. We then validate ABIDES-Economist by demonstrating its ability to generate simulated data that aligns with the relevant stylized facts for the economic scenario under consideration, following the learning of all agent behaviors via reinforcement learning. Specifically, we train our economic agents' policies under two broad configurations. The first configuration demonstrates that the learned economic agents produce system data consistent with macroeconomic and microeconomic stylized facts. The second configuration illustrates the utility of the validated simulation platform in designing regulatory policies for the central bank and government. These policies outperform standard rule-based approaches from the literature, which often overlook agent heterogeneity, shocks, and agent adaptability.", "authors": ["Kshama Dwarakanath", "Tucker Balch", "Svitlana Vyetrenko"], "categories": ["cs.MA", "econ.GN"], "fields_of_study": ["Computer Science", "Economics"], "published_date": "2024-02-14", "url": "https://arxiv.org/abs/2402.09563", "pdf_url": "https://arxiv.org/pdf/2402.09563v2", "arxiv_id": "2402.09563", "doi": null, "citation_count": 4, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.1747}
{"id": "10d3f5fae21e2347e6e2b76f83c877ab0c27cc0ce0ec86ca5cc43b681aef5760", "sources": ["arxiv", "semantic_scholar"], "title": "Simulating Human Strategic Behavior: Comparing Single and Multi-agent LLMs", "abstract": "When creating policies, plans, or designs for people, it is challenging for designers to foresee all of the ways in which people may reason and behave. Recently, Large Language Models (LLMs) have been shown to be able to simulate human reasoning. We extend this work by measuring LLMs ability to simulate strategic reasoning in the ultimatum game, a classic economics bargaining experiment. Experimental evidence shows human strategic reasoning is complex; people will often choose to punish other players to enforce social norms even at personal expense. We test if LLMs can replicate this behavior in simulation, comparing two structures: single LLMs and multi-agent systems. We compare their abilities to (1) simulate human-like reasoning in the ultimatum game, (2) simulate two player personalities, greedy and fair, and (3) create robust strategies that are logically complete and consistent with personality. Our evaluation shows that multi-agent systems are more accurate than single LLMs (88 percent vs. 50 percent) in simulating human reasoning and actions for personality pairs. Thus, there is potential to use LLMs to simulate human strategic reasoning to help decision and policy-makers perform preliminary explorations of how people behave in systems.", "authors": ["Karthik Sreedhar", "Lydia Chilton"], "categories": ["cs.HC"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-13", "url": "https://arxiv.org/abs/2402.08189", "pdf_url": "https://arxiv.org/pdf/2402.08189v2", "arxiv_id": "2402.08189", "doi": "10.24251/hicss.2025.100", "citation_count": 20, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Hawaii International Conference on System Sciences", "quality_score": 0.3306}
{"id": "fbafb9177ec52d9e256eaa795bf4b61684b4ab0260b629d1e9fff478d7bcf7ca", "sources": ["arxiv", "semantic_scholar"], "title": "Distributed Quasi-Newton Method for Multi-Agent Optimization", "abstract": "We present a distributed quasi-Newton (DQN) method, which enables a group of agents to compute an optimal solution of a separable multi-agent optimization problem locally using an approximation of the curvature of the aggregate objective function. Each agent computes a descent direction from its local estimate of the aggregate Hessian, obtained from quasi-Newton approximation schemes using the gradient of its local objective function. Moreover, we introduce a distributed quasi-Newton method for equality-constrained optimization (EC-DQN), where each agent takes Karush-Kuhn-Tucker-like update steps to compute an optimal solution. In our algorithms, each agent communicates with its one-hop neighbors over a peer-to-peer communication network to compute a common solution. We prove convergence of our algorithms to a stationary point of the optimization problem. In addition, we demonstrate the competitive empirical convergence of our algorithm in both well-conditioned and ill-conditioned optimization problems, in terms of the computation time and communication cost incurred by each agent for convergence, compared to existing distributed first-order and second-order methods. Particularly, in ill-conditioned problems, our algorithms achieve a faster computation time for convergence, while requiring a lower communication cost, across a range of communication networks with different degrees of connectedness.", "authors": ["Ola Shorinwa", "Mac Schwager"], "categories": ["math.OC", "cs.MA", "eess.SY"], "fields_of_study": ["Computer Science", "Mathematics", "Engineering"], "published_date": "2024-02-09", "url": "https://arxiv.org/abs/2402.06778", "pdf_url": "https://arxiv.org/pdf/2402.06778v2", "arxiv_id": "2402.06778", "doi": "10.1109/TSP.2024.3424436", "citation_count": 10, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Signal Processing", "quality_score": 0.2603}
{"id": "4782ba08622cb75422f3c2a82f05e408c30da9fba2eb1dc99a2a9342a976ab84", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Multi-Agent Systems: Challenges and Open Problems", "abstract": "This paper explores multi-agent systems and identify challenges that remain inadequately addressed. By leveraging the diverse capabilities and roles of individual agents, multi-agent systems can tackle complex tasks through agent collaboration. We discuss optimizing task allocation, fostering robust reasoning through iterative debates, managing complex and layered context information, and enhancing memory management to support the intricate interactions within multi-agent systems. We also explore potential applications of multi-agent systems in blockchain systems to shed light on their future development and application in real-world distributed systems.", "authors": ["Shanshan Han", "Qifan Zhang", "Weizhao Jin", "Zhaozhuo Xu"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-05", "url": "https://arxiv.org/abs/2402.03578", "pdf_url": "https://arxiv.org/pdf/2402.03578v3", "arxiv_id": "2402.03578", "doi": "10.48550/arXiv.2402.03578", "citation_count": 142, "influential_citation_count": 11, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5396}
{"id": "13b68b769f4f808c60b35bee395c621dd0be553ce4aa9c58a65ae23ea672adab", "sources": ["arxiv", "semantic_scholar"], "title": "Executable Code Actions Elicit Better LLM Agents", "abstract": "Large Language Model (LLM) agents, capable of performing a broad range of actions, such as invoking tools and controlling robots, show great potential in tackling real-world challenges. LLM agents are typically prompted to produce actions by generating JSON or text in a pre-defined format, which is usually limited by constrained action space (e.g., the scope of pre-defined tools) and restricted flexibility (e.g., inability to compose multiple tools). This work proposes to use executable Python code to consolidate LLM agents' actions into a unified action space (CodeAct). Integrated with a Python interpreter, CodeAct can execute code actions and dynamically revise prior actions or emit new actions upon new observations through multi-turn interactions. Our extensive analysis of 17 LLMs on API-Bank and a newly curated benchmark shows that CodeAct outperforms widely used alternatives (up to 20% higher success rate). The encouraging performance of CodeAct motivates us to build an open-source LLM agent that interacts with environments by executing interpretable code and collaborates with users using natural language. To this end, we collect an instruction-tuning dataset CodeActInstruct that consists of 7k multi-turn interactions using CodeAct. We show that it can be used with existing data to improve models in agent-oriented tasks without compromising their general capability. CodeActAgent, finetuned from Llama2 and Mistral, is integrated with Python interpreter and uniquely tailored to perform sophisticated tasks (e.g., model training) using existing libraries and autonomously self-debug.", "authors": ["Xingyao Wang", "Yangyi Chen", "Lifan Yuan", "Yizhe Zhang", "Yunzhu Li", "Hao Peng", "Heng Ji"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-01", "url": "https://arxiv.org/abs/2402.01030", "pdf_url": "https://arxiv.org/pdf/2402.01030v4", "arxiv_id": "2402.01030", "doi": "10.48550/arXiv.2402.01030", "citation_count": 486, "influential_citation_count": 28, "has_code": true, "code_url": "https://github.com/xingyaoww/code-act", "venue": "International Conference on Machine Learning", "quality_score": 0.7312}
{"id": "2f71dd27dc091392551203f0287bb7bce5f4141bb994f54d772122d1dbab66c2", "sources": ["arxiv", "semantic_scholar"], "title": "Formal-LLM: Integrating Formal Language and Natural Language for Controllable LLM-based Agents", "abstract": "Recent advancements on Large Language Models (LLMs) enable AI Agents to automatically generate and execute multi-step plans to solve complex tasks. However, since LLM's content generation process is hardly controllable, current LLM-based agents frequently generate invalid or non-executable plans, which jeopardizes the performance of the generated plans and corrupts users' trust in LLM-based agents. In response, this paper proposes a novel \"Formal-LLM\" framework for LLM-based agents by integrating the expressiveness of natural language and the precision of formal language. Specifically, the framework allows agent developers to express their requirements or constraints for the planning process as an automaton. A stack-based LLM plan generation process is then conducted under the supervision of the automaton to ensure that the generated plan satisfies the constraints, making the planning process controllable. We conduct experiments on both benchmark tasks and practical real-life tasks, and our framework achieves over 50% overall performance increase, which validates the feasibility and effectiveness of employing Formal-LLM to guide the plan generation of agents, preventing the agents from generating invalid and unsuccessful plans. Further, more controllable LLM-based agents can facilitate the broader utilization of LLM in application scenarios where high validity of planning is essential. The source code of this work is available at https://github.com/agiresearch/Formal-LLM.", "authors": ["Zelong Li", "Wenyue Hua", "Hao Wang", "He Zhu", "Yongfeng Zhang"], "categories": ["cs.LG", "cs.AI", "cs.CL", "cs.FL"], "fields_of_study": ["Computer Science"], "published_date": "2024-02-01", "url": "https://arxiv.org/abs/2402.00798", "pdf_url": "https://arxiv.org/pdf/2402.00798v4", "arxiv_id": "2402.00798", "doi": "10.48550/arXiv.2402.00798", "citation_count": 45, "influential_citation_count": 3, "has_code": true, "code_url": "https://github.com/agiresearch/Formal-LLM", "venue": "arXiv.org", "quality_score": 0.4157}
{"id": "8d8a2eed5d281bc919d7da7973766aedd470450b3e455e862d3c12523c126c0d", "sources": ["arxiv", "semantic_scholar"], "title": "Zero-Shot Reinforcement Learning via Function Encoders", "abstract": "Although reinforcement learning (RL) can solve many challenging sequential decision making problems, achieving zero-shot transfer across related tasks remains a challenge. The difficulty lies in finding a good representation for the current task so that the agent understands how it relates to previously seen tasks. To achieve zero-shot transfer, we introduce the function encoder, a representation learning algorithm which represents a function as a weighted combination of learned, non-linear basis functions. By using a function encoder to represent the reward function or the transition function, the agent has information on how the current task relates to previously seen tasks via a coherent vector representation. Thus, the agent is able to achieve transfer between related tasks at run time with no additional training. We demonstrate state-of-the-art data efficiency, asymptotic performance, and training stability in three RL fields by augmenting basic RL algorithms with a function encoder task representation.", "authors": ["Tyler Ingebrand", "Amy Zhang", "Ufuk Topcu"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-30", "url": "https://arxiv.org/abs/2401.17173", "pdf_url": "https://arxiv.org/pdf/2401.17173v3", "arxiv_id": "2401.17173", "doi": "10.48550/arXiv.2401.17173", "citation_count": 18, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.3197}
{"id": "8997ef5b2e5bc1be8d7bf6e5652229a9a332257e8bbe61b4cc5b749a6f89404a", "sources": ["arxiv", "semantic_scholar"], "title": "A mechanism for discovering semantic relationships among agent communication protocols", "abstract": "One relevant aspect in the development of the Semantic Web framework is the achievement of a real inter-agents communication capability at the semantic level. Agents should be able to communicate with each other freely using different communication protocols, constituted by communication acts. For that scenario, we introduce in this paper an efficient mechanism presenting the following main features: - It promotes the description of the communication acts of protocols as classes that belong to a communication acts ontology, and associates to those acts a social commitment semantics formalized through predicates in the Event Calculus. - It is sustained on the idea that different protocols can be compared semantically by looking to the set of fluents associated to each branch of the protocols. Those sets are generated using Semantic Web technology rules. - It discovers the following types of protocol relationships: equivalence, specialization, restriction, prefix, suffix, infix and complement_to_infix.", "authors": ["Idoia Berges", "Jesús Bermúdez", "Alfredo Goñi", "Arantza Illarramendi"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-29", "url": "https://arxiv.org/abs/2401.16216", "pdf_url": "https://arxiv.org/pdf/2401.16216v1", "arxiv_id": "2401.16216", "doi": "10.1007/s10458-010-9154-1", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.0753}
{"id": "6b2fd6f53e525a90ee1b3a389662ad10e205f56ac9f1d18ac3aa74fe42040987", "sources": ["arxiv", "semantic_scholar"], "title": "Mobile-Agent: Autonomous Multi-Modal Mobile Device Agent with Visual Perception", "abstract": "Mobile device agent based on Multimodal Large Language Models (MLLM) is becoming a popular application. In this paper, we introduce Mobile-Agent, an autonomous multi-modal mobile device agent. Mobile-Agent first leverages visual perception tools to accurately identify and locate both the visual and textual elements within the app's front-end interface. Based on the perceived vision context, it then autonomously plans and decomposes the complex operation task, and navigates the mobile Apps through operations step by step. Different from previous solutions that rely on XML files of Apps or mobile system metadata, Mobile-Agent allows for greater adaptability across diverse mobile operating environments in a vision-centric way, thereby eliminating the necessity for system-specific customizations. To assess the performance of Mobile-Agent, we introduced Mobile-Eval, a benchmark for evaluating mobile device operations. Based on Mobile-Eval, we conducted a comprehensive evaluation of Mobile-Agent. The experimental results indicate that Mobile-Agent achieved remarkable accuracy and completion rates. Even with challenging instructions, such as multi-app operations, Mobile-Agent can still complete the requirements. Code and model will be open-sourced at https://github.com/X-PLUG/MobileAgent.", "authors": ["Junyang Wang", "Haiyang Xu", "Jiabo Ye", "Ming Yan", "Weizhou Shen", "Ji Zhang", "Fei Huang", "Jitao Sang"], "categories": ["cs.CL", "cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-29", "url": "https://arxiv.org/abs/2401.16158", "pdf_url": "https://arxiv.org/pdf/2401.16158v2", "arxiv_id": "2401.16158", "doi": "10.48550/arXiv.2401.16158", "citation_count": 296, "influential_citation_count": 22, "has_code": true, "code_url": "https://github.com/X-PLUG/MobileAgent", "venue": "arXiv.org", "quality_score": 0.6809}
{"id": "6f483b2b431bc7112882ceeb79b606965daa613b64a2852e1a59e25eb697072c", "sources": ["arxiv", "semantic_scholar"], "title": "AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents", "abstract": "Evaluating Large Language Models (LLMs) as general-purpose agents is essential for understanding their capabilities and facilitating their integration into practical applications. However, the evaluation process presents substantial challenges. A primary obstacle is the benchmarking of agent performance across diverse scenarios within a unified framework, especially in maintaining partially-observable environments and ensuring multi-round interactions. Moreover, current evaluation frameworks mostly focus on the final success rate, revealing few insights during the process and failing to provide a deep understanding of the model abilities. To address these challenges, we introduce AgentBoard, a pioneering comprehensive benchmark and accompanied open-source evaluation framework tailored to analytical evaluation of LLM agents. AgentBoard offers a fine-grained progress rate metric that captures incremental advancements as well as a comprehensive evaluation toolkit that features easy assessment of agents for multi-faceted analysis. This not only sheds light on the capabilities and limitations of LLM agents but also propels the interpretability of their performance to the forefront. Ultimately, AgentBoard serves as a step towards demystifying agent behaviors and accelerating the development of stronger LLM agents.", "authors": ["Chang Ma", "Junlei Zhang", "Zhihao Zhu", "Cheng Yang", "Yujiu Yang", "Yaohui Jin", "Zhenzhong Lan", "Lingpeng Kong", "Junxian He"], "categories": ["cs.CL", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-24", "url": "https://arxiv.org/abs/2401.13178", "pdf_url": "https://arxiv.org/pdf/2401.13178v2", "arxiv_id": "2401.13178", "doi": "10.48550/arXiv.2401.13178", "citation_count": 225, "influential_citation_count": 30, "has_code": true, "code_url": null, "venue": "Neural Information Processing Systems", "quality_score": 0.7457}
{"id": "e482b1689fbe08e11f98aeda598ca57e7b71ded1d9687781d6c0719db6d77f27", "sources": ["arxiv", "semantic_scholar"], "title": "PsySafe: A Comprehensive Framework for Psychological-based Attack, Defense, and Evaluation of Multi-agent System Safety", "abstract": "Multi-agent systems, when enhanced with Large Language Models (LLMs), exhibit profound capabilities in collective intelligence. However, the potential misuse of this intelligence for malicious purposes presents significant risks. To date, comprehensive research on the safety issues associated with multi-agent systems remains limited. In this paper, we explore these concerns through the innovative lens of agent psychology, revealing that the dark psychological states of agents constitute a significant threat to safety. To tackle these concerns, we propose a comprehensive framework (PsySafe) grounded in agent psychology, focusing on three key areas: firstly, identifying how dark personality traits in agents can lead to risky behaviors; secondly, evaluating the safety of multi-agent systems from the psychological and behavioral perspectives, and thirdly, devising effective strategies to mitigate these risks. Our experiments reveal several intriguing phenomena, such as the collective dangerous behaviors among agents, agents' self-reflection when engaging in dangerous behavior, and the correlation between agents' psychological assessments and dangerous behaviors. We anticipate that our framework and observations will provide valuable insights for further research into the safety of multi-agent systems. We will make our data and code publicly accessible at https://github.com/AI4Good24/PsySafe.", "authors": ["Zaibin Zhang", "Yongting Zhang", "Lijun Li", "Hongzhi Gao", "Lijun Wang", "Huchuan Lu", "Feng Zhao", "Yu Qiao", "Jing Shao"], "categories": ["cs.CL", "cs.AI", "cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-22", "url": "https://arxiv.org/abs/2401.11880", "pdf_url": "https://arxiv.org/pdf/2401.11880v3", "arxiv_id": "2401.11880", "doi": "10.48550/arXiv.2401.11880", "citation_count": 88, "influential_citation_count": 8, "has_code": true, "code_url": "https://github.com/AI4Good24/PsySafe", "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4873}
{"id": "db1aa8034c5c29f0d9ccb1549117866687976d888ccb3ce5e0df86f76959f2af", "sources": ["arxiv", "semantic_scholar"], "title": "Large Language Model based Multi-Agents: A Survey of Progress and Challenges", "abstract": "Large Language Models (LLMs) have achieved remarkable success across a wide array of tasks. Due to the impressive planning and reasoning abilities of LLMs, they have been used as autonomous agents to do many tasks automatically. Recently, based on the development of using one LLM as a single planning or decision-making agent, LLM-based multi-agent systems have achieved considerable progress in complex problem-solving and world simulation. To provide the community with an overview of this dynamic field, we present this survey to offer an in-depth discussion on the essential aspects of multi-agent systems based on LLMs, as well as the challenges. Our goal is for readers to gain substantial insights on the following questions: What domains and environments do LLM-based multi-agents simulate? How are these agents profiled and how do they communicate? What mechanisms contribute to the growth of agents' capacities? For those interested in delving into this field of study, we also summarize the commonly used datasets or benchmarks for them to have convenient access. To keep researchers updated on the latest studies, we maintain an open-source GitHub repository, dedicated to outlining the research on LLM-based multi-agent systems.", "authors": ["Taicheng Guo", "Xiuying Chen", "Yaqi Wang", "Ruidi Chang", "Shichao Pei", "Nitesh V. Chawla", "Olaf Wiest", "Xiangliang Zhang"], "categories": ["cs.CL", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-21", "url": "https://arxiv.org/abs/2402.01680", "pdf_url": "https://arxiv.org/pdf/2402.01680v2", "arxiv_id": "2402.01680", "doi": "10.48550/arXiv.2402.01680", "citation_count": 951, "influential_citation_count": 45, "has_code": true, "code_url": null, "venue": "International Joint Conference on Artificial Intelligence", "quality_score": 0.8314}
{"id": "1c165d7ac979bbc6d88df396260a4e0cef92a8d57f80fbe720c001dd97f3a598", "sources": ["arxiv", "semantic_scholar"], "title": "Measuring Policy Distance for Multi-Agent Reinforcement Learning", "abstract": "Diversity plays a crucial role in improving the performance of multi-agent reinforcement learning (MARL). Currently, many diversity-based methods have been developed to overcome the drawbacks of excessive parameter sharing in traditional MARL. However, there remains a lack of a general metric to quantify policy differences among agents. Such a metric would not only facilitate the evaluation of the diversity evolution in multi-agent systems, but also provide guidance for the design of diversity-based MARL algorithms. In this paper, we propose the multi-agent policy distance (MAPD), a general tool for measuring policy differences in MARL. By learning the conditional representations of agents' decisions, MAPD can computes the policy distance between any pair of agents. Furthermore, we extend MAPD to a customizable version, which can quantify differences among agent policies on specified aspects. Based on the online deployment of MAPD, we design a multi-agent dynamic parameter sharing (MADPS) algorithm as an example of the MAPD's applications. Extensive experiments demonstrate that our method is effective in measuring differences in agent policies and specific behavioral tendencies. Moreover, in comparison to other methods of parameter sharing, MADPS exhibits superior performance.", "authors": ["Tianyi Hu", "Zhiqiang Pu", "Xiaolin Ai", "Tenghai Qiu", "Jianqiang Yi"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-20", "url": "https://arxiv.org/abs/2401.11257", "pdf_url": "https://arxiv.org/pdf/2401.11257v2", "arxiv_id": "2401.11257", "doi": "10.48550/arXiv.2401.11257", "citation_count": 5, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1945}
{"id": "a208da8d2111aabc5096b6ac560bc53d8b93fd28a63e3ea5e79db9249df550b9", "sources": ["arxiv", "semantic_scholar"], "title": "MLLM-Tool: A Multimodal Large Language Model For Tool Agent Learning", "abstract": "Recently, the astonishing performance of large language models (LLMs) in natural language comprehension and generation tasks triggered lots of exploration of using them as central controllers to build agent systems. Multiple studies focus on bridging the LLMs to external tools to extend the application scenarios. However, the current LLMs' ability to perceive tool use is limited to a single text query, which may result in ambiguity in understanding the users' real intentions. LLMs are expected to eliminate that by perceiving the information in the visual- or auditory-grounded instructions. Therefore, in this paper, we propose MLLM-Tool, a system incorporating open-source LLMs and multi-modal encoders so that the learned LLMs can be conscious of multi-modal input instruction and then select the function-matched tool correctly. To facilitate the evaluation of the model's capability, we collect a dataset featuring multi-modal input tools from HuggingFace. Another essential feature of our dataset is that it also contains multiple potential choices for the same instruction due to the existence of identical functions and synonymous functions, which provides more potential solutions for the same query. The experiments reveal that our MLLM-Tool is capable of recommending appropriate tools for multi-modal instructions. Codes and data are available at https://github.com/MLLM-Tool/MLLM-Tool.", "authors": ["Chenyu Wang", "Weixin Luo", "Sixun Dong", "Xiaohua Xuan", "Zhengxin Li", "Lin Ma", "Shenghua Gao"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-19", "url": "https://arxiv.org/abs/2401.10727", "pdf_url": "https://arxiv.org/pdf/2401.10727v3", "arxiv_id": "2401.10727", "doi": "10.1109/WACV61041.2025.00650", "citation_count": 67, "influential_citation_count": 5, "has_code": true, "code_url": "https://github.com/MLLM-Tool/MLLM-Tool", "venue": "IEEE Workshop/Winter Conference on Applications of Computer Vision", "quality_score": 0.4581}
