id
string
sources
list
title
string
abstract
string
authors
list
categories
list
fields_of_study
list
published_date
timestamp[s]
url
string
pdf_url
string
arxiv_id
string
doi
string
citation_count
int64
influential_citation_count
int64
has_code
bool
code_url
string
venue
string
quality_score
float64
7c1861f9290144204b1de13a9cb05b4c0c3e0db37d6610f72e07bcdd20446c77
[ "arxiv", "semantic_scholar" ]
TO-Agents: A Multi-Agent AI Pipeline for Preference-Guided Topology Optimization
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 conn...
[ "Isabella A. Stewart", "Hongrui Chen", "Faez Ahmed" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-20T00:00:00
https://arxiv.org/abs/2605.21622
https://arxiv.org/pdf/2605.21622v1
2605.21622
null
0
0
false
null
null
0.35
de557b2ce0d60421f5c4aa6a1c64b0bb0bb9d777d12626419e59e6ae3b6df601
[ "arxiv", "semantic_scholar" ]
From Automated to Autonomous: Hierarchical Agent-native Network Architecture (HANA)
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 hig...
[ "Binghan Wu", "Shoufeng Wang", "Yunxin Liu", "Ya-Qin Zhang", "Joseph Sifakis", "Ye Ouyang" ]
[ "cs.AI", "cs.NI" ]
[ "Computer Science" ]
2026-05-20T00:00:00
https://arxiv.org/abs/2605.20608
https://arxiv.org/pdf/2605.20608v1
2605.20608
10.1109/LNET.2026.3693226
0
0
false
null
IEEE Networking Letters
0.55
2bc6115a3e47a32f10a28b6777dedacd7924f55404133735a9d5b73f2e64f6db
[ "arxiv", "semantic_scholar" ]
MemGym: a Long-Horizon Memory Environment for LLM Agents
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 ...
[ "Wujiang Xu", "Yu Wang", "Kai Mei", "Kaiqu Liang", "Zhenting Wang", "Mingyu Jin", "Han Zhang", "Shi-Xiong Zhang", "Wenyue Hua", "Sambit Sahu", "Dimitris N. Metaxas" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-20T00:00:00
https://arxiv.org/abs/2605.20833
https://arxiv.org/pdf/2605.20833v1
2605.20833
null
0
0
false
null
null
0.35
5a3f2686f860fd6c68f956ce490947a40a43f4b8d401c914e3c12963fb3c1f3f
[ "arxiv", "semantic_scholar" ]
CASPIAN: Online Detection and Attribution of Cascade Attacks in LLM Multi-Agent Systems via Cross-Channel Causal Monitoring
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 app...
[ "Kavana Venkatesh", "Jafar Isbarov", "Saad Amin", "Murat Kantarcioglu", "Jiaming Cui" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.19240
https://arxiv.org/pdf/2605.19240v1
2605.19240
null
0
0
true
https://github.com/caspian-detector/caspian
null
0.65
ef0ec7457174e454cb2e39324d3814496beadcb5663ec786a17d1b7eaf151c74
[ "arxiv", "semantic_scholar" ]
Multi-agent Collaboration with State Management
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 (...
[ "Mengyang Liu", "Taozhi Chen", "Zhenhua Xu", "Xue Jiang", "Yihong Dong" ]
[ "cs.MA", "cs.AI", "cs.CL", "cs.LG", "cs.SE" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.20563
https://arxiv.org/pdf/2605.20563v1
2605.20563
null
0
0
false
null
null
0.35
c39ea1406741a7a1de6a94cb686478588f16a41a9f778225b485d9756bde1543
[ "arxiv", "semantic_scholar" ]
EngiAI: A Multi-Agent Framework and Benchmark Suite for LLM-Driven Engineering Design
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 benc...
[ "Gioele Molinari", "Florian Felten", "Soheyl Massoudi", "Mark Fuge" ]
[ "cs.AI", "cs.LG", "cs.MA" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.19743
https://arxiv.org/pdf/2605.19743v2
2605.19743
null
0
0
true
null
null
0.65
9e1ff3183cd1fe9e45dde36c5e042e33a209985747d984f82ddf80b2dba8172f
[ "arxiv", "semantic_scholar" ]
A Methodology for Selecting and Composing Runtime Architecture Patterns for Production LLM Agents
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, ...