{"id": "0e1504a61a7d4c12ef6b8ab967e38f0f61247824bb13096df97e409ad3649efb", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative Edge Caching Based on Elastic Federated and Multi-Agent Deep Reinforcement Learning in Next-Generation Network", "abstract": "Edge caching is a promising solution for next-generation networks by empowering caching units in small-cell base stations (SBSs), which allows user equipments (UEs) to fetch users' requested contents that have been pre-cached in SBSs. It is crucial for SBSs to predict accurate popular contents through learning while protecting users' personal information. Traditional federated learning (FL) can protect users' privacy but the data discrepancies among UEs can lead to a degradation in model quality. Therefore, it is necessary to train personalized local models for each UE to predict popular contents accurately. In addition, the cached contents can be shared among adjacent SBSs in next-generation networks, thus caching predicted popular contents in different SBSs may affect the cost to fetch contents. Hence, it is critical to determine where the popular contents are cached cooperatively. To address these issues, we propose a cooperative edge caching scheme based on elastic federated and multi-agent deep reinforcement learning (CEFMR) to optimize the cost in the network. We first propose an elastic FL algorithm to train the personalized model for each UE, where adversarial autoencoder (AAE) model is adopted for training to improve the prediction accuracy, then {a popular} content prediction algorithm is proposed to predict the popular contents for each SBS based on the trained AAE model. Finally, we propose a multi-agent deep reinforcement learning (MADRL) based algorithm to decide where the predicted popular contents are collaboratively cached among SBSs. Our experimental results demonstrate the superiority of our proposed scheme to existing baseline caching schemes.", "authors": ["Qiong Wu", "Wenhua Wang", "Pingyi Fan", "Qiang Fan", "Huiling Zhu", "Khaled B. Letaief"], "categories": ["cs.LG", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-18", "url": "https://arxiv.org/abs/2401.09886", "pdf_url": "https://arxiv.org/pdf/2401.09886v2", "arxiv_id": "2401.09886", "doi": "10.1109/TNSM.2024.3403842", "citation_count": 65, "influential_citation_count": 1, "has_code": true, "code_url": "https://github.com/qiongwu86/Edge-Caching-Based-on-Multi-Agent-Deep-Reinforcement-Learning-and-Federated-Learning", "venue": "IEEE Transactions on Network and Service Management", "quality_score": 0.4549}
{"id": "aee5f2973b114503ac3fef2978ddb4f544ffb459e207686bc036010bca581885", "sources": ["arxiv", "semantic_scholar"], "title": "R-Judge: Benchmarking Safety Risk Awareness for LLM Agents", "abstract": "Large language models (LLMs) have exhibited great potential in autonomously completing tasks across real-world applications. Despite this, these LLM agents introduce unexpected safety risks when operating in interactive environments. Instead of centering on the harmlessness of LLM-generated content in most prior studies, this work addresses the imperative need for benchmarking the behavioral safety of LLM agents within diverse environments. We introduce R-Judge, a benchmark crafted to evaluate the proficiency of LLMs in judging and identifying safety risks given agent interaction records. R-Judge comprises 569 records of multi-turn agent interaction, encompassing 27 key risk scenarios among 5 application categories and 10 risk types. It is of high-quality curation with annotated safety labels and risk descriptions. Evaluation of 11 LLMs on R-Judge shows considerable room for enhancing the risk awareness of LLMs: The best-performing model, GPT-4o, achieves 74.42% while no other models significantly exceed the random. Moreover, we reveal that risk awareness in open agent scenarios is a multi-dimensional capability involving knowledge and reasoning, thus challenging for LLMs. With further experiments, we find that fine-tuning on safety judgment significantly improve model performance while straightforward prompting mechanisms fail. R-Judge is publicly available at https://github.com/Lordog/R-Judge.", "authors": ["Tongxin Yuan", "Zhiwei He", "Lingzhong Dong", "Yiming Wang", "Ruijie Zhao", "Tian Xia", "Lizhen Xu", "Binglin Zhou", "Fangqi Li", "Zhuosheng Zhang", "Rui Wang", "Gongshen Liu"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-18", "url": "https://arxiv.org/abs/2401.10019", "pdf_url": "https://arxiv.org/pdf/2401.10019v3", "arxiv_id": "2401.10019", "doi": "10.48550/arXiv.2401.10019", "citation_count": 227, "influential_citation_count": 23, "has_code": true, "code_url": "https://github.com/Lordog/R-Judge", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.6901}
{"id": "3b5e717fc383731e127e738c5842430f05d1b04a76cc8beb48d2e9c575923e4d", "sources": ["arxiv", "semantic_scholar"], "title": "Small LLMs Are Weak Tool Learners: A Multi-LLM Agent", "abstract": "Large Language Model (LLM) agents significantly extend the capabilities of standalone LLMs, empowering them to interact with external tools (e.g., APIs, functions) and complete various tasks in a self-directed fashion. The challenge of tool use demands that LLMs not only understand user queries and generate answers accurately but also excel in task planning, tool invocation, and result summarization. While traditional works focus on training a single LLM with all these capabilities, performance limitations become apparent, particularly with smaller models. To overcome these challenges, we propose a novel approach that decomposes the aforementioned capabilities into a planner, caller, and summarizer. Each component is implemented by a single LLM that focuses on a specific capability and collaborates with others to accomplish the task. This modular framework facilitates individual updates and the potential use of smaller LLMs for building each capability. To effectively train this framework, we introduce a two-stage training paradigm. First, we fine-tune a backbone LLM on the entire dataset without discriminating sub-tasks, providing the model with a comprehensive understanding of the task. Second, the fine-tuned LLM is used to instantiate the planner, caller, and summarizer respectively, which are continually fine-tuned on respective sub-tasks. Evaluation across various tool-use benchmarks illustrates that our proposed multi-LLM framework surpasses the traditional single-LLM approach, highlighting its efficacy and advantages in tool learning.", "authors": ["Weizhou Shen", "Chenliang Li", "Hongzhan Chen", "Ming Yan", "Xiaojun Quan", "Hehong Chen", "Ji Zhang", "Fei Huang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-14", "url": "https://arxiv.org/abs/2401.07324", "pdf_url": "https://arxiv.org/pdf/2401.07324v3", "arxiv_id": "2401.07324", "doi": "10.48550/arXiv.2401.07324", "citation_count": 123, "influential_citation_count": 6, "has_code": true, "code_url": "https://github.com/X-PLUG/Multi-LLM-Agent", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.5234}
{"id": "b5a232712dad8960cee028fac736a4b3cde38263fc22cac16870b5171077222b", "sources": ["arxiv", "semantic_scholar"], "title": "Transparency as Delayed Observability in Multi-Agent Systems", "abstract": "Is transparency always beneficial in complex systems such as traffic networks and stock markets? How is transparency defined in multi-agent systems, and what is its optimal degree at which social welfare is highest? We take an agent-based view to define transparency (or its lacking) as delay in agent observability of environment states, and utilize simulations to analyze the impact of delay on social welfare. To model the adaptation of agent strategies with varying delays, we model agents as learners maximizing the same objectives under different delays in a simulated environment. Focusing on two agent types - constrained and unconstrained, we use multi-agent reinforcement learning to evaluate the impact of delay on agent outcomes and social welfare. Empirical demonstration of our framework in simulated financial markets shows opposing trends in outcomes of the constrained and unconstrained agents with delay, with an optimal partial transparency regime at which social welfare is maximal.", "authors": ["Kshama Dwarakanath", "Svitlana Vyetrenko", "Toks Oyebode", "Tucker Balch"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-10", "url": "https://arxiv.org/abs/2401.05563", "pdf_url": "https://arxiv.org/pdf/2401.05563v1", "arxiv_id": "2401.05563", "doi": "10.1109/WSC60868.2023.10407520", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Online World Conference on Soft Computing in Industrial Applications", "quality_score": 0.0}
{"id": "9fa00ec29b914b8649045e7776c96cb78f5e21fdf831d401f24767c0b60b851f", "sources": ["arxiv", "semantic_scholar"], "title": "SpeechAgents: Human-Communication Simulation with Multi-Modal Multi-Agent Systems", "abstract": "Human communication is a complex and diverse process that not only involves multiple factors such as language, commonsense, and cultural backgrounds but also requires the participation of multimodal information, such as speech. Large Language Model (LLM)-based multi-agent systems have demonstrated promising performance in simulating human society. Can we leverage LLM-based multi-agent systems to simulate human communication? However, current LLM-based multi-agent systems mainly rely on text as the primary medium. In this paper, we propose SpeechAgents, a multi-modal LLM based multi-agent system designed for simulating human communication. SpeechAgents utilizes multi-modal LLM as the control center for individual agent and employes multi-modal signals as the medium for exchanged messages among agents. Additionally, we propose Multi-Agent Tuning to enhance the multi-agent capabilities of LLM without compromising general abilities. To strengthen and evaluate the effectiveness of human communication simulation, we build the Human-Communication Simulation Benchmark. Experimental results demonstrate that SpeechAgents can simulate human communication dialogues with consistent content, authentic rhythm, and rich emotions and demonstrate excellent scalability even with up to 25 agents, which can apply to tasks such as drama creation and audio novels generation. Code and models will be open-sourced at https://github. com/0nutation/SpeechAgents", "authors": ["Dong Zhang", "Zhaowei Li", "Pengyu Wang", "Xin Zhang", "Yaqian Zhou", "Xipeng Qiu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-08", "url": "https://arxiv.org/abs/2401.03945", "pdf_url": "https://arxiv.org/pdf/2401.03945v1", "arxiv_id": "2401.03945", "doi": "10.48550/arXiv.2401.03945", "citation_count": 5, "influential_citation_count": 2, "has_code": true, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "fbe4d825ea3b1bff87b6a3641bf167183d6ad9783370276f821909f387c51260", "sources": ["arxiv", "semantic_scholar"], "title": "Decentralized Multi-Agent Active Search and Tracking when Targets Outnumber Agents", "abstract": "Multi-agent multi-target tracking has a wide range of applications, including wildlife patrolling, security surveillance or environment monitoring. Such algorithms often make restrictive assumptions: the number of targets and/or their initial locations may be assumed known, or agents may be pre-assigned to monitor disjoint partitions of the environment, reducing the burden of exploration. This also limits applicability when there are fewer agents than targets, since agents are unable to continuously follow the targets in their fields of view. Multi-agent tracking algorithms additionally assume inter-agent synchronization of observations, or the presence of a central controller to coordinate joint actions. Instead, we focus on the setting of decentralized multi-agent, multi-target, simultaneous active search-and-tracking with asynchronous inter-agent communication. Our proposed algorithm DecSTER uses a sequential monte carlo implementation of the probability hypothesis density filter for posterior inference combined with Thompson sampling for decentralized multi-agent decision making. We compare different action selection policies, focusing on scenarios where targets outnumber agents. In simulation, we demonstrate that DecSTER is robust to unreliable inter-agent communication and outperforms information-greedy baselines in terms of the Optimal Sub-Pattern Assignment (OSPA) metric for different numbers of targets and varying teamsizes.", "authors": ["Arundhati Banerjee", "Jeff Schneider"], "categories": ["cs.RO", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-06", "url": "https://arxiv.org/abs/2401.03154", "pdf_url": "https://arxiv.org/pdf/2401.03154v2", "arxiv_id": "2401.03154", "doi": "10.1109/ICRA57147.2024.10609977", "citation_count": 9, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE International Conference on Robotics and Automation", "quality_score": 0.25}
{"id": "82772fc9f4952428edd2c3e5a2299abbe9c8b4b49ca1d42d0324a4e730f7fc0c", "sources": ["arxiv", "semantic_scholar"], "title": "From LLM to Conversational Agent: A Memory Enhanced Architecture with Fine-Tuning of Large Language Models", "abstract": "This paper introduces RAISE (Reasoning and Acting through Scratchpad and Examples), an advanced architecture enhancing the integration of Large Language Models (LLMs) like GPT-4 into conversational agents. RAISE, an enhancement of the ReAct framework, incorporates a dual-component memory system, mirroring human short-term and long-term memory, to maintain context and continuity in conversations. It entails a comprehensive agent construction scenario, including phases like Conversation Selection, Scene Extraction, CoT Completion, and Scene Augmentation, leading to the LLMs Training phase. This approach appears to enhance agent controllability and adaptability in complex, multi-turn dialogues. Our preliminary evaluations in a real estate sales context suggest that RAISE has some advantages over traditional agents, indicating its potential for broader applications. This work contributes to the AI field by providing a robust framework for developing more context-aware and versatile conversational agents.", "authors": ["Na Liu", "Liangyu Chen", "Xiaoyu Tian", "Wei Zou", "Kaijiang Chen", "Ming Cui"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-05", "url": "https://arxiv.org/abs/2401.02777", "pdf_url": "https://arxiv.org/pdf/2401.02777v2", "arxiv_id": "2401.02777", "doi": "10.48550/arXiv.2401.02777", "citation_count": 49, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4247}
{"id": "5a8c2913200cad24cd251dc53d36b5cfd49aa0c90a960601fb1630f38d2abc1b", "sources": ["arxiv", "semantic_scholar"], "title": "LLM Harmony: Multi-Agent Communication for Problem Solving", "abstract": "Large Language Models (LLMs) have revolutionized Natural Language Processing but exhibit limitations, particularly in autonomously addressing novel challenges such as reasoning and problem-solving. Traditional techniques like chain-of-thought prompting necessitate explicit human guidance. This paper introduces a novel multi-agent communication framework, inspired by the CAMEL model, to enhance LLMs' autonomous problem-solving capabilities. The framework employs multiple LLM agents, each with a distinct persona, engaged in role-playing communication, offering a nuanced and adaptable approach to diverse problem scenarios. Extensive experimentation demonstrates the framework's superior performance and adaptability, providing valuable insights into the collaborative potential of multiple agents in overcoming the limitations of individual models.", "authors": ["Sumedh Rasal"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2024-01-02", "url": "https://arxiv.org/abs/2401.01312", "pdf_url": "https://arxiv.org/pdf/2401.01312v1", "arxiv_id": "2401.01312", "doi": "10.48550/arXiv.2401.01312", "citation_count": 49, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4247}
{"id": "5e523322363a1dd797bafb2d449c2733da0d813085c695cfc4b813658987aecf", "sources": ["arxiv", "semantic_scholar"], "title": "TAPE: Leveraging Agent Topology for Cooperative Multi-Agent Policy Gradient", "abstract": "Multi-Agent Policy Gradient (MAPG) has made significant progress in recent years. However, centralized critics in state-of-the-art MAPG methods still face the centralized-decentralized mismatch (CDM) issue, which means sub-optimal actions by some agents will affect other agent's policy learning. While using individual critics for policy updates can avoid this issue, they severely limit cooperation among agents. To address this issue, we propose an agent topology framework, which decides whether other agents should be considered in policy gradient and achieves compromise between facilitating cooperation and alleviating the CDM issue. The agent topology allows agents to use coalition utility as learning objective instead of global utility by centralized critics or local utility by individual critics. To constitute the agent topology, various models are studied. We propose Topology-based multi-Agent Policy gradiEnt (TAPE) for both stochastic and deterministic MAPG methods. We prove the policy improvement theorem for stochastic TAPE and give a theoretical explanation for the improved cooperation among agents. Experiment results on several benchmarks show the agent topology is able to facilitate agent cooperation and alleviate CDM issue respectively to improve performance of TAPE. Finally, multiple ablation studies and a heuristic graph search algorithm are devised to show the efficacy of the agent topology.", "authors": ["Xingzhou Lou", "Junge Zhang", "Timothy J. Norman", "Kaiqi Huang", "Yali Du"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-25", "url": "https://arxiv.org/abs/2312.15667", "pdf_url": "https://arxiv.org/pdf/2312.15667v3", "arxiv_id": "2312.15667", "doi": "10.48550/arXiv.2312.15667", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.1193}
{"id": "7c451043a262de5f04aebfeeac1664abe5dde6b3f83fdbc3d10c01122ee29ea3", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Task Multi-Agent Shared Layers are Universal Cognition of Multi-Agent Coordination", "abstract": "Multi-agent reinforcement learning shines as the pinnacle of multi-agent systems, conquering intricate real-world challenges, fostering collaboration and coordination among agents, and unleashing the potential for intelligent decision-making across domains. However, training a multi-agent reinforcement learning network is a formidable endeavor, demanding substantial computational resources to interact with diverse environmental variables, extract state representations, and acquire decision-making knowledge. The recent breakthroughs in large-scale pre-trained models ignite our curiosity: Can we uncover shared knowledge in multi-agent reinforcement learning and leverage pre-trained models to expedite training for future tasks? Addressing this issue, we present an innovative multi-task learning approach that aims to extract and harness common decision-making knowledge, like cooperation and competition, across different tasks. Our approach involves concurrent training of multiple multi-agent tasks, with each task employing independent front-end perception layers while sharing back-end decision-making layers. This effective decoupling of state representation extraction from decision-making allows for more efficient training and better transferability. To evaluate the efficacy of our proposed approach, we conduct comprehensive experiments in two distinct environments: the StarCraft Multi-agent Challenge (SMAC) and the Google Research Football (GRF) environments. The experimental results unequivocally demonstrate the smooth transferability of the shared decision-making network to other tasks, thereby significantly reducing training costs and improving final performance. Furthermore, visualizations authenticate the presence of general multi-agent decision-making knowledge within the shared network layers, further validating the effectiveness of our approach.", "authors": ["Jiawei Wang", "Jian Zhao", "Zhengtao Cao", "Ruili Feng", "Rongjun Qin", "Yang Yu"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-25", "url": "https://arxiv.org/abs/2312.15674", "pdf_url": "https://arxiv.org/pdf/2312.15674v1", "arxiv_id": "2312.15674", "doi": "10.48550/arXiv.2312.15674", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1747}
{"id": "b678be27464550f8fb840b088b3ecb1fa111159564741791ef96d1a887552433", "sources": ["arxiv", "semantic_scholar"], "title": "ConcaveQ: Non-Monotonic Value Function Factorization via Concave Representations in Deep Multi-Agent Reinforcement Learning", "abstract": "Value function factorization has achieved great success in multi-agent reinforcement learning by optimizing joint action-value functions through the maximization of factorized per-agent utilities. To ensure Individual-Global-Maximum property, existing works often focus on value factorization using monotonic functions, which are known to result in restricted representation expressiveness. In this paper, we analyze the limitations of monotonic factorization and present ConcaveQ, a novel non-monotonic value function factorization approach that goes beyond monotonic mixing functions and employs neural network representations of concave mixing functions. Leveraging the concave property in factorization, an iterative action selection scheme is developed to obtain optimal joint actions during training. It is used to update agents' local policy networks, enabling fully decentralized execution. The effectiveness of the proposed ConcaveQ is validated across scenarios involving multi-agent predator-prey environment and StarCraft II micromanagement tasks. Empirical results exhibit significant improvement of ConcaveQ over state-of-the-art multi-agent reinforcement learning approaches.", "authors": ["Huiqun Li", "Hanhan Zhou", "Yifei Zou", "Dongxiao Yu", "Tian Lan"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-24", "url": "https://arxiv.org/abs/2312.15555", "pdf_url": "https://arxiv.org/pdf/2312.15555v1", "arxiv_id": "2312.15555", "doi": "10.48550/arXiv.2312.15555", "citation_count": 19, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.3253}
{"id": "15f67a885e990c80f49677e4a7854825226fee8721e4c7f62cd26a3edde05727", "sources": ["arxiv", "semantic_scholar"], "title": "AgentCoder: Multi-Agent-based Code Generation with Iterative Testing and Optimisation", "abstract": "The advancement of natural language processing (NLP) has been significantly boosted by the development of transformer-based large language models (LLMs). These models have revolutionized NLP tasks, particularly in code generation, aiding developers in creating software with enhanced efficiency. Despite their advancements, challenges in balancing code snippet generation with effective test case generation and execution persist. To address these issues, this paper introduces Multi-Agent Assistant Code Generation (AgentCoder), a novel solution comprising a multi-agent framework with specialized agents: the programmer agent, the test designer agent, and the test executor agent. During the coding procedure, the programmer agent will focus on the code generation and refinement based on the test executor agent's feedback. The test designer agent will generate test cases for the generated code, and the test executor agent will run the code with the test cases and write the feedback to the programmer. This collaborative system ensures robust code generation, surpassing the limitations of single-agent models and traditional methodologies. Our extensive experiments on 9 code generation models and 12 enhancement approaches showcase AgentCoder's superior performance over existing code generation models and prompt engineering techniques across various benchmarks. For example, AgentCoder (GPT-4) achieves 96.3\\% and 91.8\\% pass@1 in HumanEval and MBPP datasets with an overall token overhead of 56.9K and 66.3K, while state-of-the-art obtains only 90.2\\% and 78.9\\% pass@1 with an overall token overhead of 138.2K and 206.5K.", "authors": ["Dong Huang", "Jie M. Zhang", "Michael Luck", "Qingwen Bu", "Yuhao Qing", "Heming Cui"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-20", "url": "https://arxiv.org/abs/2312.13010", "pdf_url": "https://arxiv.org/pdf/2312.13010v3", "arxiv_id": "2312.13010", "doi": null, "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0}
{"id": "951440c395b02fb2b0e52460803bc63fe7631d912249af34a1b18d585fa12d95", "sources": ["arxiv", "semantic_scholar"], "title": "Curriculum Learning for Cooperation in Multi-Agent Reinforcement Learning", "abstract": "While there has been significant progress in curriculum learning and continuous learning for training agents to generalize across a wide variety of environments in the context of single-agent reinforcement learning, it is unclear if these algorithms would still be valid in a multi-agent setting. In a competitive setting, a learning agent can be trained by making it compete with a curriculum of increasingly skilled opponents. However, a general intelligent agent should also be able to learn to act around other agents and cooperate with them to achieve common goals. When cooperating with other agents, the learning agent must (a) learn how to perform the task (or subtask), and (b) increase the overall team reward. In this paper, we aim to answer the question of what kind of cooperative teammate, and a curriculum of teammates should a learning agent be trained with to achieve these two objectives. Our results on the game Overcooked show that a pre-trained teammate who is less skilled is the best teammate for overall team reward but the worst for the learning of the agent. Moreover, somewhat surprisingly, a curriculum of teammates with decreasing skill levels performs better than other types of curricula.", "authors": ["Rupali Bhati", "Sai Krishna Gottipati", "Clodéric Mars", "Matthew E. Taylor"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-19", "url": "https://arxiv.org/abs/2312.11768", "pdf_url": "https://arxiv.org/pdf/2312.11768v1", "arxiv_id": "2312.11768", "doi": "10.48550/arXiv.2312.11768", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "fb3c0230473ef991830c790ecbc7fd6dad98d26798e82edf3cd9c9a3f8eb9081", "sources": ["arxiv", "semantic_scholar"], "title": "MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQL", "abstract": "Recent LLM-based Text-to-SQL methods usually suffer from significant performance degradation on \"huge\" databases and complex user questions that require multi-step reasoning. Moreover, most existing methods neglect the crucial significance of LLMs utilizing external tools and model collaboration. To address these challenges, we introduce MAC-SQL, a novel LLM-based multi-agent collaborative framework. Our framework comprises a core decomposer agent for Text-to-SQL generation with few-shot chain-of-thought reasoning, accompanied by two auxiliary agents that utilize external tools or models to acquire smaller sub-databases and refine erroneous SQL queries. The decomposer agent collaborates with auxiliary agents, which are activated as needed and can be expanded to accommodate new features or tools for effective Text-to-SQL parsing. In our framework, We initially leverage GPT-4 as the strong backbone LLM for all agent tasks to determine the upper bound of our framework. We then fine-tune an open-sourced instruction-followed model, SQL-Llama, by leveraging Code Llama 7B, to accomplish all tasks as GPT-4 does. Experiments show that SQL-Llama achieves a comparable execution accuracy of 43.94, compared to the baseline accuracy of 46.35 for vanilla GPT-4. At the time of writing, MAC-SQL+GPT-4 achieves an execution accuracy of 59.59 when evaluated on the BIRD benchmark, establishing a new state-of-the-art (SOTA) on its holdout test set (https://github.com/wbbeyourself/MAC-SQL).", "authors": ["Bing Wang", "Changyu Ren", "Jian Yang", "Xinnian Liang", "Jiaqi Bai", "LinZheng Chai", "Zhao Yan", "Qian-Wen Zhang", "Di Yin", "Xing Sun", "Zhoujun Li"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-18", "url": "https://arxiv.org/abs/2312.11242", "pdf_url": "https://arxiv.org/pdf/2312.11242v6", "arxiv_id": "2312.11242", "doi": "10.48550/arXiv.2312.11242", "citation_count": 222, "influential_citation_count": 42, "has_code": true, "code_url": "https://github.com/wbbeyourself/MAC-SQL", "venue": "International Conference on Computational Linguistics", "quality_score": 0.8167}