[ "Vasundra Srinivasan" ]
[ "cs.AI", "cs.SE" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.20173
https://arxiv.org/pdf/2605.20173v1
2605.20173
null
0
0
true
https://github.com/vasundras/agent-runtime-patterns
null
0.65
a9fd92403376284f90d4f45da3b75461842a9160cc00aa9cc2905f5c7b2446cf
[ "arxiv", "semantic_scholar" ]
ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning
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 wi...
[ "Zuhao Yang", "Kaichen Zhang", "Sudong Wang", "Keming Wu", "Zhongyu Yang", "Bo Li", "Xiaojuan Qi", "Shijian Lu", "Xingxuan Li", "Lidong Bing" ]
[ "cs.CV" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.20342
https://arxiv.org/pdf/2605.20342v2
2605.20342
null
1
0
false
null
null
0.35
c14a952a4a42efb330843fb7d5cdae09ff8aa24c1ce0c2ea16595a1712a5037c
[ "arxiv", "semantic_scholar" ]
Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs
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, mu...
[ "Haiquan Lu", "Zigeng Chen", "Gongfan Fang", "Xinyin Ma", "Xinchao Wang" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.20315
https://arxiv.org/pdf/2605.20315v1
2605.20315
null
0
0
false
null
null
0.35
6d565b4c9ea15c8d07c6fd6209eb28aa2ce85af8087235e3edffaceb5324ff51
[ "arxiv", "semantic_scholar" ]
APS: Bias-Controlled Adaptive Prototype Simulation for Population-Scale LLM Agents
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...
[ "Quan Zheng", "Yan Gao", "Shaobin He", "Haoxiang Guan", "Yuanhe Tian", "Jie Feng", "Ming Wang", "Shuxin Zheng", "Zhen Liu" ]
[ "cs.MA", "cs.CY" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.27419
https://arxiv.org/pdf/2605.27419v1
2605.27419
null
0
0
false
null
null
0.35
4e93119299b2bde502e7c5226ecef80199853597790c03ebe91e5e88702c00a5
[ "arxiv", "semantic_scholar" ]
Conflict-Resilient Multi-Agent Reasoning via Signed Graph Modeling
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 ...
[ "Longgang He", "Longzhu He", "Daojing He", "Chaozhuo Li" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-19T00:00:00
https://arxiv.org/abs/2605.19418
https://arxiv.org/pdf/2605.19418v1
2605.19418
null
0
0
false
null
null
0.35
fc6cac82eafe90e40fa6277d56461957ed8237b6d6afb421a66e82150a939b23
[ "arxiv", "semantic_scholar" ]
Code as Agent Harness
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 reasoni...
[ "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", "S...
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.18747
https://arxiv.org/pdf/2605.18747v1
2605.18747
null
3
0
true
https://github.com/YennNing/Awesome-Code-as-Agent-Harness-Papers
null
0.65
de5d4ba67a1868992899afe32f4bffff0b1a56682bdd6953bcc6e4e36d33e01b
[ "arxiv", "semantic_scholar" ]
PROTEA: Offline Evaluation and Iterative Refinement for Multi-Agent LLM Workflows
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 ...
[ "Kazuki Kawamura", "Satoshi Waki", "Kei Tateno" ]
[ "cs.CL", "cs.AI", "cs.HC", "cs.SE" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.18032
https://arxiv.org/pdf/2605.18032v1
2605.18032
null
0
0
false
null
null
0.35
9d6beb984b78c83f21788c676031df5c83b5e0d183528dbf74c0806cce62517f
[ "arxiv", "semantic_scholar" ]
Sequential Consensus for Multi-Agent LLM Debates: A Wald-SPRT compute governor with calibration-based failure detection
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...