{"id": "331e40322597dff7f499b24dbed4ae678b3a0bd6b05f16824e27c39c2a040fa6", "sources": ["arxiv", "semantic_scholar"], "title": "ReST meets ReAct: Self-Improvement for Multi-Step Reasoning LLM Agent", "abstract": "Answering complex natural language questions often necessitates multi-step reasoning and integrating external information. Several systems have combined knowledge retrieval with a large language model (LLM) to answer such questions. These systems, however, suffer from various failure cases, and we cannot directly train them end-to-end to fix such failures, as interaction with external knowledge is non-differentiable. To address these deficiencies, we define a ReAct-style LLM agent with the ability to reason and act upon external knowledge. We further refine the agent through a ReST-like method that iteratively trains on previous trajectories, employing growing-batch reinforcement learning with AI feedback for continuous self-improvement and self-distillation. Starting from a prompted large model and after just two iterations of the algorithm, we can produce a fine-tuned small model that achieves comparable performance on challenging compositional question-answering benchmarks with two orders of magnitude fewer parameters.", "authors": ["Renat Aksitov", "Sobhan Miryoosefi", "Zonglin Li", "Daliang Li", "Sheila Babayan", "Kavya Kopparapu", "Zachary Fisher", "Ruiqi Guo", "Sushant Prakash", "Pranesh Srinivasan", "Manzil Zaheer", "Felix Yu", "Sanjiv Kumar"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-15", "url": "https://arxiv.org/abs/2312.10003", "pdf_url": "https://arxiv.org/pdf/2312.10003v1", "arxiv_id": "2312.10003", "doi": "10.48550/arXiv.2312.10003", "citation_count": 84, "influential_citation_count": 9, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.5}
{"id": "c2311b8823fde4af6a653ed96fee2807ea8ac1d1b351dd8ac64598f9257445c9", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-MARS: Large Language Model for Behavior Tree Generation and NLP-enhanced Dialogue in Multi-Agent Robot Systems", "abstract": "This paper introduces LLM-MARS, first technology that utilizes a Large Language Model based Artificial Intelligence for Multi-Agent Robot Systems. LLM-MARS enables dynamic dialogues between humans and robots, allowing the latter to generate behavior based on operator commands and provide informative answers to questions about their actions. LLM-MARS is built on a transformer-based Large Language Model, fine-tuned from the Falcon 7B model. We employ a multimodal approach using LoRa adapters for different tasks. The first LoRa adapter was developed by fine-tuning the base model on examples of Behavior Trees and their corresponding commands. The second LoRa adapter was developed by fine-tuning on question-answering examples. Practical trials on a multi-agent system of two robots within the Eurobot 2023 game rules demonstrate promising results. The robots achieve an average task execution accuracy of 79.28% in compound commands. With commands containing up to two tasks accuracy exceeded 90%. Evaluation confirms the system's answers on operators questions exhibit high accuracy, relevance, and informativeness. LLM-MARS and similar multi-agent robotic systems hold significant potential to revolutionize logistics, enabling autonomous exploration missions and advancing Industry 5.0.", "authors": ["Artem Lykov", "Maria Dronova", "Nikolay Naglov", "Mikhail Litvinov", "Sergei Satsevich", "Artem Bazhenov", "Vladimir Berman", "Aleksei Shcherbak", "Dzmitry Tsetserukou"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-14", "url": "https://arxiv.org/abs/2312.09348", "pdf_url": "https://arxiv.org/pdf/2312.09348v1", "arxiv_id": "2312.09348", "doi": "10.48550/arXiv.2312.09348", "citation_count": 29, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.3693}
{"id": "f033cad790ee21a918b6beceec9d8c64ccac4b278cf5c4878523c7251913e005", "sources": ["arxiv", "semantic_scholar"], "title": "Large Language Model Enhanced Multi-Agent Systems for 6G Communications", "abstract": "The rapid development of the Large Language Model (LLM) presents huge opportunities for 6G communications, e.g., network optimization and management by allowing users to input task requirements to LLMs by nature language. However, directly applying native LLMs in 6G encounters various challenges, such as a lack of private communication data and knowledge, limited logical reasoning, evaluation, and refinement abilities. Integrating LLMs with the capabilities of retrieval, planning, memory, evaluation and reflection in agents can greatly enhance the potential of LLMs for 6G communications. To this end, we propose a multi-agent system with customized communication knowledge and tools for solving communication related tasks using natural language, comprising three components: (1) Multi-agent Data Retrieval (MDR), which employs the condensate and inference agents to refine and summarize communication knowledge from the knowledge base, expanding the knowledge boundaries of LLMs in 6G communications; (2) Multi-agent Collaborative Planning (MCP), which utilizes multiple planning agents to generate feasible solutions for the communication related task from different perspectives based on the retrieved knowledge; (3) Multi-agent Evaluation and Reflecxion (MER), which utilizes the evaluation agent to assess the solutions, and applies the reflexion agent and refinement agent to provide improvement suggestions for current solutions. Finally, we validate the effectiveness of the proposed multi-agent system by designing a semantic communication system, as a case study of 6G communications.", "authors": ["Feibo Jiang", "Li Dong", "Yubo Peng", "Kezhi Wang", "Kun Yang", "Cunhua Pan", "Dusit Niyato", "Octavia A. Dobre"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-13", "url": "https://arxiv.org/abs/2312.07850", "pdf_url": "https://arxiv.org/pdf/2312.07850v1", "arxiv_id": "2312.07850", "doi": "10.1109/MWC.016.2300600", "citation_count": 173, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "IEEE wireless communications", "quality_score": 0.5601}
{"id": "2c212e936a2dc1fb8e86f9426002272b6ee3c4eca8bbf6a6c223b4b8b3466173", "sources": ["arxiv", "semantic_scholar"], "title": "DCIR: Dynamic Consistency Intrinsic Reward for Multi-Agent Reinforcement Learning", "abstract": "Learning optimal behavior policy for each agent in multi-agent systems is an essential yet difficult problem. Despite fruitful progress in multi-agent reinforcement learning, the challenge of addressing the dynamics of whether two agents should exhibit consistent behaviors is still under-explored. In this paper, we propose a new approach that enables agents to learn whether their behaviors should be consistent with that of other agents by utilizing intrinsic rewards to learn the optimal policy for each agent. We begin by defining behavior consistency as the divergence in output actions between two agents when provided with the same observation. Subsequently, we introduce dynamic consistency intrinsic reward (DCIR) to stimulate agents to be aware of others' behaviors and determine whether to be consistent with them. Lastly, we devise a dynamic scale network (DSN) that provides learnable scale factors for the agent at every time step to dynamically ascertain whether to award consistent behavior and the magnitude of rewards. We evaluate DCIR in multiple environments including Multi-agent Particle, Google Research Football and StarCraft II Micromanagement, demonstrating its efficacy.", "authors": ["Kunyang Lin", "Yufeng Wang", "Peihao Chen", "Runhao Zeng", "Siyuan Zhou", "Mingkui Tan", "Chuang Gan"], "categories": ["cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-10", "url": "https://arxiv.org/abs/2312.05783", "pdf_url": "https://arxiv.org/pdf/2312.05783v1", "arxiv_id": "2312.05783", "doi": "10.48550/arXiv.2312.05783", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "61ca336f6f3dd83bcc3b625e4a794eb9adafcbafa56eaf956acce902bc4e2ab9", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Reinforcement Learning via Distributed MPC as a Function Approximator", "abstract": "This paper presents a novel approach to multi-agent reinforcement learning (RL) for linear systems with convex polytopic constraints. Existing work on RL has demonstrated the use of model predictive control (MPC) as a function approximator for the policy and value functions. The current paper is the first work to extend this idea to the multi-agent setting. We propose the use of a distributed MPC scheme as a function approximator, with a structure allowing for distributed learning and deployment. We then show that Q-learning updates can be performed distributively without introducing nonstationarity, by reconstructing a centralized learning update. The effectiveness of the approach is demonstrated on two numerical examples.", "authors": ["Samuel Mallick", "Filippo Airaldi", "Azita Dabiri", "Bart De Schutter"], "categories": ["eess.SY"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2023-12-08", "url": "https://arxiv.org/abs/2312.05166", "pdf_url": "https://arxiv.org/pdf/2312.05166v4", "arxiv_id": "2312.05166", "doi": "10.1016/j.automatica.2024.111803", "citation_count": 13, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/SamuelMallick/dmpcrl-concept/tree/paper-2023", "venue": null, "quality_score": 0.2865}
{"id": "23188aee3b3d2312fd1fbc09db502255ca58966a190177f93b8fa2bc4e769996", "sources": ["arxiv", "semantic_scholar"], "title": "An LLM Compiler for Parallel Function Calling", "abstract": "The reasoning capabilities of the recent LLMs enable them to execute external function calls to overcome their inherent limitations, such as knowledge cutoffs, poor arithmetic skills, or lack of access to private data. This development has allowed LLMs to select and coordinate multiple functions based on the context to tackle more complex problems. However, current methods for function calling often require sequential reasoning and acting for each function which can result in high latency, cost, and sometimes inaccurate behavior. To address this, we introduce LLMCompiler, which executes functions in parallel to efficiently orchestrate multiple function calls. Drawing inspiration from the principles of classical compilers, LLMCompiler enables parallel function calling with three components: (i) a Function Calling Planner, formulating execution plans for function calling; (ii) a Task Fetching Unit, dispatching function calling tasks; and (iii) an Executor, executing these tasks in parallel. LLMCompiler automatically generates an optimized orchestration for the function calls and can be used with both open-source and closed-source models. We have benchmarked LLMCompiler on a range of tasks with different patterns of function calling. We observe consistent latency speedup of up to 3.7x, cost savings of up to 6.7x, and accuracy improvement of up to ~9% compared to ReAct. Our code is available at https://github.com/SqueezeAILab/LLMCompiler.", "authors": ["Sehoon Kim", "Suhong Moon", "Ryan Tabrizi", "Nicholas Lee", "Michael W. Mahoney", "Kurt Keutzer", "Amir Gholami"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-07", "url": "https://arxiv.org/abs/2312.04511", "pdf_url": "https://arxiv.org/pdf/2312.04511v3", "arxiv_id": "2312.04511", "doi": "10.48550/arXiv.2312.04511", "citation_count": 163, "influential_citation_count": 10, "has_code": true, "code_url": "https://github.com/SqueezeAILab/LLMCompiler", "venue": "International Conference on Machine Learning", "quality_score": 0.5537}
{"id": "eab2b0ab956e9d49265036142d1be644934901dfe45086164b009ea4e40d91c5", "sources": ["arxiv", "semantic_scholar"], "title": "LLM as OS, Agents as Apps: Envisioning AIOS, Agents and the AIOS-Agent Ecosystem", "abstract": "This paper envisions a revolutionary AIOS-Agent ecosystem, where Large Language Model (LLM) serves as the (Artificial) Intelligent Operating System (IOS, or AIOS)--an operating system \"with soul\". Upon this foundation, a diverse range of LLM-based AI Agent Applications (Agents, or AAPs) are developed, enriching the AIOS-Agent ecosystem and signaling a paradigm shift from the traditional OS-APP ecosystem. We envision that LLM's impact will not be limited to the AI application level, instead, it will in turn revolutionize the design and implementation of computer system, architecture, software, and programming language, featured by several main concepts: LLM as OS (system-level), Agents as Applications (application-level), Natural Language as Programming Interface (user-level), and Tools as Devices/Libraries (hardware/middleware-level). We begin by introducing the architecture of traditional OS. Then we formalize a conceptual framework for AIOS through \"LLM as OS (LLMOS)\", drawing analogies between AIOS and traditional OS: LLM is likened to OS kernel, context window to memory, external storage to file system, hardware tools to peripheral devices, software tools to programming libraries, and user prompts to user commands. Subsequently, we introduce the new AIOS-Agent Ecosystem, where users can easily program Agent Applications (AAPs) using natural language, democratizing the development of software, which is different from the traditional OS-APP ecosystem. Following this, we explore the diverse scope of Agent Applications. We delve into both single-agent and multi-agent systems, as well as human-agent interaction. Lastly, drawing on the insights from traditional OS-APP ecosystem, we propose a roadmap for the evolution of the AIOS-Agent ecosystem. This roadmap is designed to guide the future research and development, suggesting systematic progresses of AIOS and its Agent applications.", "authors": ["Yingqiang Ge", "Yujie Ren", "Wenyue Hua", "Shuyuan Xu", "Juntao Tan", "Yongfeng Zhang"], "categories": ["cs.OS", "cs.AI", "cs.CL", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-06", "url": "https://arxiv.org/abs/2312.03815", "pdf_url": "https://arxiv.org/pdf/2312.03815v2", "arxiv_id": "2312.03815", "doi": "10.48550/arXiv.2312.03815", "citation_count": 43, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4109}
{"id": "7588058d50f7ca2939fd9e798aee85579855db1b8551cf37ae493cf6f1df3636", "sources": ["arxiv", "semantic_scholar"], "title": "Deciphering Digital Detectives: Understanding LLM Behaviors and Capabilities in Multi-Agent Mystery Games", "abstract": "In this study, we explore the application of Large Language Models (LLMs) in \\textit{Jubensha}, a Chinese detective role-playing game and a novel area in Artificial Intelligence (AI) driven gaming. We introduce the first dataset specifically for Jubensha, including character scripts and game rules, to foster AI agent development in this complex narrative environment. Our work also presents a unique multi-agent interaction framework using LLMs, allowing AI agents to autonomously engage in this game. To evaluate the gaming performance of these AI agents, we developed novel methods measuring their mastery of case information and reasoning skills. Furthermore, we incorporated the latest advancements in in-context learning to improve the agents' performance in information gathering, murderer identification, and logical reasoning. The experimental results validate the effectiveness of our proposed methods. This work aims to offer a novel perspective on understanding LLM capabilities and establish a new benchmark for evaluating large language model-based agents.", "authors": ["Dekun Wu", "Haochen Shi", "Zhiyuan Sun", "Bang Liu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-12-01", "url": "https://arxiv.org/abs/2312.00746", "pdf_url": "https://arxiv.org/pdf/2312.00746v2", "arxiv_id": "2312.00746", "doi": "10.48550/arXiv.2312.00746", "citation_count": 39, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Annual Meeting of the Association for Computational Linguistics", "quality_score": 0.4005}
{"id": "cfdf43324070fc6738f470ac4090be1ecf3046249026a9110713d78b7f07d041", "sources": ["arxiv", "semantic_scholar"], "title": "Should we be going MAD? A Look at Multi-Agent Debate Strategies for LLMs", "abstract": "Recent advancements in large language models (LLMs) underscore their potential for responding to inquiries in various domains. However, ensuring that generative agents provide accurate and reliable answers remains an ongoing challenge. In this context, multi-agent debate (MAD) has emerged as a promising strategy for enhancing the truthfulness of LLMs. We benchmark a range of debating and prompting strategies to explore the trade-offs between cost, time, and accuracy. Importantly, we find that multi-agent debating systems, in their current form, do not reliably outperform other proposed prompting strategies, such as self-consistency and ensembling using multiple reasoning paths. However, when performing hyperparameter tuning, several MAD systems, such as Multi-Persona, perform better. This suggests that MAD protocols might not be inherently worse than other approaches, but that they are more sensitive to different hyperparameter settings and difficult to optimize. We build on these results to offer insights into improving debating strategies, such as adjusting agent agreement levels, which can significantly enhance performance and even surpass all other non-debate protocols we evaluated. We provide an open-source repository to the community with several state-of-the-art protocols together with evaluation scripts to benchmark across popular research datasets.", "authors": ["Andries Smit", "Paul Duckworth", "Nathan Grinsztajn", "Thomas D. Barrett", "Arnu Pretorius"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-29", "url": "https://arxiv.org/abs/2311.17371", "pdf_url": "https://arxiv.org/pdf/2311.17371v3", "arxiv_id": "2311.17371", "doi": null, "citation_count": 98, "influential_citation_count": 5, "has_code": true, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.4989}
{"id": "ad6f54783d336d4c6ce0bda21168f68edce78cfc745967d258478e34a8ed10df", "sources": ["arxiv", "semantic_scholar"], "title": "Towards Top-Down Reasoning: An Explainable Multi-Agent Approach for Visual Question Answering", "abstract": "Recently, to comprehensively improve Vision Language Models (VLMs) for Visual Question Answering (VQA), several methods have been proposed to further reinforce the inference capabilities of VLMs to independently tackle VQA tasks rather than some methods that only utilize VLMs as aids to Large Language Models (LLMs). However, these methods ignore the rich common-sense knowledge inside the given VQA image sampled from the real world. Thus, they cannot fully use the powerful VLM for the given VQA question to achieve optimal performance. Attempt to overcome this limitation and inspired by the human top-down reasoning process, i.e., systematically exploring relevant issues to derive a comprehensive answer, this work introduces a novel, explainable multi-agent collaboration framework by leveraging the expansive knowledge of Large Language Models (LLMs) to enhance the capabilities of VLMs themselves. Specifically, our framework comprises three agents, i.e., Responder, Seeker, and Integrator, to collaboratively answer the given VQA question by seeking its relevant issues and generating the final answer in such a top-down reasoning process. The VLM-based Responder agent generates the answer candidates for the question and responds to other relevant issues. The Seeker agent, primarily based on LLM, identifies relevant issues related to the question to inform the Responder agent and constructs a Multi-View Knowledge Base (MVKB) for the given visual scene by leveraging the build-in world knowledge of LLM. The Integrator agent combines knowledge from the Seeker agent and the Responder agent to produce the final VQA answer. Extensive and comprehensive evaluations on diverse VQA datasets with a variety of VLMs demonstrate the superior performance and interpretability of our framework over the baseline method in the zero-shot setting without extra training cost.", "authors": ["Zeqing Wang", "Wentao Wan", "Qiqing Lao", "Runmeng Chen", "Minjie Lang", "Xiao Wang", "Keze Wang", "Liang Lin"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-29", "url": "https://arxiv.org/abs/2311.17331", "pdf_url": "https://arxiv.org/pdf/2311.17331v4", "arxiv_id": "2311.17331", "doi": "10.48550/arXiv.2311.17331", "citation_count": 13, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2865}
{"id": "47981ef162aa630cbb70a780ce6f775d6e37b9d690f4fa2082dadbcfc48dac5f", "sources": ["arxiv", "semantic_scholar"], "title": "Embodied Multi-Modal Agent trained by an LLM from a Parallel TextWorld", "abstract": "While large language models (LLMs) excel in a simulated world of texts, they struggle to interact with the more realistic world without perceptions of other modalities such as visual or audio signals. Although vision-language models (VLMs) integrate LLM modules (1) aligned with static image features, and (2) may possess prior knowledge of world dynamics (as demonstrated in the text world), they have not been trained in an embodied visual world and thus cannot align with its dynamics. On the other hand, training an embodied agent in a noisy visual world without expert guidance is often challenging and inefficient. In this paper, we train a VLM agent living in a visual world using an LLM agent excelling in a parallel text world. Specifically, we distill LLM's reflection outcomes (improved actions by analyzing mistakes) in a text world's tasks to finetune the VLM on the same tasks of the visual world, resulting in an Embodied Multi-Modal Agent (EMMA) quickly adapting to the visual world dynamics. Such cross-modality imitation learning between the two parallel worlds is achieved by a novel DAgger-DPO algorithm, enabling EMMA to generalize to a broad scope of new tasks without any further guidance from the LLM expert. Extensive evaluations on the ALFWorld benchmark's diverse tasks highlight EMMA's superior performance to SOTA VLM-based agents, e.g., 20%-70% improvement in the success rate.", "authors": ["Yijun Yang", "Tianyi Zhou", "Kanxue Li", "Dapeng Tao", "Lusong Li", "Li Shen", "Xiaodong He", "Jing Jiang", "Yuhui Shi"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-28", "url": "https://arxiv.org/abs/2311.16714", "pdf_url": "https://arxiv.org/pdf/2311.16714v2", "arxiv_id": "2311.16714", "doi": "10.1109/CVPR52733.2024.02482", "citation_count": 94, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "Computer Vision and Pattern Recognition", "quality_score": 0.4944}
{"id": "ff0a6cdc9fa0663d63bc6042ca96f745222a790f1153b84f82389b69d043b1bb", "sources": ["arxiv", "semantic_scholar"], "title": "An Industrial Perspective on Multi-Agent Decision Making for Interoperable Robot Navigation following the VDA5050 Standard", "abstract": "This paper provides a perspective on the literature and current challenges in Multi-Agent Systems for interoperable robot navigation in industry. The focus is on the multi-agent decision stack for Autonomous Mobile Robots operating in mixed environments with humans, manually driven vehicles, and legacy Automated Guided Vehicles. We provide typical characteristics of such Multi-Agent Systems observed today and how these are expected to change on the short term due to the new standard VDA5050 and the interoperability framework OpenRMF. We present recent changes in fleet management standards and the role of open middleware frameworks like ROS2 reaching industrial-grade quality. Approaches to increase the robustness and performance of multi-robot navigation systems for transportation are discussed, and research opportunities are derived.", "authors": ["Niels van Duijkeren", "Luigi Palmieri", "Ralph Lange", "Alexander Kleiner"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-24", "url": "https://arxiv.org/abs/2311.14615", "pdf_url": "https://arxiv.org/pdf/2311.14615v1", "arxiv_id": "2311.14615", "doi": "10.48550/arXiv.2311.14615", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "2fd3a78ca9e1a1f12ccd76aa31857d21db349d87ca9f317a4ac593b5fac28e57", "sources": ["arxiv", "semantic_scholar"], "title": "FinMem: A Performance-Enhanced LLM Trading Agent with Layered Memory and Character Design", "abstract": "Recent advancements in Large Language Models (LLMs) have exhibited notable efficacy in question-answering (QA) tasks across diverse domains. Their prowess in integrating extensive web knowledge has fueled interest in developing LLM-based autonomous agents. While LLMs are efficient in decoding human instructions and deriving solutions by holistically processing historical inputs, transitioning to purpose-driven agents requires a supplementary rational architecture to process multi-source information, establish reasoning chains, and prioritize critical tasks. Addressing this, we introduce \\textsc{FinMem}, a novel LLM-based agent framework devised for financial decision-making. It encompasses three core modules: Profiling, to customize the agent's characteristics; Memory, with layered message processing, to aid the agent in assimilating hierarchical financial data; and Decision-making, to convert insights gained from memories into investment decisions. Notably, \\textsc{FinMem}'s memory module aligns closely with the cognitive structure of human traders, offering robust interpretability and real-time tuning. Its adjustable cognitive span allows for the retention of critical information beyond human perceptual limits, thereby enhancing trading outcomes. This framework enables the agent to self-evolve its professional knowledge, react agilely to new investment cues, and continuously refine trading decisions in the volatile financial environment. We first compare \\textsc{FinMem} with various algorithmic agents on a scalable real-world financial dataset, underscoring its leading trading performance in stocks. We then fine-tuned the agent's perceptual span and character setting to achieve a significantly enhanced trading performance. Collectively, \\textsc{FinMem} presents a cutting-edge LLM agent framework for automated trading, boosting cumulative investment returns.", "authors": ["Yangyang Yu", "Haohang Li", "Zhi Chen", "Yuechen Jiang", "Yang Li", "Denghui Zhang", "Rong Liu", "Jordan W. Suchow", "Khaldoun Khashanah"], "categories": ["q-fin.CP", "cs.AI", "cs.CE", "cs.LG"], "fields_of_study": ["Economics", "Computer Science"], "published_date": "2023-11-23", "url": "https://arxiv.org/abs/2311.13743", "pdf_url": "https://arxiv.org/pdf/2311.13743v2", "arxiv_id": "2311.13743", "doi": "10.1109/TBDATA.2025.3593370", "citation_count": 193, "influential_citation_count": 14, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Big Data", "quality_score": 0.588}
{"id": "455b98ed73eb7761b9ff0af6ff37a1cc3750228d79aa70cb34f63c4aec5c0790", "sources": ["arxiv", "semantic_scholar"], "title": "Equitable Coordination in Multi-agent Power Systems: Impacts of Computation Granularity", "abstract": "The growing integration of distributed energy resources drives the centralized power system towards a decentralized multi-agent network. Operating multi-agent networks significantly relies on inter-agent communications. Computation granularity in this context refers to the number of nodes overseen by an agent. The impact of granularity on equitable power coordination, particularly among marginalized customers with limited communication bandwidth (e.g., intermittent internet connectivity) is not well studied. This work explores different levels of computation granularity for agent-based energy dispatch and studies their impact on equitable coordination. We will leverage and utilize the Consensus + Innovations approach to model the equitable coordination of a multi-agent power system.", "authors": ["Yuhan Du", "Javad Mohammadi"], "categories": ["eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-11-20", "url": "https://arxiv.org/abs/2311.12190", "pdf_url": "https://arxiv.org/pdf/2311.12190v1", "arxiv_id": "2311.12190", "doi": "10.1109/PESGM51994.2024.10688739", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Power & Energy Society General Meeting", "quality_score": 0.1747}
{"id": "da99e3a9b541796729c238715188e7ec07ec41cf30ed19636bc93ec4a3ecf45a", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperative AI via Decentralized Commitment Devices", "abstract": "Credible commitment devices have been a popular approach for robust multi-agent coordination. However, existing commitment mechanisms face limitations like privacy, integrity, and susceptibility to mediator or user strategic behavior. It is unclear if the cooperative AI techniques we study are robust to real-world incentives and attack vectors. However, decentralized commitment devices that utilize cryptography have been deployed in the wild, and numerous studies have shown their ability to coordinate algorithmic agents facing adversarial opponents with significant economic incentives, currently in the order of several million to billions of dollars. In this paper, we use examples in the decentralization and, in particular, Maximal Extractable Value (MEV) (arXiv:1904.05234) literature to illustrate the potential security issues in cooperative AI. We call for expanded research into decentralized commitments to advance cooperative AI capabilities for secure coordination in open environments and empirical testing frameworks to evaluate multi-agent coordination ability given real-world commitment constraints.", "authors": ["Xinyuan Sun", "Davide Crapis", "Matt Stephenson", "Barnabé Monnot", "Thomas Thiery", "Jonathan Passerat-Palmbach"], "categories": ["cs.AI", "cs.CR", "cs.GT", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-14", "url": "https://arxiv.org/abs/2311.07815", "pdf_url": "https://arxiv.org/pdf/2311.07815v1", "arxiv_id": "2311.07815", "doi": "10.48550/arXiv.2311.07815", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.301}
{"id": "980e526d082236d47c67ab600031b99f406250ef0e90ba99ddf159c3d7f14d1a", "sources": ["arxiv", "semantic_scholar"], "title": "Stability analysis for large-scale multi-agent molecular communication systems", "abstract": "Molecular communication (MC) is recently featured as a novel communication tool to connect individual biological nanorobots. It is expected that a large number of nanorobots can form large multi-agent MC systems through MC to accomplish complex and large-scale tasks that cannot be achieved by a single nanorobot. However, most previous models for MC systems assume a unidirectional diffusion communication channel and cannot capture the feedback between each nanorobot, which is important for multi-agent MC systems. In this paper, we introduce a system theoretic model for large-scale multi-agent MC systems using transfer functions, and then propose a method to analyze the stability for multi-agent MC systems. The proposed method decomposes the multi-agent MC system into multiple single-input and single-output (SISO) systems, which facilitates the application of simple analysis technique for SISO systems to the large-scale multi-agent MC system. Finally, we demonstrate the proposed method by analyzing the stability of a specific large-scale multi-agent MC system and clarify a parameter region to synchronize the states of nanorobots, which is important to make cooperative behaviors at a population level.", "authors": ["Taishi Kotsuka", "Yutaka Hori"], "categories": ["eess.SY", "cs.IT", "q-bio.MN"], "fields_of_study": ["Medicine", "Computer Science", "Engineering", "Mathematics", "Biology"], "published_date": "2023-11-12", "url": "https://arxiv.org/abs/2311.06730", "pdf_url": "https://arxiv.org/pdf/2311.06730v2", "arxiv_id": "2311.06730", "doi": "10.1109/TNB.2024.3404592", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Nanobioscience", "quality_score": 0.0753}
{"id": "c8365fea3aa702f4c750f4d386903c35c001dcea100b5ad743b40eaed8d6e13b", "sources": ["arxiv", "semantic_scholar"], "title": "TrainerAgent: Customizable and Efficient Model Training through LLM-Powered Multi-Agent System", "abstract": "Training AI models has always been challenging, especially when there is a need for custom models to provide personalized services. Algorithm engineers often face a lengthy process to iteratively develop models tailored to specific business requirements, making it even more difficult for non-experts. The quest for high-quality and efficient model development, along with the emergence of Large Language Model (LLM) Agents, has become a key focus in the industry. Leveraging the powerful analytical, planning, and decision-making capabilities of LLM, we propose a TrainerAgent system comprising a multi-agent framework including Task, Data, Model and Server agents. These agents analyze user-defined tasks, input data, and requirements (e.g., accuracy, speed), optimizing them comprehensively from both data and model perspectives to obtain satisfactory models, and finally deploy these models as online service. Experimental evaluations on classical discriminative and generative tasks in computer vision and natural language processing domains demonstrate that our system consistently produces models that meet the desired criteria. Furthermore, the system exhibits the ability to critically identify and reject unattainable tasks, such as fantastical scenarios or unethical requests, ensuring robustness and safety. This research presents a significant advancement in achieving desired models with increased efficiency and quality as compared to traditional model development, facilitated by the integration of LLM-powered analysis, decision-making, and execution capabilities, as well as the collaboration among four agents. We anticipate that our work will contribute to the advancement of research on TrainerAgent in both academic and industry communities, potentially establishing it as a new paradigm for model development in the field of AI.", "authors": ["Haoyuan Li", "Hao Jiang", "Tianke Zhang", "Zhelun Yu", "Aoxiong Yin", "Hao Cheng", "Siming Fu", "Yuhao Zhang", "Wanggui He"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-11", "url": "https://arxiv.org/abs/2311.06622", "pdf_url": "https://arxiv.org/pdf/2311.06622v2", "arxiv_id": "2311.06622", "doi": "10.48550/arXiv.2311.06622", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "2f381e709e89eaad8d6644152be3693fee4169fd821506b348d126c6a2b98b67", "sources": ["arxiv", "semantic_scholar"], "title": "Adversarial Attacks on Cooperative Multi-agent Bandits", "abstract": "Cooperative multi-agent multi-armed bandits (CMA2B) consider the collaborative efforts of multiple agents in a shared multi-armed bandit game. We study latent vulnerabilities exposed by this collaboration and consider adversarial attacks on a few agents with the goal of influencing the decisions of the rest. More specifically, we study adversarial attacks on CMA2B in both homogeneous settings, where agents operate with the same arm set, and heterogeneous settings, where agents have distinct arm sets. In the homogeneous setting, we propose attack strategies that, by targeting just one agent, convince all agents to select a particular target arm $T-o(T)$ times while incurring $o(T)$ attack costs in $T$ rounds. In the heterogeneous setting, we prove that a target arm attack requires linear attack costs and propose attack strategies that can force a maximum number of agents to suffer linear regrets while incurring sublinear costs and only manipulating the observations of a few target agents. Numerical experiments validate the effectiveness of our proposed attack strategies.", "authors": ["Jinhang Zuo", "Zhiyao Zhang", "Xuchuang Wang", "Cheng Chen", "Shuai Li", "John C. S. Lui", "Mohammad Hajiesmaili", "Adam Wierman"], "categories": ["cs.LG", "cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-03", "url": "https://arxiv.org/abs/2311.01698", "pdf_url": "https://arxiv.org/pdf/2311.01698v1", "arxiv_id": "2311.01698", "doi": "10.48550/arXiv.2311.01698", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "4959eceebef4f05c9b244678c47787c60a5401136545be5f81e4305defdaaf00", "sources": ["arxiv", "semantic_scholar"], "title": "Responsible Emergent Multi-Agent Behavior", "abstract": "Responsible AI has risen to the forefront of the AI research community. As neural network-based learning algorithms continue to permeate real-world applications, the field of Responsible AI has played a large role in ensuring that such systems maintain a high-level of human-compatibility. Despite this progress, the state of the art in Responsible AI has ignored one crucial point: human problems are multi-agent problems. Predominant approaches largely consider the performance of a single AI system in isolation, but human problems are, by their very nature, multi-agent. From driving in traffic to negotiating economic policy, human problem-solving involves interaction and the interplay of the actions and motives of multiple individuals. This dissertation develops the study of responsible emergent multi-agent behavior, illustrating how researchers and practitioners can better understand and shape multi-agent learning with respect to three pillars of Responsible AI: interpretability, fairness, and robustness. First, I investigate multi-agent interpretability, presenting novel techniques for understanding emergent multi-agent behavior at multiple levels of granularity. With respect to low-level interpretability, I examine the extent to which implicit communication emerges as an aid to coordination in multi-agent populations. I introduce a novel curriculum-driven method for learning high-performing policies in difficult, sparse reward environments and show through a measure of position-based social influence that multi-agent teams that learn sophisticated coordination strategies exchange significantly more information through implicit signals than lesser-coordinated agents. Then, at a high-level, I study concept-based interpretability in the context of multi-agent learning. I propose a novel method for learning intrinsically interpretable, concept-based policies and show that it enables...", "authors": ["Niko A. Grupen"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-02", "url": "https://arxiv.org/abs/2311.01609", "pdf_url": "https://arxiv.org/pdf/2311.01609v1", "arxiv_id": "2311.01609", "doi": "10.48550/arXiv.2311.01609", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "6470db7c42cb08b0e8195cea819c656431b467ffce2aebc0bddb64ce0cc3c609", "sources": ["arxiv", "semantic_scholar"], "title": "QFree: A Universal Value Function Factorization for Multi-Agent Reinforcement Learning", "abstract": "Centralized training is widely utilized in the field of multi-agent reinforcement learning (MARL) to assure the stability of training process. Once a joint policy is obtained, it is critical to design a value function factorization method to extract optimal decentralized policies for the agents, which needs to satisfy the individual-global-max (IGM) principle. While imposing additional limitations on the IGM function class can help to meet the requirement, it comes at the cost of restricting its application to more complex multi-agent environments. In this paper, we propose QFree, a universal value function factorization method for MARL. We start by developing mathematical equivalent conditions of the IGM principle based on the advantage function, which ensures that the principle holds without any compromise, removing the conservatism of conventional methods. We then establish a more expressive mixing network architecture that can fulfill the equivalent factorization. In particular, the novel loss function is developed by considering the equivalent conditions as regularization term during policy evaluation in the MARL algorithm. Finally, the effectiveness of the proposed method is verified in a nonmonotonic matrix game scenario. Moreover, we show that QFree achieves the state-of-the-art performance in a general-purpose complex MARL benchmark environment, Starcraft Multi-Agent Challenge (SMAC).", "authors": ["Rizhong Wang", "Huiping Li", "Di Cui", "Demin Xu"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-11-01", "url": "https://arxiv.org/abs/2311.00356", "pdf_url": "https://arxiv.org/pdf/2311.00356v1", "arxiv_id": "2311.00356", "doi": "10.48550/arXiv.2311.00356", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "3d582b0d692b3cc32ba22b9f4f775987fedef7dac02d6f0d4a2d80e1cee6009a", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Consensus Seeking via Large Language Models", "abstract": "Multi-agent systems driven by large language models (LLMs) have shown promising abilities for solving complex tasks in a collaborative manner. This work considers a fundamental problem in multi-agent collaboration: consensus seeking. When multiple agents work together, we are interested in how they can reach a consensus through inter-agent negotiation. To that end, this work studies a consensus-seeking task where the state of each agent is a numerical value and they negotiate with each other to reach a consensus value. It is revealed that when not explicitly directed on which strategy should be adopted, the LLM-driven agents primarily use the average strategy for consensus seeking although they may occasionally use some other strategies. Moreover, this work analyzes the impact of the agent number, agent personality, and network topology on the negotiation process. The findings reported in this work can potentially lay the foundations for understanding the behaviors of LLM-driven multi-agent systems for solving more complex tasks. Furthermore, LLM-driven consensus seeking is applied to a multi-robot aggregation task. This application demonstrates the potential of LLM-driven agents to achieve zero-shot autonomous planning for multi-robot collaboration tasks. Project website: windylab.github.io/ConsensusLLM/.", "authors": ["Huaben Chen", "Wenkang Ji", "Lufeng Xu", "Shiyu Zhao"], "categories": ["cs.CL", "cs.RO", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-10-31", "url": "https://arxiv.org/abs/2310.20151", "pdf_url": "https://arxiv.org/pdf/2310.20151v2", "arxiv_id": "2310.20151", "doi": "10.48550/arXiv.2310.20151", "citation_count": 62, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4498}
{"id": "f6ec7c73aa21c8149fe116dc5bb6c6d8689a475c60e69c170f8f12153b20cad0", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Based Agent Society Investigation: Collaboration and Confrontation in Avalon Gameplay", "abstract": "This paper explores the open research problem of understanding the social behaviors of LLM-based agents. Using Avalon as a testbed, we employ system prompts to guide LLM agents in gameplay. While previous studies have touched on gameplay with LLM agents, research on their social behaviors is lacking. We propose a novel framework, tailored for Avalon, features a multi-agent system facilitating efficient communication and interaction. We evaluate its performance based on game success and analyze LLM agents' social behaviors. Results affirm the framework's effectiveness in creating adaptive agents and suggest LLM-based agents' potential in navigating dynamic social interactions. By examining collaboration and confrontation behaviors, we offer insights into this field's research and applications. Our code is publicly available at https://github.com/3DAgentWorld/LLM-Game-Agent.", "authors": ["Yihuai Lan", "Zhiqiang Hu", "Lei Wang", "Yang Wang", "Deheng Ye", "Peilin Zhao", "Ee-Peng Lim", "Hui Xiong", "Hao Wang"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-23", "url": "https://arxiv.org/abs/2310.14985", "pdf_url": "https://arxiv.org/pdf/2310.14985v4", "arxiv_id": "2310.14985", "doi": "10.48550/arXiv.2310.14985", "citation_count": 74, "influential_citation_count": 7, "has_code": true, "code_url": "https://github.com/3DAgentWorld/LLM-Game-Agent", "venue": "Conference on Empirical Methods in Natural Language Processing", "quality_score": 0.4688}
{"id": "2b06db31bd59d4ea70ebcfe909850e0b4e4746d4343aa93de8b91e757b9839f2", "sources": ["arxiv", "semantic_scholar"], "title": "Influence of Team Interactions on Multi-Robot Cooperation: A Relational Network Perspective", "abstract": "Relational networks within a team play a critical role in the performance of many real-world multi-robot systems. To successfully accomplish tasks that require cooperation and coordination, different agents (e.g., robots) necessitate different priorities based on their positioning within the team. Yet, many of the existing multi-robot cooperation algorithms regard agents as interchangeable and lack a mechanism to guide the type of cooperation strategy the agents should exhibit. To account for the team structure in cooperative tasks, we propose a novel algorithm that uses a relational network comprising inter-agent relationships to prioritize certain agents over others. Through appropriate design of the team's relational network, we can guide the cooperation strategy, resulting in the emergence of new behaviors that accomplish the specified task. We conducted six experiments in a multi-robot setting with a cooperative task. Our results demonstrate that the proposed method can effectively influence the type of solution that the algorithm converges to by specifying the relationships between the agents, making it a promising approach for tasks that require cooperation among agents with a specified team structure.", "authors": ["Yasin Findik", "Hamid Osooli", "Paul Robinette", "Kshitij Jerath", "S. Reza Ahmadzadeh"], "categories": ["cs.RO", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-19", "url": "https://arxiv.org/abs/2310.12910", "pdf_url": "https://arxiv.org/pdf/2310.12910v1", "arxiv_id": "2310.12910", "doi": "10.1109/MRS60187.2023.10416779", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Symposium on Multi-Robot and Multi-Agent Systems", "quality_score": 0.2113}
{"id": "91c6421b0fe692b2975087eb5476e0dc0d63a2c57b5bb79a427ffa63a9fc94f8", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Adaptation: Learning to Recover from Unforeseen Malfunctions in Multi-Robot Teams", "abstract": "Cooperative multi-agent reinforcement learning (MARL) approaches tackle the challenge of finding effective multi-agent cooperation strategies for accomplishing individual or shared objectives in multi-agent teams. In real-world scenarios, however, agents may encounter unforeseen failures due to constraints like battery depletion or mechanical issues. Existing state-of-the-art methods in MARL often recover slowly -- if at all -- from such malfunctions once agents have already converged on a cooperation strategy. To address this gap, we present the Collaborative Adaptation (CA) framework. CA introduces a mechanism that guides collaboration and accelerates adaptation from unforeseen failures by leveraging inter-agent relationships. Our findings demonstrate that CA enables agents to act on the knowledge of inter-agent relations, recovering from unforeseen agent failures and selecting appropriate cooperative strategies.", "authors": ["Yasin Findik", "Paul Robinette", "Kshitij Jerath", "S. Reza Ahmadzadeh"], "categories": ["cs.RO", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-19", "url": "https://arxiv.org/abs/2310.12909", "pdf_url": "https://arxiv.org/pdf/2310.12909v1", "arxiv_id": "2310.12909", "doi": "10.48550/arXiv.2310.12909", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "a134ebbe69580bacf41ccaf7c8a05e0189992346b2b2bcc38343f959fb95131a", "sources": ["arxiv", "semantic_scholar"], "title": "Fact-based Agent modeling for Multi-Agent Reinforcement Learning", "abstract": "In multi-agent systems, agents need to interact and collaborate with other agents in environments. Agent modeling is crucial to facilitate agent interactions and make adaptive cooperation strategies. However, it is challenging for agents to model the beliefs, behaviors, and intentions of other agents in non-stationary environment where all agent policies are learned simultaneously. In addition, the existing methods realize agent modeling through behavior cloning which assume that the local information of other agents can be accessed during execution or training. However, this assumption is infeasible in unknown scenarios characterized by unknown agents, such as competition teams, unreliable communication and federated learning due to privacy concerns. To eliminate this assumption and achieve agent modeling in unknown scenarios, Fact-based Agent modeling (FAM) method is proposed in which fact-based belief inference (FBI) network models other agents in partially observable environment only based on its local information. The reward and observation obtained by agents after taking actions are called facts, and FAM uses facts as reconstruction target to learn the policy representation of other agents through a variational autoencoder. We evaluate FAM on various Multiagent Particle Environment (MPE) and compare the results with several state-of-the-art MARL algorithms. Experimental results show that compared with baseline methods, FAM can effectively improve the efficiency of agent policy learning by making adaptive cooperation strategies in multi-agent reinforcement learning tasks, while achieving higher returns in complex competitive-cooperative mixed scenarios.", "authors": ["Baofu Fang", "Caiming Zheng", "Hao Wang"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-18", "url": "https://arxiv.org/abs/2310.12290", "pdf_url": "https://arxiv.org/pdf/2310.12290v1", "arxiv_id": "2310.12290", "doi": "10.48550/arXiv.2310.12290", "citation_count": 1, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "737ed4ba7b82522ce45ce407c97bb449b2d158ed82b81031f7eae6e9a6bcf7b8", "sources": ["arxiv", "semantic_scholar"], "title": "Malicious Agent Detection for Robust Multi-Agent Collaborative Perception", "abstract": "Recently, multi-agent collaborative (MAC) perception has been proposed and outperformed the traditional single-agent perception in many applications, such as autonomous driving. However, MAC perception is more vulnerable to adversarial attacks than single-agent perception due to the information exchange. The attacker can easily degrade the performance of a victim agent by sending harmful information from a malicious agent nearby. In this paper, we extend adversarial attacks to an important perception task -- MAC object detection, where generic defenses such as adversarial training are no longer effective against these attacks. More importantly, we propose Malicious Agent Detection (MADE), a reactive defense specific to MAC perception that can be deployed by each agent to accurately detect and then remove any potential malicious agent in its local collaboration network. In particular, MADE inspects each agent in the network independently using a semi-supervised anomaly detector based on a double-hypothesis test with the Benjamini-Hochberg procedure to control the false positive rate of the inference. For the two hypothesis tests, we propose a match loss statistic and a collaborative reconstruction loss statistic, respectively, both based on the consistency between the agent to be inspected and the ego agent where our detector is deployed. We conduct comprehensive evaluations on a benchmark 3D dataset V2X-sim and a real-road dataset DAIR-V2X and show that with the protection of MADE, the drops in the average precision compared with the best-case \"oracle\" defender against our attack are merely 1.28% and 0.34%, respectively, much lower than 8.92% and 10.00% for adversarial training, respectively.", "authors": ["Yangheng Zhao", "Zhen Xiang", "Sheng Yin", "Xianghe Pang", "Siheng Chen", "Yanfeng Wang"], "categories": ["cs.CR"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-18", "url": "https://arxiv.org/abs/2310.11901", "pdf_url": "https://arxiv.org/pdf/2310.11901v2", "arxiv_id": "2310.11901", "doi": "10.48550/arXiv.2310.11901", "citation_count": 12, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2785}