[ "Andrea Morandi" ]
[ "cs.LG" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.19193
https://arxiv.org/pdf/2605.19193v1
2605.19193
null
0
0
false
null
null
0.35
5111e7da38474f17944b806c2c9e8887d06ed088809c1ce76b0512b4cf9023d5
[ "arxiv", "semantic_scholar" ]
Harnessing LLM Agents with Skill Programs
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 age...
[ "Hongjun Liu", "Yifei Ming", "Shafiq Joty", "Chen Zhao" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.17734
https://arxiv.org/pdf/2605.17734v1
2605.17734
null
1
0
false
null
null
0.35
c1b2aaf032e7abb83c3247cb58d55608c964cdcc523e1fd00b81ac35851ddd9e
[ "arxiv", "semantic_scholar" ]
LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning
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 leverag...
[ "Sangjun Bae", "Yisak Park", "Sanghyeon Lee", "Seungyul Han" ]
[ "cs.AI", "cs.LG", "cs.MA" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.18077
https://arxiv.org/pdf/2605.18077v2
2605.18077
null
1
0
false
null
null
0.35
636f8f454c78b86eab408aca1a28299e02b0daa322dd64a94d5b9cd531ab8603
[ "arxiv", "semantic_scholar" ]
PPAI: Enabling Personalized LLM Agent Interoperability for Collaborative Edge Intelligence
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...
[ "Zile Wang", "Qianli Liu", "Kaibin Guo", "Haodong Wang", "Jian Lin", "Zicong Hong", "Song Guo" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.18067
https://arxiv.org/pdf/2605.18067v1
2605.18067
null
0
0
false
null
null
0.35
1355e0aba3dbd9f6a4388cb534677a790f575edfe2e4b6a1d22ab2e486bbdf18
[ "arxiv", "semantic_scholar" ]
Position: A Three-Layer Probabilistic Assume-Guarantee Architecture Is Structurally Required for Safe LLM Agent Deployment
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 constit...
[ "S. Bensalem", "Y. Dong", "M. Franzle", "X. Huang", "J. Kroger", "D. Nickovic", "A. Nouri", "R. Roy", "C. Wu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.18672
https://arxiv.org/pdf/2605.18672v1
2605.18672
null
0
0
false
null
null
0.35
57b034743e4188b6f0183e6e991d6605f88fc1096e1bc3a2f5e36aaf950cc145
[ "arxiv", "semantic_scholar" ]
Time to REFLECT: Can We Trust LLM Judges for Evidence-based Research Agents?
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, ...
[ "Leyao Wang", "Yanan He", "Peng Chen", "Asaf Yehudai", "Yixin Liu", "Rex Ying", "Michal Shmueli-Scheuer", "Arman Cohan" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-18T00:00:00
https://arxiv.org/abs/2605.19196
https://arxiv.org/pdf/2605.19196v1
2605.19196
null
1
0
false
null
null
0.35
7294407b9c92bd8bcd0c1108288b8c212007d67d15609d091b1578a5f23f23e4
[ "arxiv", "semantic_scholar" ]
VerifyMAS: Hypothesis Verification for Failure Attribution in LLM Multi-Agent Systems
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, r...
[ "Hezhe Qiao", "Hanghang Tong", "Ee-Peng Lim", "Bing Liu", "Guansong Pang" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17467
https://arxiv.org/pdf/2605.17467v1
2605.17467
null
0
0
true
null
null
0.65
7b8e0ba5a4867f9ca24596343dff2174443b65b1115466a5727eca83762168de
[ "arxiv", "semantic_scholar" ]
MetaCogAgent: A Metacognitive Multi-Agent LLM Framework with Self-Aware Task Delegation
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 tas...
[ "Chenyu Wang", "Yang Shu" ]
[ "cs.AI", "cs.MA" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17292
https://arxiv.org/pdf/2605.17292v1
2605.17292
null
0
0
false
null
null
0.35
6656e88d69f28a5156dbb8e96e7be5ae3cec2d68529b3cb752ede5041fb4ffe4
[ "arxiv", "semantic_scholar" ]
Trust No Tool: Evaluating and Defending LLM Agents under Untrusted Tool Feedback
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...