{"id": "7528c420785508d71f1e1053f27481c98ec48a3bd398a5ad46354a31a5a112be", "sources": ["arxiv", "semantic_scholar"], "title": "Balancing Autonomy and Alignment: A Multi-Dimensional Taxonomy for Autonomous LLM-powered Multi-Agent Architectures", "abstract": "Large language models (LLMs) have revolutionized the field of artificial intelligence, endowing it with sophisticated language understanding and generation capabilities. However, when faced with more complex and interconnected tasks that demand a profound and iterative thought process, LLMs reveal their inherent limitations. Autonomous LLM-powered multi-agent systems represent a strategic response to these challenges. Such systems strive for autonomously tackling user-prompted goals by decomposing them into manageable tasks and orchestrating their execution and result synthesis through a collective of specialized intelligent agents. Equipped with LLM-powered reasoning capabilities, these agents harness the cognitive synergy of collaborating with their peers, enhanced by leveraging contextual resources such as tools and datasets. While these architectures hold promising potential in amplifying AI capabilities, striking the right balance between different levels of autonomy and alignment remains the crucial challenge for their effective operation. This paper proposes a comprehensive multi-dimensional taxonomy, engineered to analyze how autonomous LLM-powered multi-agent systems balance the dynamic interplay between autonomy and alignment across various aspects inherent to architectural viewpoints such as goal-driven task management, agent composition, multi-agent collaboration, and context interaction. It also includes a domain-ontology model specifying fundamental architectural concepts. Our taxonomy aims to empower researchers, engineers, and AI practitioners to systematically analyze the architectural dynamics and balancing strategies employed by these increasingly prevalent AI systems. The exploratory taxonomic classification of selected representative LLM-powered multi-agent systems illustrates its practical utility and reveals potential for future research and development.", "authors": ["Thorsten Händler"], "categories": ["cs.AI", "cs.MA", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-05", "url": "https://arxiv.org/abs/2310.03659", "pdf_url": "https://arxiv.org/pdf/2310.03659v1", "arxiv_id": "2310.03659", "doi": "10.48550/arXiv.2310.03659", "citation_count": 42, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4084}
{"id": "2d19bac643a69e8954aaeed51ea68bfb3cb29090a22eb1956914293c89f13929", "sources": ["arxiv", "semantic_scholar"], "title": "LLM-Coordination: Evaluating and Analyzing Multi-agent Coordination Abilities in Large Language Models", "abstract": "Large Language Models (LLMs) have demonstrated emergent common-sense reasoning and Theory of Mind (ToM) capabilities, making them promising candidates for developing coordination agents. This study introduces the LLM-Coordination Benchmark, a novel benchmark for analyzing LLMs in the context of Pure Coordination Settings, where agents must cooperate to maximize gains. Our benchmark evaluates LLMs through two distinct tasks. The first is Agentic Coordination, where LLMs act as proactive participants in four pure coordination games. The second is Coordination Question Answering (CoordQA), which tests LLMs on 198 multiple-choice questions across these games to evaluate three key abilities: Environment Comprehension, ToM Reasoning, and Joint Planning. Results from Agentic Coordination experiments reveal that LLM-Agents excel in multi-agent coordination settings where decision-making primarily relies on environmental variables but face challenges in scenarios requiring active consideration of partners' beliefs and intentions. The CoordQA experiments further highlight significant room for improvement in LLMs' Theory of Mind reasoning and joint planning capabilities. Zero-Shot Coordination (ZSC) experiments in the Agentic Coordination setting demonstrate that LLM agents, unlike RL methods, exhibit robustness to unseen partners. These findings indicate the potential of LLMs as Agents in pure coordination setups and underscore areas for improvement. Code Available at https://github.com/eric-ai-lab/llm_coordination.", "authors": ["Saaket Agashe", "Yue Fan", "Anthony Reyna", "Xin Eric Wang"], "categories": ["cs.CL", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-05", "url": "https://arxiv.org/abs/2310.03903", "pdf_url": "https://arxiv.org/pdf/2310.03903v3", "arxiv_id": "2310.03903", "doi": "10.18653/v1/2025.findings-naacl.448", "citation_count": 63, "influential_citation_count": 4, "has_code": true, "code_url": "https://github.com/eric-ai-lab/llm_coordination", "venue": "North American Chapter of the Association for Computational Linguistics", "quality_score": 0.4515}
{"id": "3deb185969dd9bac66c4e5b99fbc8fe74fe32b4dfba58686ebd7b7d21f9b85ab", "sources": ["arxiv", "semantic_scholar"], "title": "A Game Approach to Multi-dimensional Opinion Dynamics in Social Networks with Stubborn Strategist Agents", "abstract": "In a social network, individuals express their opinions on several interdependent topics, and therefore the evolution of their opinions on these topics is also mutually dependent. In this work, we propose a differential game model for the multi-dimensional opinion formation of a social network whose population of agents interacts according to a communication graph. Each individual's opinion evolves according to an aggregation of disagreements between the agent's opinions and its graph neighbors on multiple interdependent topics exposed to an unknown extraneous disturbance. For a social network with strategist agents the opinions evolve over time with respect to the minimization of a quadratic cost function that solely represents each individual's motives against the disturbance. We find the unique Nash/worst-case equilibrium solution for the proposed differential game model of coupled multi-dimensional opinions under an open-loop information structure. Moreover, we propose a distributed implementation of the Nash/worst-case equilibrium solution. We examine the non-distributed and proposed distributed open-loop Nash/worst-case strategies on a small social network with strategist agents in a two-dimensional opinion space. Then we compare the opinions evolved based on the Nash/worst-case strategy with the opinions corresponding to social optimality actions for non-strategist agents.", "authors": ["Hossein B. Jond", "Aykut Yıldız"], "categories": ["cs.SI"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-05", "url": "https://arxiv.org/abs/2310.03900", "pdf_url": "https://arxiv.org/pdf/2310.03900v3", "arxiv_id": "2310.03900", "doi": "10.1016/j.ejcon.2023.100941", "citation_count": 5, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "European Journal of Control", "quality_score": 0.1945}
{"id": "47fc465c890f03bda0122731a648d5767bcd3e313b43283659e9688d65c24f8b", "sources": ["arxiv", "semantic_scholar"], "title": "Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View", "abstract": "As Natural Language Processing (NLP) systems are increasingly employed in intricate social environments, a pressing query emerges: Can these NLP systems mirror human-esque collaborative intelligence, in a multi-agent society consisting of multiple large language models (LLMs)? This paper probes the collaboration mechanisms among contemporary NLP systems by melding practical experiments with theoretical insights. We fabricate four unique `societies' comprised of LLM agents, where each agent is characterized by a specific `trait' (easy-going or overconfident) and engages in collaboration with a distinct `thinking pattern' (debate or reflection). Through evaluating these multi-agent societies on three benchmark datasets, we discern that certain collaborative strategies not only outshine previous top-tier approaches, but also optimize efficiency (using fewer API tokens). Moreover, our results further illustrate that LLM agents manifest human-like social behaviors, such as conformity and consensus reaching, mirroring foundational social psychology theories. In conclusion, we integrate insights from social psychology to contextualize the collaboration of LLM agents, inspiring further investigations into the collaboration mechanism for LLMs. We commit to sharing our code and datasets\\footnote{\\url{https://github.com/zjunlp/MachineSoM}.}, hoping to catalyze further research in this promising avenue.", "authors": ["Jintian Zhang", "Xin Xu", "Ningyu Zhang", "Ruibo Liu", "Bryan Hooi", "Shumin Deng"], "categories": ["cs.CL", "cs.AI", "cs.CY", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-03", "url": "https://arxiv.org/abs/2310.02124", "pdf_url": "https://arxiv.org/pdf/2310.02124v3", "arxiv_id": "2310.02124", "doi": "10.48550/arXiv.2310.02124", "citation_count": 260, "influential_citation_count": 9, "has_code": true, "code_url": "https://github.com/zjunlp/MachineSoM}.}", "venue": "arXiv.org", "quality_score": 0.6042}
{"id": "19e869fae27c0893cb8ad401984b57c38b7a53d2de40b72e5148580e44fdf19c", "sources": ["arxiv", "semantic_scholar"], "title": "Adapting LLM Agents with Universal Feedback in Communication", "abstract": "Recent advances in large language models (LLMs) have demonstrated potential for LLM agents. To facilitate the training for these agents with both linguistic feedback and non-linguistic reward signals, we introduce Learning through Communication (LTC). We design a universal buffer to store all the feedback, and an iterative pipeline to enable an LLM agent to explore and update its policy in an given environment. To optimize agent interactions for task-specific learning with our universal buffer and pipeline, we introduce diverse communication patterns tailored for both single-agent and multi-agent environments. We evaluate the efficacy of our LTC approach on four diverse datasets: ALFWorld (single-agent), HotpotQA (multi-agent collaboration), Chameleon (multi-agent competition), and GSM8k (multi-agent teacher-student). On these data sets, LTC outperforms the supervised instruction fine-tuning baselines by 3.6% to 12%. These results highlight the versatility and efficiency of LTC in facilitating online adaptation for LLM agents.", "authors": ["Kuan Wang", "Yadong Lu", "Michael Santacroce", "Yeyun Gong", "Chao Zhang", "Yelong Shen"], "categories": ["cs.CL", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-10-01", "url": "https://arxiv.org/abs/2310.01444", "pdf_url": "https://arxiv.org/pdf/2310.01444v3", "arxiv_id": "2310.01444", "doi": null, "citation_count": 16, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.3076}
{"id": "daf7b9e8f120a5a191621f984a786cafd39159fb2e6df63130f7fe53959c8215", "sources": ["arxiv", "semantic_scholar"], "title": "Cooperation Dynamics in Multi-Agent Systems: Exploring Game-Theoretic Scenarios with Mean-Field Equilibria", "abstract": "Cooperation is fundamental in Multi-Agent Systems (MAS) and Multi-Agent Reinforcement Learning (MARL), often requiring agents to balance individual gains with collective rewards. In this regard, this paper aims to investigate strategies to invoke cooperation in game-theoretic scenarios, namely the Iterated Prisoner's Dilemma, where agents must optimize both individual and group outcomes. Existing cooperative strategies are analyzed for their effectiveness in promoting group-oriented behavior in repeated games. Modifications are proposed where encouraging group rewards will also result in a higher individual gain, addressing real-world dilemmas seen in distributed systems. The study extends to scenarios with exponentially growing agent populations ($N \\longrightarrow +\\infty$), where traditional computation and equilibrium determination are challenging. Leveraging mean-field game theory, equilibrium solutions and reward structures are established for infinitely large agent sets in repeated games. Finally, practical insights are offered through simulations using the Multi Agent-Posthumous Credit Assignment trainer, and the paper explores adapting simulation algorithms to create scenarios favoring cooperation for group rewards. These practical implementations bridge theoretical concepts with real-world applications.", "authors": ["Vaigarai Sathi", "Sabahat Shaik", "Jaswanth Nidamanuri"], "categories": ["cs.GT", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-09-28", "url": "https://arxiv.org/abs/2309.16263", "pdf_url": "https://arxiv.org/pdf/2309.16263v3", "arxiv_id": "2309.16263", "doi": "10.48550/arXiv.2309.16263", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1505}
{"id": "208d277e2d9ac4b4c9b532037f3725609f00fdf18ee86607a8dfad730bc2d985", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Deep Reinforcement Learning for Cooperative and Competitive Autonomous Vehicles using AutoDRIVE Ecosystem", "abstract": "This work presents a modular and parallelizable multi-agent deep reinforcement learning framework for imbibing cooperative as well as competitive behaviors within autonomous vehicles. We introduce AutoDRIVE Ecosystem as an enabler to develop physically accurate and graphically realistic digital twins of Nigel and F1TENTH, two scaled autonomous vehicle platforms with unique qualities and capabilities, and leverage this ecosystem to train and deploy multi-agent reinforcement learning policies. We first investigate an intersection traversal problem using a set of cooperative vehicles (Nigel) that share limited state information with each other in single as well as multi-agent learning settings using a common policy approach. We then investigate an adversarial head-to-head autonomous racing problem using a different set of vehicles (F1TENTH) in a multi-agent learning setting using an individual policy approach. In either set of experiments, a decentralized learning architecture was adopted, which allowed robust training and testing of the approaches in stochastic environments, since the agents were mutually independent and exhibited asynchronous motion behavior. The problems were further aggravated by providing the agents with sparse observation spaces and requiring them to sample control commands that implicitly satisfied the imposed kinodynamic as well as safety constraints. The experimental results for both problem statements are reported in terms of quantitative metrics and qualitative remarks for training as well as deployment phases.", "authors": ["Tanmay Vilas Samak", "Chinmay Vilas Samak", "Venkat Krovi"], "categories": ["cs.RO", "cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-09-18", "url": "https://arxiv.org/abs/2309.10007", "pdf_url": "https://arxiv.org/pdf/2309.10007v2", "arxiv_id": "2309.10007", "doi": "10.48550/arXiv.2309.10007", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "48f598240e609ea1ed01d1665f66bd1ad3bd53777861e01e6c3730d4616239b6", "sources": ["arxiv", "semantic_scholar"], "title": "SMART-LLM: Smart Multi-Agent Robot Task Planning using Large Language Models", "abstract": "In this work, we introduce SMART-LLM, an innovative framework designed for embodied multi-robot task planning. SMART-LLM: Smart Multi-Agent Robot Task Planning using Large Language Models (LLMs), harnesses the power of LLMs to convert high-level task instructions provided as input into a multi-robot task plan. It accomplishes this by executing a series of stages, including task decomposition, coalition formation, and task allocation, all guided by programmatic LLM prompts within the few-shot prompting paradigm. We create a benchmark dataset designed for validating the multi-robot task planning problem, encompassing four distinct categories of high-level instructions that vary in task complexity. Our evaluation experiments span both simulation and real-world scenarios, demonstrating that the proposed model can achieve promising results for generating multi-robot task plans. The experimental videos, code, and datasets from the work can be found at https://sites.google.com/view/smart-llm/.", "authors": ["Shyam Sundar Kannan", "Vishnunandan L. N. Venkatesh", "Byung-Cheol Min"], "categories": ["cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-09-18", "url": "https://arxiv.org/abs/2309.10062", "pdf_url": "https://arxiv.org/pdf/2309.10062v2", "arxiv_id": "2309.10062", "doi": "10.1109/IROS58592.2024.10802322", "citation_count": 272, "influential_citation_count": 23, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.6901}
{"id": "0aee26de3da068e08690f27651b0c6d18177e1c7c0c75650edf4998fe3c51512", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Collective Construction using 3D Decomposition", "abstract": "This paper addresses a Multi-Agent Collective Construction (MACC) problem that aims to build a three-dimensional structure comprised of cubic blocks. We use cube-shaped robots that can carry one cubic block at a time, and move forward, reverse, left, and right to an adjacent cell of the same height or climb up and down one cube height. To construct structures taller than one cube, the robots must build supporting stairs made of blocks and remove the stairs once the structure is built. Conventional techniques solve for the entire structure at once and quickly become intractable for larger workspaces and complex structures, especially in a multi-agent setting. To this end, we present a decomposition algorithm that computes valid substructures based on intrinsic structural dependencies. We use Mixed Integer Linear Programming (MILP) to solve for each of these substructures and then aggregate the solutions to construct the entire structure. Extensive testing on 200 randomly generated structures shows an order of magnitude improvement in the solution computation time compared to an MILP approach without decomposition. Additionally, compared to Reinforcement Learning (RL) based and heuristics-based approaches drawn from the literature, our solution indicates orders of magnitude improvement in the number of pick-up and drop-off actions required to construct a structure. Furthermore, we leverage the independence between substructures to detect which sub-structures can be built in parallel. With this parallelization technique, we illustrate a further improvement in the number of time steps required to complete building the structure. This work is a step towards applying multi-agent collective construction for real-world structures by significantly reducing solution computation time with a bounded increase in the number of time steps required to build the structure.", "authors": ["Akshaya Kesarimangalam Srinivasan", "Shambhavi Singh", "Geordan Gutow", "Howie Choset", "Bhaskar Vundurthy"], "categories": ["cs.RO", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-09-02", "url": "https://arxiv.org/abs/2309.00985", "pdf_url": "https://arxiv.org/pdf/2309.00985v1", "arxiv_id": "2309.00985", "doi": "10.1109/IROS55552.2023.10341964", "citation_count": 7, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.2258}
{"id": "fb6526f44ca0729d1be6a17ede4c6b751afa59abd8aa5a7bd58a5f08f490e60e", "sources": ["arxiv", "semantic_scholar"], "title": "ZeroLeak: Using LLMs for Scalable and Cost Effective Side-Channel Patching", "abstract": "Security critical software, e.g., OpenSSL, comes with numerous side-channel leakages left unpatched due to a lack of resources or experts. The situation will only worsen as the pace of code development accelerates, with developers relying on Large Language Models (LLMs) to automatically generate code. In this work, we explore the use of LLMs in generating patches for vulnerable code with microarchitectural side-channel leakages. For this, we investigate the generative abilities of powerful LLMs by carefully crafting prompts following a zero-shot learning approach. All generated code is dynamically analyzed by leakage detection tools, which are capable of pinpointing information leakage at the instruction level leaked either from secret dependent accesses or branches or vulnerable Spectre gadgets, respectively. Carefully crafted prompts are used to generate candidate replacements for vulnerable code, which are then analyzed for correctness and for leakage resilience. From a cost/performance perspective, the GPT4-based configuration costs in API calls a mere few cents per vulnerability fixed. Our results show that LLM-based patching is far more cost-effective and thus provides a scalable solution. Finally, the framework we propose will improve in time, especially as vulnerability detection tools and LLMs mature.", "authors": ["M. Caner Tol", "Berk Sunar"], "categories": ["cs.CR", "cs.LG", "cs.SE"], "fields_of_study": ["Computer Science"], "published_date": "2023-08-24", "url": "https://arxiv.org/abs/2308.13062", "pdf_url": "https://arxiv.org/pdf/2308.13062v1", "arxiv_id": "2308.13062", "doi": "10.48550/arXiv.2308.13062", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "8c40d577895ec8a8336fd6980206761fd6f1644a8ec846c1524d38c0b8464659", "sources": ["arxiv", "semantic_scholar"], "title": "Scalable δ-Level Coherent State Synchronization of Multi-Agent Systems in the Presence of Bounded Disturbances", "abstract": "In this paper, we study scalable $δ-$level coherent state synchronization for multi-agent systems (MAS) where the agents are subject to bounded disturbances/noises. We propose a scale-free framework designed solely based on the knowledge of agent models and agnostic to the communication graph and the size of the network. We define the level of coherency for each agent as the norm of the weighted sum of the disagreement dynamics with its neighbors. The objective is to restrict the network's coherency level to $δ$ without a-priori information about the disturbance.", "authors": ["Donya Nojavanzadeh", "Zhenwei Liu", "Ali Saberi", "Anton A. Stoorvogel"], "categories": ["eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-08-23", "url": "https://arxiv.org/abs/2308.11959", "pdf_url": "https://arxiv.org/pdf/2308.11959v5", "arxiv_id": "2308.11959", "doi": "10.48550/arXiv.2308.11959", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "86d39a648006ba7ab7973d641da17765ed0d1271126204034af9634c1d86c921", "sources": ["arxiv", "semantic_scholar"], "title": "AgentVerse: Facilitating Multi-Agent Collaboration and Exploring Emergent Behaviors", "abstract": "Autonomous agents empowered by Large Language Models (LLMs) have undergone significant improvements, enabling them to generalize across a broad spectrum of tasks. However, in real-world scenarios, cooperation among individuals is often required to enhance the efficiency and effectiveness of task accomplishment. Hence, inspired by human group dynamics, we propose a multi-agent framework \\framework that can collaboratively and dynamically adjust its composition as a greater-than-the-sum-of-its-parts system. Our experiments demonstrate that \\framework framework can effectively deploy multi-agent groups that outperform a single agent. Furthermore, we delve into the emergence of social behaviors among individual agents within a group during collaborative task accomplishment. In view of these behaviors, we discuss some possible strategies to leverage positive ones and mitigate negative ones for improving the collaborative potential of multi-agent groups. Our codes for \\framework will soon be released at \\url{https://github.com/OpenBMB/AgentVerse}.", "authors": ["Weize Chen", "Yusheng Su", "Jingwei Zuo", "Cheng Yang", "Chenfei Yuan", "Chi-Min Chan", "Heyang Yu", "Yaxi Lu", "Yi-Hsin Hung", "Chen Qian", "Yujia Qin", "Xin Cong", "Ruobing Xie", "Zhiyuan Liu", "Maosong Sun", "Jie Zhou"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-08-21", "url": "https://arxiv.org/abs/2308.10848", "pdf_url": "https://arxiv.org/pdf/2308.10848v3", "arxiv_id": "2308.10848", "doi": "10.48550/arXiv.2308.10848", "citation_count": 643, "influential_citation_count": 52, "has_code": true, "code_url": "https://github.com/OpenBMB/AgentVerse/", "venue": "International Conference on Learning Representations", "quality_score": 0.8621}
{"id": "bc55dd7a1498259793b99884c947fabc4c21f836bdaa18c812ca9cbd6610e2e5", "sources": ["arxiv", "semantic_scholar"], "title": "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation", "abstract": "AutoGen is an open-source framework that allows developers to build LLM applications via multiple agents that can converse with each other to accomplish tasks. AutoGen agents are customizable, conversable, and can operate in various modes that employ combinations of LLMs, human inputs, and tools. Using AutoGen, developers can also flexibly define agent interaction behaviors. Both natural language and computer code can be used to program flexible conversation patterns for different applications. AutoGen serves as a generic infrastructure to build diverse applications of various complexities and LLM capacities. Empirical studies demonstrate the effectiveness of the framework in many example applications, with domains ranging from mathematics, coding, question answering, operations research, online decision-making, entertainment, etc.", "authors": ["Qingyun Wu", "Gagan Bansal", "Jieyu Zhang", "Yiran Wu", "Beibin Li", "Erkang Zhu", "Li Jiang", "Xiaoyun Zhang", "Shaokun Zhang", "Jiale Liu", "Ahmed Hassan Awadallah", "Ryen W White", "Doug Burger", "Chi Wang"], "categories": ["cs.AI", "cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-08-16", "url": "https://arxiv.org/abs/2308.08155", "pdf_url": "https://arxiv.org/pdf/2308.08155v2", "arxiv_id": "2308.08155", "doi": null, "citation_count": 1801, "influential_citation_count": 99, "has_code": true, "code_url": null, "venue": null, "quality_score": 1.0}
{"id": "087b4ec2205c60f985c82fdcfc6a148160774fac20200dfbed58d8ce367a86fa", "sources": ["arxiv", "semantic_scholar"], "title": "ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate", "abstract": "Text evaluation has historically posed significant challenges, often demanding substantial labor and time cost. With the emergence of large language models (LLMs), researchers have explored LLMs' potential as alternatives for human evaluation. While these single-agent-based approaches show promise, experimental results suggest that further advancements are needed to bridge the gap between their current effectiveness and human-level evaluation quality. Recognizing that best practices of human evaluation processes often involve multiple human annotators collaborating in the evaluation, we resort to a multi-agent debate framework, moving beyond single-agent prompting strategies. The multi-agent-based approach enables a group of LLMs to synergize with an array of intelligent counterparts, harnessing their distinct capabilities and expertise to enhance efficiency and effectiveness in handling intricate tasks. In this paper, we construct a multi-agent referee team called ChatEval to autonomously discuss and evaluate the quality of generated responses from different models on open-ended questions and traditional natural language generation (NLG) tasks. Our analysis shows that ChatEval transcends mere textual scoring, offering a human-mimicking evaluation process for reliable assessments. Our code is available at https://github.com/chanchimin/ChatEval.", "authors": ["Chi-Min Chan", "Weize Chen", "Yusheng Su", "Jianxuan Yu", "Wei Xue", "Shanghang Zhang", "Jie Fu", "Zhiyuan Liu"], "categories": ["cs.CL"], "fields_of_study": ["Computer Science"], "published_date": "2023-08-14", "url": "https://arxiv.org/abs/2308.07201", "pdf_url": "https://arxiv.org/pdf/2308.07201v1", "arxiv_id": "2308.07201", "doi": "10.48550/arXiv.2308.07201", "citation_count": 939, "influential_citation_count": 80, "has_code": true, "code_url": "https://github.com/chanchimin/ChatEval", "venue": "arXiv.org", "quality_score": 0.9542}
{"id": "c2db5b571e8f7b6e6eaa9f179e31ac7ed9b65c24563b995de6e1ce0c26abf80a", "sources": ["arxiv", "semantic_scholar"], "title": "MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework", "abstract": "Remarkable progress has been made on automated problem solving through societies of agents based on large language models (LLMs). Existing LLM-based multi-agent systems can already solve simple dialogue tasks. Solutions to more complex tasks, however, are complicated through logic inconsistencies due to cascading hallucinations caused by naively chaining LLMs. Here we introduce MetaGPT, an innovative meta-programming framework incorporating efficient human workflows into LLM-based multi-agent collaborations. MetaGPT encodes Standardized Operating Procedures (SOPs) into prompt sequences for more streamlined workflows, thus allowing agents with human-like domain expertise to verify intermediate results and reduce errors. MetaGPT utilizes an assembly line paradigm to assign diverse roles to various agents, efficiently breaking down complex tasks into subtasks involving many agents working together. On collaborative software engineering benchmarks, MetaGPT generates more coherent solutions than previous chat-based multi-agent systems. Our project can be found at https://github.com/geekan/MetaGPT", "authors": ["Sirui Hong", "Mingchen Zhuge", "Jiaqi Chen", "Xiawu Zheng", "Yuheng Cheng", "Ceyao Zhang", "Jinlin Wang", "Zili Wang", "Steven Ka Shing Yau", "Zijuan Lin", "Liyang Zhou", "Chenyu Ran", "Lingfeng Xiao", "Chenglin Wu", "Jürgen Schmidhuber"], "categories": ["cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-08-01", "url": "https://arxiv.org/abs/2308.00352", "pdf_url": "https://arxiv.org/pdf/2308.00352v7", "arxiv_id": "2308.00352", "doi": "10.48550/arXiv.2308.00352", "citation_count": 1925, "influential_citation_count": 150, "has_code": true, "code_url": "https://github.com/geekan/MetaGPT", "venue": "arXiv.org", "quality_score": 1.0}
{"id": "1c1670a2704574ebb8d2a1d304d4cb5cecef53dffbd869ec7a8dcdbaf601f24c", "sources": ["arxiv", "semantic_scholar"], "title": "Using Multi-Agent MicroServices (MAMS) for Agent Based Modelling", "abstract": "This paper demonstrates the use of the Multi-Agent MicroServices (MAMS) architectural style through a case study based around the development of a prototype traffic simulation in which agents model a population of individuals who travel from home to work and vice versa by car.", "authors": ["Martynas Jagutis", "Sean Russell", "Rem Collier"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-07-27", "url": "https://arxiv.org/abs/2307.14745", "pdf_url": "https://arxiv.org/pdf/2307.14745v1", "arxiv_id": "2307.14745", "doi": "10.48550/arXiv.2307.14745", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Workshop on Engineering Multi-Agent Systems", "quality_score": 0.0753}
{"id": "7d69882f836db659386e351d657380a4e3d025bb088602ed826a8886c5610932", "sources": ["arxiv", "semantic_scholar"], "title": "Heterogeneous Embodied Multi-Agent Collaboration", "abstract": "Multi-agent embodied tasks have recently been studied in complex indoor visual environments. Collaboration among multiple agents can improve work efficiency and has significant practical value. However, most of the existing research focuses on homogeneous multi-agent tasks. Compared with homogeneous agents, heterogeneous agents can leverage their different capabilities to allocate corresponding sub-tasks and cooperate to complete complex tasks. Heterogeneous multi-agent tasks are common in real-world scenarios, and the collaboration strategy among heterogeneous agents is a challenging and important problem to be solved. To study collaboration among heterogeneous agents, we propose the heterogeneous multi-agent tidying-up task, in which multiple heterogeneous agents with different capabilities collaborate with each other to detect misplaced objects and place them in reasonable locations. This is a demanding task since it requires agents to make the best use of their different capabilities to conduct reasonable task planning and complete the whole task. To solve this task, we build a heterogeneous multi-agent tidying-up benchmark dataset in a large number of houses with multiple rooms based on ProcTHOR-10K. We propose the hierarchical decision model based on misplaced object detection, reasonable receptacle prediction, as well as the handshake-based group communication mechanism. Extensive experiments are conducted to demonstrate the effectiveness of the proposed model. The project's website and videos of experiments can be found at https://hetercol.github.io/.", "authors": ["Xinzhu Liu", "Di Guo", "Huaping Liu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2023-07-26", "url": "https://arxiv.org/abs/2307.13957", "pdf_url": "https://arxiv.org/pdf/2307.13957v2", "arxiv_id": "2307.13957", "doi": "10.1109/LRA.2024.3390588", "citation_count": 18, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Robotics and Automation Letters", "quality_score": 0.3197}
{"id": "17d946f4bd28973520c5d87f98013efdb31b4e03a56f4ae3694f82d07ffe513b", "sources": ["arxiv", "semantic_scholar"], "title": "Deep and Decentralized Multi-Agent Coverage of a Target with Unknown Distribution", "abstract": "This paper proposes a new architecture for multi-agent systems to cover an unknowingly distributed fast, safely, and decentralizedly. The inter-agent communication is organized by a directed graph with fixed topology, and we model agent coordination as a decentralized leader-follower problem with time-varying communication weights. Given this problem setting, we first present a method for converting communication graph into a neural network, where an agent can be represented by a unique node of the communication graph but multiple neurons of the corresponding neural network. We then apply a mass-cetric strategy to train time-varying communication weights of the neural network in a decentralized fashion which in turn implies that the observation zone of every follower agent is independently assigned by the follower based on positions of in-neighbors. By training the neural network, we can ensure safe and decentralized multi-agent coordination of coverage control. Despite the target is unknown to the agent team, we provide a proof for convergence of the proposed multi-agent coverage method.", "authors": ["Hossein Rastgoftar"], "categories": ["eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-07-10", "url": "https://arxiv.org/abs/2307.04407", "pdf_url": "https://arxiv.org/pdf/2307.04407v1", "arxiv_id": "2307.04407", "doi": "10.1109/TCNS.2025.3525802", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Control of Network Systems", "quality_score": 0.1747}
{"id": "ea29edc07dc6af64bd27f4d7700d73d250bc503bedc189a558ae10e6577f31c9", "sources": ["arxiv", "semantic_scholar"], "title": "Wireless Multi-Agent Generative AI: From Connected Intelligence to Collective Intelligence", "abstract": "The convergence of generative large language models (LLMs), edge networks, and multi-agent systems represents a groundbreaking synergy that holds immense promise for future wireless generations, harnessing the power of collective intelligence and paving the way for self-governed networks where intelligent decision-making happens right at the edge. This article puts the stepping-stone for incorporating multi-agent generative artificial intelligence (AI) in wireless networks, and sets the scene for realizing on-device LLMs, where multi-agent LLMs are collaboratively planning and solving tasks to achieve a number of network goals. We further investigate the profound limitations of cloud-based LLMs, and explore multi-agent LLMs from a game theoretic perspective, where agents collaboratively solve tasks in competitive environments. Moreover, we establish the underpinnings for the architecture design of wireless multi-agent generative AI systems at the network level and the agent level, and we identify the wireless technologies that are envisioned to play a key role in enabling on-device LLM. To demonstrate the promising potentials of wireless multi-agent generative AI networks, we highlight the benefits that can be achieved when implementing wireless generative agents in intent-based networking, and we provide a case study to showcase how on-device LLMs can contribute to solving network intents in a collaborative fashion. We finally shed lights on potential challenges and sketch a research roadmap towards realizing the vision of wireless collective intelligence.", "authors": ["Hang Zou", "Qiyang Zhao", "Lina Bariah", "Mehdi Bennis", "Merouane Debbah"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-07-06", "url": "https://arxiv.org/abs/2307.02757", "pdf_url": "https://arxiv.org/pdf/2307.02757v1", "arxiv_id": "2307.02757", "doi": "10.48550/arXiv.2307.02757", "citation_count": 70, "influential_citation_count": 4, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4628}
{"id": "348bf2afaef5b1ed0d0904db9eb7eb0fb3b08b7aa4b9faf06156148d72098369", "sources": ["arxiv", "semantic_scholar"], "title": "SACHA: Soft Actor-Critic with Heuristic-Based Attention for Partially Observable Multi-Agent Path Finding", "abstract": "Multi-Agent Path Finding (MAPF) is a crucial component for many large-scale robotic systems, where agents must plan their collision-free paths to their given goal positions. Recently, multi-agent reinforcement learning has been introduced to solve the partially observable variant of MAPF by learning a decentralized single-agent policy in a centralized fashion based on each agent's partial observation. However, existing learning-based methods are ineffective in achieving complex multi-agent cooperation, especially in congested environments, due to the non-stationarity of this setting. To tackle this challenge, we propose a multi-agent actor-critic method called Soft Actor-Critic with Heuristic-Based Attention (SACHA), which employs novel heuristic-based attention mechanisms for both the actors and critics to encourage cooperation among agents. SACHA learns a neural network for each agent to selectively pay attention to the shortest path heuristic guidance from multiple agents within its field of view, thereby allowing for more scalable learning of cooperation. SACHA also extends the existing multi-agent actor-critic framework by introducing a novel critic centered on each agent to approximate $Q$-values. Compared to existing methods that use a fully observable critic, our agent-centered multi-agent actor-critic method results in more impartial credit assignment and better generalizability of the learned policy to MAPF instances with varying numbers of agents and types of environments. We also implement SACHA(C), which embeds a communication module in the agent's policy network to enable information exchange among agents. We evaluate both SACHA and SACHA(C) on a variety of MAPF instances and demonstrate decent improvements over several state-of-the-art learning-based MAPF methods with respect to success rate and solution quality.", "authors": ["Qiushi Lin", "Hang Ma"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-07-05", "url": "https://arxiv.org/abs/2307.02691", "pdf_url": "https://arxiv.org/pdf/2307.02691v1", "arxiv_id": "2307.02691", "doi": "10.1109/LRA.2023.3292004", "citation_count": 32, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "IEEE Robotics and Automation Letters", "quality_score": 0.3796}
{"id": "bfa3b3943866e9bbb2f6c5cfb4b291d0c4d0a5825d12976913dc5dbec511f87e", "sources": ["arxiv", "semantic_scholar"], "title": "Adversarial Search and Tracking with Multiagent Reinforcement Learning in Sparsely Observable Environment", "abstract": "We study a search and tracking (S&T) problem where a team of dynamic search agents must collaborate to track an adversarial, evasive agent. The heterogeneous search team may only have access to a limited number of past adversary trajectories within a large search space. This problem is challenging for both model-based searching and reinforcement learning (RL) methods since the adversary exhibits reactionary and deceptive evasive behaviors in a large space leading to sparse detections for the search agents. To address this challenge, we propose a novel Multi-Agent RL (MARL) framework that leverages the estimated adversary location from our learnable filtering model. We show that our MARL architecture can outperform all baselines and achieves a 46% increase in detection rate.", "authors": ["Zixuan Wu", "Sean Ye", "Manisha Natarajan", "Letian Chen", "Rohan Paleja", "Matthew C. Gombolay"], "categories": ["cs.LG", "cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-20", "url": "https://arxiv.org/abs/2306.11301", "pdf_url": "https://arxiv.org/pdf/2306.11301v2", "arxiv_id": "2306.11301", "doi": "10.1109/MRS60187.2023.10416776", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Symposium on Multi-Robot and Multi-Agent Systems", "quality_score": 0.1505}
{"id": "ff7a797179a51dcd1437ae561cb304b4ba9e2c4e809cd372a37bf3d1d66e83d3", "sources": ["arxiv", "semantic_scholar"], "title": "QCNeXt: A Next-Generation Framework For Joint Multi-Agent Trajectory Prediction", "abstract": "Estimating the joint distribution of on-road agents' future trajectories is essential for autonomous driving. In this technical report, we propose a next-generation framework for joint multi-agent trajectory prediction called QCNeXt. First, we adopt the query-centric encoding paradigm for the task of joint multi-agent trajectory prediction. Powered by this encoding scheme, our scene encoder is equipped with permutation equivariance on the set elements, roto-translation invariance in the space dimension, and translation invariance in the time dimension. These invariance properties not only enable accurate multi-agent forecasting fundamentally but also empower the encoder with the capability of streaming processing. Second, we propose a multi-agent DETR-like decoder, which facilitates joint multi-agent trajectory prediction by modeling agents' interactions at future time steps. For the first time, we show that a joint prediction model can outperform marginal prediction models even on the marginal metrics, which opens up new research opportunities in trajectory prediction. Our approach ranks 1st on the Argoverse 2 multi-agent motion forecasting benchmark, winning the championship of the Argoverse Challenge at the CVPR 2023 Workshop on Autonomous Driving.", "authors": ["Zikang Zhou", "Zihao Wen", "Jianping Wang", "Yung-Hui Li", "Yu-Kai Huang"], "categories": ["cs.CV", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-18", "url": "https://arxiv.org/abs/2306.10508", "pdf_url": "https://arxiv.org/pdf/2306.10508v1", "arxiv_id": "2306.10508", "doi": "10.48550/arXiv.2306.10508", "citation_count": 57, "influential_citation_count": 6, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.4409}
{"id": "aede10c96d70a113f6b140168fa9e40313b1d1dc778b5d187aaf3d20a9d9f26b", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Reinforcement Learning Guided by Signal Temporal Logic Specifications", "abstract": "Reward design is a key component of deep reinforcement learning, yet some tasks and designer's objectives may be unnatural to define as a scalar cost function. Among the various techniques, formal methods integrated with DRL have garnered considerable attention due to their expressiveness and flexibility to define the reward and requirements for different states and actions of the agent. However, how to leverage Signal Temporal Logic (STL) to guide multi-agent reinforcement learning reward design remains unexplored. Complex interactions, heterogeneous goals and critical safety requirements in multi-agent systems make this problem even more challenging. In this paper, we propose a novel STL-guided multi-agent reinforcement learning framework. The STL requirements are designed to include both task specifications according to the objective of each agent and safety specifications, and the robustness values of the STL specifications are leveraged to generate rewards. We validate the advantages of our method through empirical studies. The experimental results demonstrate significant reward performance improvements compared to MARL without STL guidance, along with a remarkable increase in the overall safety rate of the multi-agent systems.", "authors": ["Jiangwei Wang", "Shuo Yang", "Ziyan An", "Songyang Han", "Zhili Zhang", "Rahul Mangharam", "Meiyi Ma", "Fei Miao"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-11", "url": "https://arxiv.org/abs/2306.06808", "pdf_url": "https://arxiv.org/pdf/2306.06808v2", "arxiv_id": "2306.06808", "doi": "10.1109/IROS60139.2025.11246629", "citation_count": 15, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE/RJS International Conference on Intelligent RObots and Systems", "quality_score": 0.301}
{"id": "c046152e5b522c6e2a374ee09afd6bbb58b9948f25d281d8bdb9c44f1878f80c", "sources": ["arxiv", "semantic_scholar"], "title": "Robustness Testing for Multi-Agent Reinforcement Learning: State Perturbations on Critical Agents", "abstract": "Multi-Agent Reinforcement Learning (MARL) has been widely applied in many fields such as smart traffic and unmanned aerial vehicles. However, most MARL algorithms are vulnerable to adversarial perturbations on agent states. Robustness testing for a trained model is an essential step for confirming the trustworthiness of the model against unexpected perturbations. This work proposes a novel Robustness Testing framework for MARL that attacks states of Critical Agents (RTCA). The RTCA has two innovations: 1) a Differential Evolution (DE) based method to select critical agents as victims and to advise the worst-case joint actions on them; and 2) a team cooperation policy evaluation method employed as the objective function for the optimization of DE. Then, adversarial state perturbations of the critical agents are generated based on the worst-case joint actions. This is the first robustness testing framework with varying victim agents. RTCA demonstrates outstanding performance in terms of the number of victim agents and destroying cooperation policies.", "authors": ["Ziyuan Zhou", "Guanjun Liu"], "categories": ["cs.LG", "cs.AI", "cs.CR", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-09", "url": "https://arxiv.org/abs/2306.06136", "pdf_url": "https://arxiv.org/pdf/2306.06136v1", "arxiv_id": "2306.06136", "doi": "10.48550/arXiv.2306.06136", "citation_count": 17, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "European Conference on Artificial Intelligence", "quality_score": 0.3138}
{"id": "fa2a145705a2c3d8c5b44da0e4c009b156f4631c2e17bce813439de015c1bb4f", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Collaboration: Harnessing the Power of Intelligent LLM Agents", "abstract": "In this paper, we present a novel framework for enhancing the capabilities of large language models (LLMs) by leveraging the power of multi-agent systems. Our framework introduces a collaborative environment where multiple intelligent agent components, each with distinctive attributes and roles, work together to handle complex tasks more efficiently and effectively. We demonstrate the practicality and versatility of our framework through case studies in artificial general intelligence (AGI), specifically focusing on the Auto-GPT and BabyAGI models. We also examine the \"Gorilla\" model, which integrates external APIs into the LLM. Our framework addresses limitations and challenges such as looping issues, security risks, scalability, system evaluation, and ethical considerations. By modeling various domains such as courtroom simulations and software development scenarios, we showcase the potential applications and benefits of our proposed multi-agent system. Our framework provides an avenue for advancing the capabilities and performance of LLMs through collaboration and knowledge exchange among intelligent agents.", "authors": ["Yashar Talebirad", "Amirhossein Nadiri"], "categories": ["cs.AI", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-05", "url": "https://arxiv.org/abs/2306.03314", "pdf_url": "https://arxiv.org/pdf/2306.03314v1", "arxiv_id": "2306.03314", "doi": "10.48550/arXiv.2306.03314", "citation_count": 438, "influential_citation_count": 13, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.6606}
{"id": "71f3d26cc759e1398a412cf29ef1bb5b8063ad2373b1868c77bace9577b23e8e", "sources": ["arxiv", "semantic_scholar"], "title": "A Unified Framework for Factorizing Distributional Value Functions for Multi-Agent Reinforcement Learning", "abstract": "In fully cooperative multi-agent reinforcement learning (MARL) settings, environments are highly stochastic due to the partial observability of each agent and the continuously changing policies of other agents. To address the above issues, we proposed a unified framework, called DFAC, for integrating distributional RL with value function factorization methods. This framework generalizes expected value function factorization methods to enable the factorization of return distributions. To validate DFAC, we first demonstrate its ability to factorize the value functions of a simple matrix game with stochastic rewards. Then, we perform experiments on all Super Hard maps of the StarCraft Multi-Agent Challenge and six self-designed Ultra Hard maps, showing that DFAC is able to outperform a number of baselines.", "authors": ["Wei-Fang Sun", "Cheng-Kuang Lee", "Simon See", "Chun-Yi Lee"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-04", "url": "https://arxiv.org/abs/2306.02430", "pdf_url": "https://arxiv.org/pdf/2306.02430v1", "arxiv_id": "2306.02430", "doi": "10.48550/arXiv.2306.02430", "citation_count": 3, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Journal of machine learning research", "quality_score": 0.1505}
{"id": "bb277ce53e5624d5a94a8998a568f44e8ea4abde39887fafcc00ff730ec9f674", "sources": ["arxiv", "semantic_scholar"], "title": "Achieving Fairness in Multi-Agent Markov Decision Processes Using Reinforcement Learning", "abstract": "Fairness plays a crucial role in various multi-agent systems (e.g., communication networks, financial markets, etc.). Many multi-agent dynamical interactions can be cast as Markov Decision Processes (MDPs). While existing research has focused on studying fairness in known environments, the exploration of fairness in such systems for unknown environments remains open. In this paper, we propose a Reinforcement Learning (RL) approach to achieve fairness in multi-agent finite-horizon episodic MDPs. Instead of maximizing the sum of individual agents' value functions, we introduce a fairness function that ensures equitable rewards across agents. Since the classical Bellman's equation does not hold when the sum of individual value functions is not maximized, we cannot use traditional approaches. Instead, in order to explore, we maintain a confidence bound of the unknown environment and then propose an online convex optimization based approach to obtain a policy constrained to this confidence region. We show that such an approach achieves sub-linear regret in terms of the number of episodes. Additionally, we provide a probably approximately correct (PAC) guarantee based on the obtained regret bound. We also propose an offline RL algorithm and bound the optimality gap with respect to the optimal fair solution. To mitigate computational complexity, we introduce a policy-gradient type method for the fair objective. Simulation experiments also demonstrate the efficacy of our approach.", "authors": ["Peizhong Ju", "Arnob Ghosh", "Ness B. Shroff"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-06-01", "url": "https://arxiv.org/abs/2306.00324", "pdf_url": "https://arxiv.org/pdf/2306.00324v1", "arxiv_id": "2306.00324", "doi": "10.48550/arXiv.2306.00324", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.2386}
{"id": "92660930e7c78cb92dafd276a0aaf4aa660091f9c6e944b027c5d91744326843", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Multi-Agent Heterogeneous Multi-Armed Bandits", "abstract": "The study of collaborative multi-agent bandits has attracted significant attention recently. In light of this, we initiate the study of a new collaborative setting, consisting of $N$ agents such that each agent is learning one of $M$ stochastic multi-armed bandits to minimize their group cumulative regret. We develop decentralized algorithms which facilitate collaboration between the agents under two scenarios. We characterize the performance of these algorithms by deriving the per agent cumulative regret and group regret upper bounds. We also prove lower bounds for the group regret in this setting, which demonstrates the near-optimal behavior of the proposed algorithms.", "authors": ["Ronshee Chawla", "Daniel Vial", "Sanjay Shakkottai", "R. Srikant"], "categories": ["cs.LG", "cs.DC", "cs.MA", "cs.SI", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2023-05-30", "url": "https://arxiv.org/abs/2305.18784", "pdf_url": "https://arxiv.org/pdf/2305.18784v2", "arxiv_id": "2305.18784", "doi": "10.48550/arXiv.2305.18784", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.2386}
{"id": "948c1b5c4a8828a8050c86b87d69d2b0699132477da8532e967df17a7b2a28ec", "sources": ["arxiv", "semantic_scholar"], "title": "Collaborative Multi-Agent Video Fast-Forwarding", "abstract": "Multi-agent applications have recently gained significant popularity. In many computer vision tasks, a network of agents, such as a team of robots with cameras, could work collaboratively to perceive the environment for efficient and accurate situation awareness. However, these agents often have limited computation, communication, and storage resources. Thus, reducing resource consumption while still providing an accurate perception of the environment becomes an important goal when deploying multi-agent systems. To achieve this goal, we identify and leverage the overlap among different camera views in multi-agent systems for reducing the processing, transmission and storage of redundant/unimportant video frames. Specifically, we have developed two collaborative multi-agent video fast-forwarding frameworks in distributed and centralized settings, respectively. In these frameworks, each individual agent can selectively process or skip video frames at adjustable paces based on multiple strategies via reinforcement learning. Multiple agents then collaboratively sense the environment via either 1) a consensus-based distributed framework called DMVF that periodically updates the fast-forwarding strategies of agents by establishing communication and consensus among connected neighbors, or 2) a centralized framework called MFFNet that utilizes a central controller to decide the fast-forwarding strategies for agents based on collected data. We demonstrate the efficacy and efficiency of our proposed frameworks on a real-world surveillance video dataset VideoWeb and a new simulated driving dataset CarlaSim, through extensive simulations and deployment on an embedded platform with TCP communication. We show that compared with other approaches in the literature, our frameworks achieve better coverage of important frames, while significantly reducing the number of frames processed at each agent.", "authors": ["Shuyue Lan", "Zhilu Wang", "Ermin Wei", "Amit K. Roy-Chowdhury", "Qi Zhu"], "categories": ["cs.CV"], "fields_of_study": ["Computer Science"], "published_date": "2023-05-27", "url": "https://arxiv.org/abs/2305.17569", "pdf_url": "https://arxiv.org/pdf/2305.17569v1", "arxiv_id": "2305.17569", "doi": "10.1109/TMM.2023.3275853", "citation_count": 6, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE transactions on multimedia", "quality_score": 0.2113}