[ "Lecheng Yan", "Ruizhe Li", "Xicheng Han", "Wenxi Li", "Binwu Wang", "Longyue Wang", "Chenyang Lyu", "Guanhua Chen" ]
[ "cs.CR", "cs.CL" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17453
https://arxiv.org/pdf/2605.17453v1
2605.17453
null
0
0
false
null
null
0.35
8e529ead37b8ebc3c3d4039f1b84e721f6bac0019bfd481a14fb32e07d35f8fc
[ "arxiv", "semantic_scholar" ]
Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains
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 substa...
[ "Taolin Zhang", "Zijie Zhou", "Jiuheng Wan", "Tingyuan Hu", "Chengyu Wang", "Xiaofeng He", "Richang Hong" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17359
https://arxiv.org/pdf/2605.17359v1
2605.17359
null
0
0
false
null
null
0.35
5549d89545f6aa150768a9f0cc4f081c5451a3c0d78117c094c2927309e753e2
[ "arxiv", "semantic_scholar" ]
Agent Bazaar: Enabling Economic Alignment in Multi-Agent Marketplaces
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 B...
[ "Seth Karten", "Cameron Crow", "Chi Jin" ]
[ "cs.LG", "cs.MA" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17698
https://arxiv.org/pdf/2605.17698v1
2605.17698
null
0
0
false
null
null
0.35
86030df69f07adcaa0f8cc28c3d3a5dbcde90b9e3e495f30e2c4172a289c4fe9
[ "arxiv", "semantic_scholar" ]
Taming "Zombie'' Agents: A Markov State-Aware Framework for Resilient Multi-Agent Evolution
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...
[ "Taolin Zhang", "Pukun Zhao", "Qizhou Chen", "Jiuheng Wan", "Chen Chen", "Xiaofeng He", "Chengyu Wang", "Richang Hong" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17348
https://arxiv.org/pdf/2605.17348v1
2605.17348
null
0
0
false
null
null
0.35
682cfebab4f97c339cacc339b6f0475bcd7ae0c72acdbfae410a7b4164ce7175
[ "arxiv", "semantic_scholar" ]
RooAgent: An LLM Agent for Root-Based High Energy Physics Analysis
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 ...
[ "Aman Desai" ]
[ "hep-ph" ]
[ "Physics" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17318
https://arxiv.org/pdf/2605.17318v2
2605.17318
null
0
0
true
https://github.com/amanmdesai/RooAgent
null
0.65
42b07c895e4407830b29fbbde66de3c2bbdab59b321ee950ab51fc344358d89d
[ "arxiv", "semantic_scholar" ]
AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering
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 AMA...
[ "Taolin Zhang", "Dongyang Li", "Chen Chen", "Qizhou Chen", "Jiuheng Wan", "Xiaofeng He", "Chengyu Wang", "Richang Hong" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-17T00:00:00
https://arxiv.org/abs/2605.17352
https://arxiv.org/pdf/2605.17352v1
2605.17352
null
0
0
false
null
null
0.35
8c1300afc52492790790c1c2cc4f0302d280a9b1a3f051775269ae35c2ca5c47
[ "arxiv", "semantic_scholar" ]
Multi-Paradigm Agent Interaction in Practice:A Systematic Analysis of Generator-Evaluator, ReAct Loop,and Adversarial Evaluation in the buddyMe Framework
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 (Gener...
[ "Xiaohua Wang", "Chao Han", "Kai Yu", "XiaoLiang Xu", "Liang Wang" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-16T00:00:00
https://arxiv.org/abs/2605.16821
https://arxiv.org/pdf/2605.16821v1
2605.16821
null
0
0
true
null
null
0.65
e17ebd701cb2661353c7939c1412ad26fcd7243c47aee4533682ab79ee390a92
[ "arxiv", "semantic_scholar" ]
S-Bus: Automatic Read-Set Reconstruction for Multi-Agent LLM State Coordination
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 ...