{"id": "f7bf76332e0f5e61f65bc9fdd3eb59ec0659831b8626fb25309e107564d01936", "sources": ["arxiv", "semantic_scholar"], "title": "Local Optimization Achieves Global Optimality in Multi-Agent Reinforcement Learning", "abstract": "Policy optimization methods with function approximation are widely used in multi-agent reinforcement learning. However, it remains elusive how to design such algorithms with statistical guarantees. Leveraging a multi-agent performance difference lemma that characterizes the landscape of multi-agent policy optimization, we find that the localized action value function serves as an ideal descent direction for each local policy. Motivated by the observation, we present a multi-agent PPO algorithm in which the local policy of each agent is updated similarly to vanilla PPO. We prove that with standard regularity conditions on the Markov game and problem-dependent quantities, our algorithm converges to the globally optimal policy at a sublinear rate. We extend our algorithm to the off-policy setting and introduce pessimism to policy evaluation, which aligns with experiments. To our knowledge, this is the first provably convergent multi-agent PPO algorithm in cooperative Markov games.", "authors": ["Yulai Zhao", "Zhuoran Yang", "Zhaoran Wang", "Jason D. Lee"], "categories": ["cs.LG", "cs.GT", "cs.MA", "stat.ML"], "fields_of_study": ["Computer Science", "Mathematics"], "published_date": "2023-05-08", "url": "https://arxiv.org/abs/2305.04819", "pdf_url": "https://arxiv.org/pdf/2305.04819v1", "arxiv_id": "2305.04819", "doi": "10.48550/arXiv.2305.04819", "citation_count": 9, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.25}
{"id": "f2a41bdc20e857ba7315294a5ed8f737faec48d9461ab5f5c79603617ed5cd40", "sources": ["arxiv", "semantic_scholar"], "title": "Model-free Motion Planning of Autonomous Agents for Complex Tasks in Partially Observable Environments", "abstract": "Motion planning of autonomous agents in partially known environments with incomplete information is a challenging problem, particularly for complex tasks. This paper proposes a model-free reinforcement learning approach to address this problem. We formulate motion planning as a probabilistic-labeled partially observable Markov decision process (PL-POMDP) problem and use linear temporal logic (LTL) to express the complex task. The LTL formula is then converted to a limit-deterministic generalized Büchi automaton (LDGBA). The problem is redefined as finding an optimal policy on the product of PL-POMDP with LDGBA based on model-checking techniques to satisfy the complex task. We implement deep Q learning with long short-term memory (LSTM) to process the observation history and task recognition. Our contributions include the proposed method, the utilization of LTL and LDGBA, and the LSTM-enhanced deep Q learning. We demonstrate the applicability of the proposed method by conducting simulations in various environments, including grid worlds, a virtual office, and a multi-agent warehouse. The simulation results demonstrate that our proposed method effectively addresses environment, action, and observation uncertainties. This indicates its potential for real-world applications, including the control of unmanned aerial vehicles (UAVs).", "authors": ["Junchao Li", "Mingyu Cai", "Zhen Kan", "Shaoping Xiao"], "categories": ["cs.AI", "cs.FL", "cs.MA", "cs.RO", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-04-30", "url": "https://arxiv.org/abs/2305.00561", "pdf_url": "https://arxiv.org/pdf/2305.00561v1", "arxiv_id": "2305.00561", "doi": "10.48550/arXiv.2305.00561", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "2730c2392a766064ac8415fefeae13cac9133ec42919cd7980fcff86d67a479f", "sources": ["arxiv", "semantic_scholar"], "title": "Partially Observable Mean Field Multi-Agent Reinforcement Learning Based on Graph-Attention", "abstract": "Traditional multi-agent reinforcement learning algorithms are difficultly applied in a large-scale multi-agent environment. The introduction of mean field theory has enhanced the scalability of multi-agent reinforcement learning in recent years. This paper considers partially observable multi-agent reinforcement learning (MARL), where each agent can only observe other agents within a fixed range. This partial observability affects the agent's ability to assess the quality of the actions of surrounding agents. This paper focuses on developing a method to capture more effective information from local observations in order to select more effective actions. Previous work in this field employs probability distributions or weighted mean field to update the average actions of neighborhood agents, but it does not fully consider the feature information of surrounding neighbors and leads to a local optimum. In this paper, we propose a novel multi-agent reinforcement learning algorithm, Partially Observable Mean Field Multi-Agent Reinforcement Learning based on Graph-Attention (GAMFQ) to remedy this flaw. GAMFQ uses a graph attention module and a mean field module to describe how an agent is influenced by the actions of other agents at each time step. This graph attention module consists of a graph attention encoder and a differentiable attention mechanism, and this mechanism outputs a dynamic graph to represent the effectiveness of neighborhood agents against central agents. The mean-field module approximates the effect of a neighborhood agent on a central agent as the average effect of effective neighborhood agents. Experiments show that GAMFQ outperforms baselines including the state-of-the-art partially observable mean-field reinforcement learning algorithms. The code for this paper is here \\url{https://github.com/yangmin32/GPMF}.", "authors": ["Min Yang", "Guanjun Liu", "Ziyuan Zhou"], "categories": ["cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-04-25", "url": "https://arxiv.org/abs/2304.12653", "pdf_url": "https://arxiv.org/pdf/2304.12653v4", "arxiv_id": "2304.12653", "doi": "10.3390/drones7070476", "citation_count": 24, "influential_citation_count": 0, "has_code": true, "code_url": "https://github.com/yangmin32/GPMF}", "venue": "Drones", "quality_score": 0.3495}
{"id": "13b94c9db946217a2a8c9e3ca4a443fd1a590ec1fc75a2db3631604b7875520c", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Policy Reciprocity with Theoretical Guarantee", "abstract": "Modern multi-agent reinforcement learning (RL) algorithms hold great potential for solving a variety of real-world problems. However, they do not fully exploit cross-agent knowledge to reduce sample complexity and improve performance. Although transfer RL supports knowledge sharing, it is hyperparameter sensitive and complex. To solve this problem, we propose a novel multi-agent policy reciprocity (PR) framework, where each agent can fully exploit cross-agent policies even in mismatched states. We then define an adjacency space for mismatched states and design a plug-and-play module for value iteration, which enables agents to infer more precise returns. To improve the scalability of PR, deep PR is proposed for continuous control tasks. Moreover, theoretical analysis shows that agents can asymptotically reach consensus through individual perceived rewards and converge to an optimal value function, which implies the stability and effectiveness of PR, respectively. Experimental results on discrete and continuous environments demonstrate that PR outperforms various existing RL and transfer RL methods.", "authors": ["Haozhi Wang", "Yinchuan Li", "Qing Wang", "Yunfeng Shao", "Jianye Hao"], "categories": ["cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-04-12", "url": "https://arxiv.org/abs/2304.05632", "pdf_url": "https://arxiv.org/pdf/2304.05632v1", "arxiv_id": "2304.05632", "doi": "10.48550/arXiv.2304.05632", "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0753}
{"id": "1d4a21a3bc6317d3655a6565ad40882017a4ea556cf0f7c39e0aaae9d6b24f57", "sources": ["arxiv", "semantic_scholar"], "title": "Adaptive parallelization of multi-agent simulations with localized dynamics", "abstract": "Agent-based modelling constitutes a versatile approach to representing and simulating complex systems. Studying large-scale systems is challenging because of the computational time required for the simulation runs: scaling is at least linear in system size (number of agents). Given the inherently modular nature of MABSs, parallel computing is a natural approach to overcoming this challenge. However, because of the shared information and communication between agents, parellelization is not simple. We present a protocol for shared-memory, parallel execution of MABSs. This approach is useful for models that can be formulated in terms of sequential computations, and that involve updates that are localized, in the sense of involving small numbers of agents. The protocol has a bottom-up and asynchronous nature, allowing it to deal with heterogeneous computation in an adaptive, yet graceful manner. We illustrate the potential performance gains on exemplar cultural dynamics and disease spreading MABSs.", "authors": ["Alexandru-Ionuţ Băbeanu", "Tatiana Filatova", "Jan H. Kwakkel", "Neil Yorke-Smith"], "categories": ["cs.DC", "cs.CE", "cs.MA", "physics.soc-ph"], "fields_of_study": ["Computer Science", "Physics"], "published_date": "2023-04-04", "url": "https://arxiv.org/abs/2304.01724", "pdf_url": "https://arxiv.org/pdf/2304.01724v1", "arxiv_id": "2304.01724", "doi": "10.48550/arXiv.2304.01724", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "5da36cc2c3c5f46f18b8b61195df2d3da1219b6f3c1c69923042ceab48ba9400", "sources": ["arxiv", "semantic_scholar"], "title": "Attrition-Aware Adaptation for Multi-Agent Patrolling", "abstract": "Multi-agent patrolling is a key problem in a variety of domains such as intrusion detection, area surveillance, and policing which involves repeated visits by a group of agents to specified points in an environment. While the problem is well-studied, most works do not provide performance guarantees and either do not consider agent attrition or impose significant communication requirements to enable adaptation. In this work, we present the Adaptive Heuristic-based Patrolling Algorithm, which is capable of adaptation to agent loss using minimal communication by taking advantage of Voronoi partitioning, and which meets guaranteed performance bounds. Additionally, we provide new centralized and distributed mathematical programming formulations of the patrolling problem, analyze the properties of Voronoi partitioning, and finally, show the value of our adaptive heuristic algorithm by comparison with various benchmark algorithms using physical robots and simulation based on the Robot Operating System (ROS) 2.", "authors": ["Anthony Goeckner", "Xinliang Li", "Ermin Wei", "Qi Zhu"], "categories": ["cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-04-03", "url": "https://arxiv.org/abs/2304.01386", "pdf_url": "https://arxiv.org/pdf/2304.01386v3", "arxiv_id": "2304.01386", "doi": "10.1109/LRA.2024.3421793", "citation_count": 8, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "IEEE Robotics and Automation Letters", "quality_score": 0.2386}
{"id": "184b23f10c3b44ef608b715fb581b7148e967113b925dd57394a442095ad811b", "sources": ["arxiv", "semantic_scholar"], "title": "A Hierarchical Game-Theoretic Decision-Making for Cooperative Multi-Agent Systems Under the Presence of Adversarial Agents", "abstract": "Underlying relationships among Multi-Agent Systems (MAS) in hazardous scenarios can be represented as Game-theoretic models. This paper proposes a new hierarchical network-based model called Game-theoretic Utility Tree (GUT), which decomposes high-level strategies into executable low-level actions for cooperative MAS decisions. It combines with a new payoff measure based on agent needs for real-time strategy games. We present an Explore game domain, where we measure the performance of MAS achieving tasks from the perspective of balancing the success probability and system costs. We evaluate the GUT approach against state-of-the-art methods that greedily rely on rewards of the composite actions. Conclusive results on extensive numerical simulations indicate that GUT can organize more complex relationships among MAS cooperation, helping the group achieve challenging tasks with lower costs and higher winning rates. Furthermore, we demonstrated the applicability of the GUT using the simulator-hardware testbed - Robotarium. The performances verified the effectiveness of the GUT in the real robot application and validated that the GUT could effectively organize MAS cooperation strategies, helping the group with fewer advantages achieve higher performance.", "authors": ["Qin Yang", "Ramviyas Parasuraman"], "categories": ["cs.MA", "cs.AI", "cs.LG", "cs.RO", "eess.SY"], "fields_of_study": ["Computer Science", "Engineering"], "published_date": "2023-03-28", "url": "https://arxiv.org/abs/2303.16641", "pdf_url": "https://arxiv.org/pdf/2303.16641v1", "arxiv_id": "2303.16641", "doi": "10.1145/3555776.3577642", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "ACM Symposium on Applied Computing", "quality_score": 0.2386}
{"id": "7acf53b8be42a7274ef35e510b12a8f269afae8c593ebbb72f82de18729bb233", "sources": ["arxiv", "semantic_scholar"], "title": "Projected Multi-Agent Consensus Equilibrium (PMACE) with Application to Ptychography", "abstract": "Multi-Agent Consensus Equilibrium (MACE) formulates an inverse imaging problem as a balance among multiple update agents such as data-fitting terms and denoisers. However, each such agent operates on a separate copy of the full image, leading to redundant memory use and slow convergence when each agent affects only a small subset of the full image. In this paper, we extend MACE to Projected Multi-Agent Consensus Equilibrium (PMACE), in which each agent updates only a projected component of the full image, thus greatly reducing memory use for some applications.We describe PMACE in terms of an equilibrium problem and an equivalent fixed point problem and show that in most cases the PMACE equilibrium is not the solution of an optimization problem. To demonstrate the value of PMACE, we apply it to the problem of ptychography, in which a sample is reconstructed from the diffraction patterns resulting from coherent X-ray illumination at multiple overlapping spots. In our PMACE formulation, each spot corresponds to a separate data-fitting agent, with the final solution found as an equilibrium among all the agents. Our results demonstrate that the PMACE reconstruction algorithm generates more accurate reconstructions at a lower computational cost than existing ptychography algorithms when the spots are sparsely sampled.", "authors": ["Qiuchen Zhai", "Gregery T. Buzzard", "Kevin Mertes", "Brendt Wohlberg", "Charles A. Bouman"], "categories": ["math.OC", "eess.IV"], "fields_of_study": ["Computer Science", "Mathematics", "Engineering"], "published_date": "2023-03-28", "url": "https://arxiv.org/abs/2303.15679", "pdf_url": "https://arxiv.org/pdf/2303.15679v2", "arxiv_id": "2303.15679", "doi": "10.1109/TCI.2023.3328288", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "IEEE Transactions on Computational Imaging", "quality_score": 0.2386}
{"id": "25beb45b3864434f9b7e79acd9b9edff2f4619766ec08613a67a935db993c7fe", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-agent Black-box Optimization using a Bayesian Approach to Alternating Direction Method of Multipliers", "abstract": "Bayesian optimization (BO) is a powerful black-box optimization framework that looks to efficiently learn the global optimum of an unknown system by systematically trading-off between exploration and exploitation. However, the use of BO as a tool for coordinated decision-making in multi-agent systems with unknown structure has not been widely studied. This paper investigates a black-box optimization problem over a multi-agent network coupled via shared variables or constraints, where each subproblem is formulated as a BO that uses only its local data. The proposed multi-agent BO (MABO) framework adds a penalty term to traditional BO acquisition functions to account for coupling between the subsystems without data sharing. We derive a suitable form for this penalty term using alternating directions method of multipliers (ADMM), which enables the local decision-making problems to be solved in parallel (and potentially asynchronously). The effectiveness of the proposed MABO method is demonstrated on an intelligent transport system for fuel efficient vehicle platooning.", "authors": ["Dinesh Krishnamoorthy", "Joel A. Paulson"], "categories": ["math.OC", "cs.DC", "cs.MA", "eess.SY"], "fields_of_study": ["Mathematics", "Computer Science", "Engineering"], "published_date": "2023-03-25", "url": "https://arxiv.org/abs/2303.14414", "pdf_url": "https://arxiv.org/pdf/2303.14414v1", "arxiv_id": "2303.14414", "doi": "10.48550/arXiv.2303.14414", "citation_count": 8, "influential_citation_count": 2, "has_code": false, "code_url": null, "venue": "IFAC-PapersOnLine", "quality_score": 0.2386}
{"id": "05750e536ca90f918e0368a143195bf5f4aaec0f8158f2474f7d84e0382d53f1", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Reward Machines in Cooperative Multi-Agent Tasks", "abstract": "This paper presents a novel approach to Multi-Agent Reinforcement Learning (MARL) that combines cooperative task decomposition with the learning of reward machines (RMs) encoding the structure of the sub-tasks. The proposed method helps deal with the non-Markovian nature of the rewards in partially observable environments and improves the interpretability of the learnt policies required to complete the cooperative task. The RMs associated with each sub-task are learnt in a decentralised manner and then used to guide the behaviour of each agent. By doing so, the complexity of a cooperative multi-agent problem is reduced, allowing for more effective learning. The results suggest that our approach is a promising direction for future research in MARL, especially in complex environments with large state spaces and multiple agents.", "authors": ["Leo Ardon", "Daniel Furelos-Blanco", "Alessandra Russo"], "categories": ["cs.AI", "cs.MA", "cs.SC"], "fields_of_study": ["Computer Science"], "published_date": "2023-03-24", "url": "https://arxiv.org/abs/2303.14061", "pdf_url": "https://arxiv.org/pdf/2303.14061v4", "arxiv_id": "2303.14061", "doi": "10.1007/978-3-031-56255-6_3", "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "7e0bbf9f77918044313c3b6d39fc33b98d8ca629a73a44446d025c4a88a840fb", "sources": ["arxiv", "semantic_scholar"], "title": "Coordinating Fully-Cooperative Agents Using Hierarchical Learning Anticipation", "abstract": "Learning anticipation is a reasoning paradigm in multi-agent reinforcement learning, where agents, during learning, consider the anticipated learning of other agents. There has been substantial research into the role of learning anticipation in improving cooperation among self-interested agents in general-sum games. Two primary examples are Learning with Opponent-Learning Awareness (LOLA), which anticipates and shapes the opponent's learning process to ensure cooperation among self-interested agents in various games such as iterated prisoner's dilemma, and Look-Ahead (LA), which uses learning anticipation to guarantee convergence in games with cyclic behaviors. So far, the effectiveness of applying learning anticipation to fully-cooperative games has not been explored. In this study, we aim to research the influence of learning anticipation on coordination among common-interested agents. We first illustrate that both LOLA and LA, when applied to fully-cooperative games, degrade coordination among agents, causing worst-case outcomes. Subsequently, to overcome this miscoordination behavior, we propose Hierarchical Learning Anticipation (HLA), where agents anticipate the learning of other agents in a hierarchical fashion. Specifically, HLA assigns agents to several hierarchy levels to properly regulate their reasonings. Our theoretical and empirical findings confirm that HLA can significantly improve coordination among common-interested agents in fully-cooperative normal-form games. With HLA, to the best of our knowledge, we are the first to unlock the benefits of learning anticipation for fully-cooperative games.", "authors": ["Ariyan Bighashdel", "Daan de Geus", "Pavol Jancura", "Gijs Dubbelman"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-03-15", "url": "https://arxiv.org/abs/2303.08307", "pdf_url": "https://arxiv.org/pdf/2303.08307v2", "arxiv_id": "2303.08307", "doi": "10.48550/arXiv.2303.08307", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.1193}
{"id": "26fa1b2e8c71d10601097d31c3f3503ddcb86996fe34950c79aa1c4cc96a84ee", "sources": ["arxiv", "semantic_scholar"], "title": "Online Control Barrier Functions for Decentralized Multi-Agent Navigation", "abstract": "Control barrier functions (CBFs) enable guaranteed safe multi-agent navigation in the continuous domain. The resulting navigation performance, however, is highly sensitive to the underlying hyperparameters. Traditional approaches consider fixed CBFs (where parameters are tuned apriori), and hence, typically do not perform well in cluttered and highly dynamic environments: conservative parameter values can lead to inefficient agent trajectories, or even failure to reach goal positions, whereas aggressive parameter values can lead to infeasible controls. To overcome these issues, in this paper, we propose online CBFs, whereby hyperparameters are tuned in real-time, as a function of what agents perceive in their immediate neighborhood. Since the explicit relationship between CBFs and navigation performance is hard to model, we leverage reinforcement learning to learn CBF-tuning policies in a model-free manner. Because we parameterize the policies with graph neural networks (GNNs), we are able to synthesize decentralized agent controllers that adjust parameter values locally, varying the degree of conservative and aggressive behaviors across agents. Simulations as well as real-world experiments show that (i) online CBFs are capable of solving navigation scenarios that are infeasible for fixed CBFs, and (ii), that they improve navigation performance by adapting to other agents and changes in the environment.", "authors": ["Zhan Gao", "Guang Yang", "Amanda Prorok"], "categories": ["cs.RO", "cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-03-08", "url": "https://arxiv.org/abs/2303.04313", "pdf_url": "https://arxiv.org/pdf/2303.04313v2", "arxiv_id": "2303.04313", "doi": "10.1109/MRS60187.2023.10416796", "citation_count": 28, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "International Symposium on Multi-Robot and Multi-Agent Systems", "quality_score": 0.3656}
{"id": "915fc6329948e021f2b9843da1904357dd6623b21190f4e9b805065aa35a17fe", "sources": ["arxiv", "semantic_scholar"], "title": "Coordination of Multiple Robots along Given Paths with Bounded Junction Complexity", "abstract": "We study a fundamental NP-hard motion coordination problem for multi-robot/multi-agent systems: We are given a graph $G$ and set of agents, where each agent has a given directed path in $G$. Each agent is initially located on the first vertex of its path. At each time step an agent can move to the next vertex on its path, provided that the vertex is not occupied by another agent. The goal is to find a sequence of such moves along the given paths so that each reaches its target, or to report that no such sequence exists. The problem models guidepath-based transport systems, which is a pertinent abstraction for traffic in a variety of contemporary applications, ranging from train networks or Automated Guided Vehicles (AGVs) in factories, through computer game animations, to qubit transport in quantum computing. It also arises as a sub-problem in the more general multi-robot motion-planning problem. We provide a fine-grained tractability analysis of the problem by considering new assumptions and identifying minimal values of key parameters for which the problem remains NP-hard. Our analysis identifies a critical parameter called vertex multiplicity (VM), defined as the maximum number of paths passing through the same vertex. We show that a prevalent variant of the problem, which is equivalent to Sequential Resource Allocation (concerning deadlock prevention for concurrent processes), is NP-hard even when VM is 3. On the positive side, for VM $\\le$ 2 we give an efficient algorithm that iteratively resolves cycles of blocking relations among agents. We also present a variant that is NP-hard when the VM is 2 even when $G$ is a 2D grid and each path lies in a single grid row or column. By studying highly distilled yet NP-hard variants, we deepen the understanding of what makes the problem intractable and thereby guide the search for efficient solutions under practical assumptions.", "authors": ["Mikkel Abrahamsen", "Tzvika Geft", "Dan Halperin", "Barak Ugav"], "categories": ["cs.RO", "cs.CG", "cs.DS", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-03-01", "url": "https://arxiv.org/abs/2303.00745", "pdf_url": "https://arxiv.org/pdf/2303.00745v1", "arxiv_id": "2303.00745", "doi": "10.48550/arXiv.2303.00745", "citation_count": 6, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.2113}
{"id": "ba338771136c539d9423a2eda5bf50e6bb77d9277192a11a1a1b05353ccba2fb", "sources": ["arxiv", "semantic_scholar"], "title": "Causal Explanations for Sequential Decision-Making in Multi-Agent Systems", "abstract": "We present CEMA: Causal Explanations in Multi-Agent systems; a framework for creating causal natural language explanations of an agent's decisions in dynamic sequential multi-agent systems to build more trustworthy autonomous agents. Unlike prior work that assumes a fixed causal structure, CEMA only requires a probabilistic model for forward-simulating the state of the system. Using such a model, CEMA simulates counterfactual worlds that identify the salient causes behind the agent's decisions. We evaluate CEMA on the task of motion planning for autonomous driving and test it in diverse simulated scenarios. We show that CEMA correctly and robustly identifies the causes behind the agent's decisions, even when a large number of other agents is present, and show via a user study that CEMA's explanations have a positive effect on participants' trust in autonomous vehicles and are rated as high as high-quality baseline explanations elicited from other participants. We release the collected explanations with annotations as the HEADD dataset.", "authors": ["Balint Gyevnar", "Cheng Wang", "Christopher G. Lucas", "Shay B. Cohen", "Stefano V. Albrecht"], "categories": ["cs.AI", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-21", "url": "https://arxiv.org/abs/2302.10809", "pdf_url": "https://arxiv.org/pdf/2302.10809v4", "arxiv_id": "2302.10809", "doi": "10.5555/3635637.3662930", "citation_count": 17, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.3138}
{"id": "8454ab4622bd6879f8a1ca8d2d5d401031aaeaf0593348355fdf0be9fbdc7823", "sources": ["arxiv", "semantic_scholar"], "title": "MANSA: Learning Fast and Slow in Multi-Agent Systems", "abstract": "In multi-agent reinforcement learning (MARL), independent learning (IL) often shows remarkable performance and easily scales with the number of agents. Yet, using IL can be inefficient and runs the risk of failing to successfully train, particularly in scenarios that require agents to coordinate their actions. Using centralised learning (CL) enables MARL agents to quickly learn how to coordinate their behaviour but employing CL everywhere is often prohibitively expensive in real-world applications. Besides, using CL in value-based methods often needs strong representational constraints (e.g. individual-global-max condition) that can lead to poor performance if violated. In this paper, we introduce a novel plug & play IL framework named Multi-Agent Network Selection Algorithm (MANSA) which selectively employs CL only at states that require coordination. At its core, MANSA has an additional agent that uses switching controls to quickly learn the best states to activate CL during training, using CL only where necessary and vastly reducing the computational burden of CL. Our theory proves MANSA preserves cooperative MARL convergence properties, boosts IL performance and can optimally make use of a fixed budget on the number CL calls. We show empirically in Level-based Foraging (LBF) and StarCraft Multi-agent Challenge (SMAC) that MANSA achieves fast, superior and more reliable performance while making 40% fewer CL calls in SMAC and using CL at only 1% CL calls in LBF.", "authors": ["David Mguni", "Haojun Chen", "Taher Jafferjee", "Jianhong Wang", "Long Fei", "Xidong Feng", "Stephen McAleer", "Feifei Tong", "Jun Wang", "Yaodong Yang"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-12", "url": "https://arxiv.org/abs/2302.05910", "pdf_url": "https://arxiv.org/pdf/2302.05910v3", "arxiv_id": "2302.05910", "doi": "10.48550/arXiv.2302.05910", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Machine Learning", "quality_score": 0.1747}
{"id": "2f3223fd40829153a60a2a149ddd568e15936126ef773b8fde9a9d5afab36c53", "sources": ["arxiv", "semantic_scholar"], "title": "Learning cooperative behaviours in adversarial multi-agent systems", "abstract": "This work extends an existing virtual multi-agent platform called RoboSumo to create TripleSumo -- a platform for investigating multi-agent cooperative behaviors in continuous action spaces, with physical contact in an adversarial environment. In this paper we investigate a scenario in which two agents, namely `Bug' and `Ant', must team up and push another agent `Spider' out of the arena. To tackle this goal, the newly added agent `Bug' is trained during an ongoing match between `Ant' and `Spider'. `Bug' must develop awareness of the other agents' actions, infer the strategy of both sides, and eventually learn an action policy to cooperate. The reinforcement learning algorithm Deep Deterministic Policy Gradient (DDPG) is implemented with a hybrid reward structure combining dense and sparse rewards. The cooperative behavior is quantitatively evaluated by the mean probability of winning the match and mean number of steps needed to win.", "authors": ["Ni Wang", "Gautham P. Das", "Alan G. Millard"], "categories": ["cs.AI", "cs.MA", "cs.RO"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-10", "url": "https://arxiv.org/abs/2302.05528", "pdf_url": "https://arxiv.org/pdf/2302.05528v1", "arxiv_id": "2302.05528", "doi": "10.1007/978-3-031-15908-4_15", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Towards Autonomous Robotic Systems", "quality_score": 0.0}
{"id": "6da096c39005dedb3c406251fec7ffe1fa09f53456a6b7800a6fab3b981627a3", "sources": ["arxiv", "semantic_scholar"], "title": "Learning Graph-Enhanced Commander-Executor for Multi-Agent Navigation", "abstract": "This paper investigates the multi-agent navigation problem, which requires multiple agents to reach the target goals in a limited time. Multi-agent reinforcement learning (MARL) has shown promising results for solving this issue. However, it is inefficient for MARL to directly explore the (nearly) optimal policy in the large search space, which is exacerbated as the agent number increases (e.g., 10+ agents) or the environment is more complex (e.g., 3D simulator). Goal-conditioned hierarchical reinforcement learning (HRL) provides a promising direction to tackle this challenge by introducing a hierarchical structure to decompose the search space, where the low-level policy predicts primitive actions in the guidance of the goals derived from the high-level policy. In this paper, we propose Multi-Agent Graph-Enhanced Commander-Executor (MAGE-X), a graph-based goal-conditioned hierarchical method for multi-agent navigation tasks. MAGE-X comprises a high-level Goal Commander and a low-level Action Executor. The Goal Commander predicts the probability distribution of goals and leverages them to assign each agent the most appropriate final target. The Action Executor utilizes graph neural networks (GNN) to construct a subgraph for each agent that only contains crucial partners to improve cooperation. Additionally, the Goal Encoder in the Action Executor captures the relationship between the agent and the designated goal to encourage the agent to reach the final target. The results show that MAGE-X outperforms the state-of-the-art MARL baselines with a 100% success rate with only 3 million training steps in multi-agent particle environments (MPE) with 50 agents, and at least a 12% higher success rate and 2x higher data efficiency in a more complicated quadrotor 3D navigation task.", "authors": ["Xinyi Yang", "Shiyu Huang", "Yiwen Sun", "Yuxiang Yang", "Chao Yu", "Wei-Wei Tu", "Huazhong Yang", "Yu Wang"], "categories": ["cs.RO", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-08", "url": "https://arxiv.org/abs/2302.04094", "pdf_url": "https://arxiv.org/pdf/2302.04094v1", "arxiv_id": "2302.04094", "doi": "10.48550/arXiv.2302.04094", "citation_count": 13, "influential_citation_count": 3, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.301}
{"id": "667b501df0ae21307846c6c9c216a2dfe148d450cff45783b9030bbb442344af", "sources": ["arxiv", "semantic_scholar"], "title": "Uncoupled Learning of Differential Stackelberg Equilibria with Commitments", "abstract": "In multi-agent problems requiring a high degree of cooperation, success often depends on the ability of the agents to adapt to each other's behavior. A natural solution concept in such settings is the Stackelberg equilibrium, in which the ``leader'' agent selects the strategy that maximizes its own payoff given that the ``follower'' agent will choose their best response to this strategy. Recent work has extended this solution concept to two-player differentiable games, such as those arising from multi-agent deep reinforcement learning, in the form of the \\textit{differential} Stackelberg equilibrium. While this previous work has presented learning dynamics which converge to such equilibria, these dynamics are ``coupled'' in the sense that the learning updates for the leader's strategy require some information about the follower's payoff function. As such, these methods cannot be applied to truly decentralised multi-agent settings, particularly ad hoc cooperation, where each agent only has access to its own payoff function. In this work we present ``uncoupled'' learning dynamics based on zeroth-order gradient estimators, in which each agent's strategy update depends only on their observations of the other's behavior. We analyze the convergence of these dynamics in general-sum games, and prove that they converge to differential Stackelberg equilibria under the same conditions as previous coupled methods. Furthermore, we present an online mechanism by which symmetric learners can negotiate leader-follower roles. We conclude with a discussion of the implications of our work for multi-agent reinforcement learning and ad hoc collaboration more generally.", "authors": ["Robert Loftin", "Mustafa Mert Çelikok", "Herke van Hoof", "Samuel Kaski", "Frans A. Oliehoek"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-07", "url": "https://arxiv.org/abs/2302.03438", "pdf_url": "https://arxiv.org/pdf/2302.03438v2", "arxiv_id": "2302.03438", "doi": "10.5555/3635637.3662984", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.1193}
{"id": "22174c264b96bc4483b4fca287bb0d2b2b8cd8a331db1c9c367078e62467c446", "sources": ["arxiv", "semantic_scholar"], "title": "Ensemble Value Functions for Efficient Exploration in Multi-Agent Reinforcement Learning", "abstract": "Multi-agent reinforcement learning (MARL) requires agents to explore within a vast joint action space to find joint actions that lead to coordination. Existing value-based MARL algorithms commonly rely on random exploration, such as $ε$-greedy, to explore the environment which is not systematic and inefficient at identifying effective actions in multi-agent problems. Additionally, the concurrent training of the policies of multiple agents during training can render the optimisation non-stationary. This can lead to unstable value estimates, highly variant gradients, and ultimately hinder coordination between agents. To address these challenges, we propose ensemble value functions for multi-agent exploration (EMAX). EMAX is a framework to seamlessly extend value-based MARL algorithms. EMAX leverages an ensemble of value functions for each agent to guide their exploration, reduce the variance of their optimisation, and makes their policies more robust to miscoordination. EMAX achieves these benefits by (1) systematically guiding the exploration of agents with a UCB policy towards parts of the environment that require multiple agents to coordinate. (2) EMAX computes average value estimates across the ensemble as target values to reduce the variance of gradients and make optimisation more stable. (3) During evaluation, EMAX selects actions following a majority vote across the ensemble to reduce the likelihood of miscoordination. We first instantiate independent DQN with EMAX and evaluate it in 11 general-sum tasks with sparse rewards. We show that EMAX improves final evaluation returns by 185% across all tasks. We then evaluate EMAX on top of IDQN, VDN and QMIX in 21 common-reward tasks, and show that EMAX improves sample efficiency and final evaluation returns across all tasks over all three vanilla algorithms by 60%, 47%, and 538%, respectively.", "authors": ["Lukas Schäfer", "Oliver Slumbers", "Stephen McAleer", "Yali Du", "Stefano V. Albrecht", "David Mguni"], "categories": ["cs.MA", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-07", "url": "https://arxiv.org/abs/2302.03439", "pdf_url": "https://arxiv.org/pdf/2302.03439v7", "arxiv_id": "2302.03439", "doi": "10.48550/arXiv.2302.03439", "citation_count": 10, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.2603}
{"id": "6c7edca278dc978c0ee44dcb63315a3e82417849f1f63394dbf60b533c5cb07e", "sources": ["arxiv", "semantic_scholar"], "title": "Off-the-Grid MARL: Datasets with Baselines for Offline Multi-Agent Reinforcement Learning", "abstract": "Being able to harness the power of large datasets for developing cooperative multi-agent controllers promises to unlock enormous value for real-world applications. Many important industrial systems are multi-agent in nature and are difficult to model using bespoke simulators. However, in industry, distributed processes can often be recorded during operation, and large quantities of demonstrative data stored. Offline multi-agent reinforcement learning (MARL) provides a promising paradigm for building effective decentralised controllers from such datasets. However, offline MARL is still in its infancy and therefore lacks standardised benchmark datasets and baselines typically found in more mature subfields of reinforcement learning (RL). These deficiencies make it difficult for the community to sensibly measure progress. In this work, we aim to fill this gap by releasing off-the-grid MARL (OG-MARL): a growing repository of high-quality datasets with baselines for cooperative offline MARL research. Our datasets provide settings that are characteristic of real-world systems, including complex environment dynamics, heterogeneous agents, non-stationarity, many agents, partial observability, suboptimality, sparse rewards and demonstrated coordination. For each setting, we provide a range of different dataset types (e.g. Good, Medium, Poor, and Replay) and profile the composition of experiences for each dataset. We hope that OG-MARL will serve the community as a reliable source of datasets and help drive progress, while also providing an accessible entry point for researchers new to the field.", "authors": ["Claude Formanek", "Asad Jeewa", "Jonathan Shock", "Arnu Pretorius"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-02-01", "url": "https://arxiv.org/abs/2302.00521", "pdf_url": "https://arxiv.org/pdf/2302.00521v2", "arxiv_id": "2302.00521", "doi": null, "citation_count": 8, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.2386}
{"id": "85f333a655c0610b240094eb962574d9a0669d88a4c1b1ef8ce58672adb0ed0e", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Contract Design: How to Commission Multiple Agents with Individual Outcome", "abstract": "We study hidden-action principal-agent problems with multiple agents. These are problems in which a principal commits to an outcome-dependent payment scheme in order to incentivize some agents to take costly, unobservable actions that lead to favorable outcomes. Previous works on multi-agent problems study models where the principal observes a single outcome determined by the actions of all the agents. Such models considerably limit the contracting power of the principal, since payments can only depend on the joint result of all the agents' actions, and there is no way of paying each agent for their individual result. In this paper, we consider a model in which each agent determines their own individual outcome as an effect of their action only, the principal observes all the individual outcomes separately, and they perceive a reward that jointly depends on all these outcomes. This considerably enhances the principal's contracting capabilities, by allowing them to pay each agent on the basis of their individual result. We analyze the computational complexity of finding principal-optimal contracts, revolving around two newly-introduced properties of principal's rewards, which we call IR-supermodularity and DR-submodularity. Intuitively, the former captures settings with increasing returns, where the rewards grow faster as the agents' effort increases, while the latter models the case of diminishing returns, in which rewards grow slower instead. These two properties naturally model two common real-world phenomena, namely diseconomies and economies of scale. In this paper, we first address basic instances in which the principal knows everything about the agents, and, then, more general Bayesian instances where each agent has their own private type determining their features, such as action costs and how actions stochastically determine individual outcomes.", "authors": ["Matteo Castiglioni", "Alberto Marchesi", "Nicola Gatti"], "categories": ["cs.GT"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-31", "url": "https://arxiv.org/abs/2301.13654", "pdf_url": "https://arxiv.org/pdf/2301.13654v1", "arxiv_id": "2301.13654", "doi": "10.1145/3580507.3597793", "citation_count": 38, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "ACM Conference on Economics and Computation", "quality_score": 0.3978}
{"id": "90a4b0f6ea6f2b7a4c1fae1ac304e51c263abdec892917251801462b1fffe699", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Congestion Cost Minimization With Linear Function Approximations", "abstract": "This work considers multiple agents traversing a network from a source node to the goal node. The cost to an agent for traveling a link has a private as well as a congestion component. The agent's objective is to find a path to the goal node with minimum overall cost in a decentralized way. We model this as a fully decentralized multi-agent reinforcement learning problem and propose a novel multi-agent congestion cost minimization (MACCM) algorithm. Our MACCM algorithm uses linear function approximations of transition probabilities and the global cost function. In the absence of a central controller and to preserve privacy, agents communicate the cost function parameters to their neighbors via a time-varying communication network. Moreover, each agent maintains its estimate of the global state-action value, which is updated via a multi-agent extended value iteration (MAEVI) sub-routine. We show that our MACCM algorithm achieves a sub-linear regret. The proof requires the convergence of cost function parameters, the MAEVI algorithm, and analysis of the regret bounds induced by the MAEVI triggering condition for each agent. We implement our algorithm on a two node network with multiple links to validate it. We first identify the optimal policy, the optimal number of agents going to the goal node in each period. We observe that the average regret is close to zero for 2 and 3 agents. The optimal policy captures the trade-off between the minimum cost of staying at a node and the congestion cost of going to the goal node. Our work is a generalization of learning the stochastic shortest path problem.", "authors": ["Prashant Trivedi", "Nandyala Hemachandra"], "categories": ["cs.LG", "cs.AI", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-26", "url": "https://arxiv.org/abs/2301.10993", "pdf_url": "https://arxiv.org/pdf/2301.10993v2", "arxiv_id": "2301.10993", "doi": "10.48550/arXiv.2301.10993", "citation_count": 2, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "International Conference on Artificial Intelligence and Statistics", "quality_score": 0.1193}
{"id": "fdb973443104861e8de3e5b74aa2ab4505e5cdde5c84c8908f50275041cd8fe2", "sources": ["arxiv", "semantic_scholar"], "title": "Periodic Multi-Agent Path Planning", "abstract": "Multi-agent path planning (MAPP) is the problem of planning collision-free trajectories from start to goal locations for a team of agents. This work explores a relatively unexplored setting of MAPP where streams of agents have to go through the starts and goals with high throughput. We tackle this problem by formulating a new variant of MAPP called periodic MAPP in which the timing of agent appearances is periodic. The objective with periodic MAPP is to find a periodic plan, a set of collision-free trajectories that the agent streams can use repeatedly over periods, with periods that are as small as possible. To meet this objective, we propose a solution method that is based on constraint relaxation and optimization. We show that the periodic plans once found can be used for a more practical case in which agents in a stream can appear at random times. We confirm the effectiveness of our method compared with baseline methods in terms of throughput in several scenarios that abstract autonomous intersection management tasks.", "authors": ["Kazumi Kasaura", "Ryo Yonetani", "Mai Nishimura"], "categories": ["cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-26", "url": "https://arxiv.org/abs/2301.10910", "pdf_url": "https://arxiv.org/pdf/2301.10910v2", "arxiv_id": "2301.10910", "doi": "10.1609/aaai.v37i5.25762", "citation_count": 4, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "AAAI Conference on Artificial Intelligence", "quality_score": 0.1747}
{"id": "82be53695dd8f057fd755d0aa83a30ea1de1e2d319ee042a3635f4ef76ba52c3", "sources": ["arxiv", "semantic_scholar"], "title": "Decentralized Multi-agent Filtering", "abstract": "This paper addresses the considerations that comes along with adopting decentralized communication for multi-agent localization applications in discrete state spaces. In this framework, we extend the original formulation of the Bayes filter, a foundational probabilistic tool for discrete state estimation, by appending a step of greedy belief sharing as a method to propagate information and improve local estimates' posteriors. We apply our work in a model-based multi-agent grid-world setting, where each agent maintains a belief distribution for every agents' state. Our results affirm the utility of our proposed extensions for decentralized collaborative tasks. The code base for this work is available in the following repo", "authors": ["Dom Huh", "Prasant Mohapatra"], "categories": ["cs.MA", "cs.AI"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-21", "url": "https://arxiv.org/abs/2301.08864", "pdf_url": "https://arxiv.org/pdf/2301.08864v1", "arxiv_id": "2301.08864", "doi": "10.48550/arXiv.2301.08864", "citation_count": 0, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "arXiv.org", "quality_score": 0.0}
{"id": "ad02fef0c1a6198bdd5a378cbcd1d72bb2b6b12f53a1642764fabedb4a72fe44", "sources": ["arxiv", "semantic_scholar"], "title": "Investigating the Impact of Direct Punishment on the Emergence of Cooperation in Multi-Agent Reinforcement Learning Systems", "abstract": "Solving the problem of cooperation is fundamentally important for the creation and maintenance of functional societies. Problems of cooperation are omnipresent within human society, with examples ranging from navigating busy road junctions to negotiating treaties. As the use of AI becomes more pervasive throughout society, the need for socially intelligent agents capable of navigating these complex cooperative dilemmas is becoming increasingly evident. Direct punishment is a ubiquitous social mechanism that has been shown to foster the emergence of cooperation in both humans and non-humans. In the natural world, direct punishment is often strongly coupled with partner selection and reputation and used in conjunction with third-party punishment. The interactions between these mechanisms could potentially enhance the emergence of cooperation within populations. However, no previous work has evaluated the learning dynamics and outcomes emerging from Multi-Agent Reinforcement Learning (MARL) populations that combine these mechanisms. This paper addresses this gap. It presents a comprehensive analysis and evaluation of the behaviors and learning dynamics associated with direct punishment, third-party punishment, partner selection, and reputation. Finally, we discuss the implications of using these mechanisms on the design of cooperative AI systems.", "authors": ["Nayana Dasgupta", "Mirco Musolesi"], "categories": ["cs.MA", "cs.AI", "cs.LG"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-19", "url": "https://arxiv.org/abs/2301.08278", "pdf_url": "https://arxiv.org/pdf/2301.08278v3", "arxiv_id": "2301.08278", "doi": "10.1007/s10458-025-09698-5", "citation_count": 12, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": "Autonomous Agents and Multi-Agent Systems", "quality_score": 0.2785}
{"id": "dd310da5f8ed205e8184b20263d109b4614ee5a056fdb6cd26480bafa4095fea", "sources": ["arxiv", "semantic_scholar"], "title": "Multi-Agent Coordination Fluid Flow Modeling and Experimental Evaluation", "abstract": "Reliability is a critical aspect of multi-agent system coordination as it ensures that the system functions correctly and consistently. If one agent in the system fails or behaves unexpectedly, it can negatively impact the performance and effectiveness of the entire system. Therefore, it is important to design and implement multi-agent systems with a high level of reliability to ensure that they can operate safely and move smoothly in the presence of unforeseen agent failure or lack of communication with some agent teams moving in a shared motion space. This paper presents a novel fluid flow navigation model that, in an ideal fluid flow, divides agents into cooperative (non-singular) and noncooperative (singular) agents, with cooperative agents sliding along streamlines safely enclosing noncooperative agents in a shared motion space. A series of flight experiments utilizing crazyflie quadcopters will experimentally validate the suggested model.", "authors": ["Harshvardhan Uppaluru", "Mohammad Ghuran", "Hossein Rastgoftar"], "categories": ["eess.SY"], "fields_of_study": ["Engineering", "Computer Science"], "published_date": "2023-01-14", "url": "https://arxiv.org/abs/2301.05833", "pdf_url": "https://arxiv.org/pdf/2301.05833v3", "arxiv_id": "2301.05833", "doi": null, "citation_count": 1, "influential_citation_count": 0, "has_code": false, "code_url": null, "venue": null, "quality_score": 0.0753}
{"id": "84933c2328030669510ca0cac34782acff329f4a1ab57792f893fcef55b611bf", "sources": ["arxiv", "semantic_scholar"], "title": "Self-Motivated Multi-Agent Exploration", "abstract": "In cooperative multi-agent reinforcement learning (CMARL), it is critical for agents to achieve a balance between self-exploration and team collaboration. However, agents can hardly accomplish the team task without coordination and they would be trapped in a local optimum where easy cooperation is accessed without enough individual exploration. Recent works mainly concentrate on agents' coordinated exploration, which brings about the exponentially grown exploration of the state space. To address this issue, we propose Self-Motivated Multi-Agent Exploration (SMMAE), which aims to achieve success in team tasks by adaptively finding a trade-off between self-exploration and team cooperation. In SMMAE, we train an independent exploration policy for each agent to maximize their own visited state space. Each agent learns an adjustable exploration probability based on the stability of the joint team policy. The experiments on highly cooperative tasks in StarCraft II micromanagement benchmark (SMAC) demonstrate that SMMAE can explore task-related states more efficiently, accomplish coordinated behaviours and boost the learning performance.", "authors": ["Shaowei Zhang", "Jiahan Cao", "Lei Yuan", "Yang Yu", "De-Chuan Zhan"], "categories": ["cs.LG", "cs.MA"], "fields_of_study": ["Computer Science"], "published_date": "2023-01-05", "url": "https://arxiv.org/abs/2301.02083", "pdf_url": "https://arxiv.org/pdf/2301.02083v2", "arxiv_id": "2301.02083", "doi": "10.48550/arXiv.2301.02083", "citation_count": 11, "influential_citation_count": 1, "has_code": false, "code_url": null, "venue": "Adaptive Agents and Multi-Agent Systems", "quality_score": 0.2698}