[ "Sajjad Khan" ]
[ "cs.LG", "cs.AI", "cs.DC", "cs.MA" ]
[ "Computer Science" ]
2026-05-16T00:00:00
https://arxiv.org/abs/2605.17076
https://arxiv.org/pdf/2605.17076v2
2605.17076
null
1
0
true
https://github.com/sajjadanwar0/sbus
null
0.65
77880cdb6661616827844a3339e2a112023dc9fe2eec013b5e080da9dc1984fe
[ "arxiv", "semantic_scholar" ]
NeuroMAS: Multi-Agent Systems as Neural Networks with Joint Reinforcement Learning
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 agen...
[ "Haoran Lu", "Luyang Fang", "Wenxuan Zhong", "Ping Ma" ]
[ "cs.AI", "cs.MA", "stat.ME", "stat.ML" ]
[ "Computer Science", "Mathematics" ]
2026-05-16T00:00:00
https://arxiv.org/abs/2605.16757
https://arxiv.org/pdf/2605.16757v1
2605.16757
null
0
0
false
null
null
0.35
9cc54f12e19ad8542c56882b77994e7ca3fddbd740c697f1042153e4fc171235
[ "arxiv", "semantic_scholar" ]
BioXArena: Benchmarking LLM Agents on Multi-Modal Biomedical Machine Learning Tasks
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 learn...
[ "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" ]
[ "cs.CE" ]
[ "Computer Science" ]
2026-05-15T00:00:00
https://arxiv.org/abs/2605.15766
https://arxiv.org/pdf/2605.15766v1
2605.15766
null
0
0
false
null
null
0.35
1acdb65f7a03f600cf2bcb3dee403ba62d003c11e4ac08400f4c507124b1f306
[ "arxiv", "semantic_scholar" ]
Cattle Trade: A Multi-Agent Benchmark for LLM Bluffing, Bidding, and Bargaining
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 mod...
[ "Robert Müller", "Clemens Müller" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14537
https://arxiv.org/pdf/2605.14537v1
2605.14537
null
0
0
false
null
null
0.35
39fc5b80727965d7820af2d4ed4ef7eb7d21beeef498519355275c2dc873662e
[ "arxiv", "semantic_scholar" ]
Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems
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...
[ "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" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14892
https://arxiv.org/pdf/2605.14892v2
2605.14892
null
0
0
false
null
null
0.35
a72fe7030dd79e5175761308fd3ad74a1dcf2de4decccbbda5112fe516d60e48
[ "arxiv", "semantic_scholar" ]
Concurrency without Model Changes: Future-based Asynchronous Function Calling for LLMs
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 e...
[ "Guangyu Feng", "Huanzhi Mao", "Prabal Dutta", "Joseph E. Gonzalez" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.15077
https://arxiv.org/pdf/2605.15077v1
2605.15077
null
1
0
false
null
null
0.35
bd2fc440df137d37625dedc149737ba12ef155263197653bf40604ce2af6999a
[ "arxiv", "semantic_scholar" ]
Making OpenAPI Documentation Agent-Ready: Detecting Documentation and REST Smells with a Multi-Agent LLM System
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 wi...
[ "Rayfran Rocha Lima", "Davi G. Assunção Pinheiro", "Thiago Medeiros de Menezes" ]
[ "cs.SE" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14312
https://arxiv.org/pdf/2605.14312v1
2605.14312
null
0
0
false
null
null
0.35
53f486dd8cac88c931604833d0c7173ddd9f57d2f1763eddf4139a42647cbed0
[ "arxiv", "semantic_scholar" ]
From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents
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 framew...
[ "Md Tahmid Rahman Laskar", "Xue-Yong Fu", "Seyyed Saeed Sarfjoo", "Quinten McNamara", "Jonas Robertson", "Shashi Bhushan TN" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.15104
https://arxiv.org/pdf/2605.15104v2
2605.15104
null
0
0
true
null
null
0.65
2d9e60909ce1e6a4d179789e5bc87074612d4553d2fed2ea6a2da13f1847efdb
[ "arxiv", "semantic_scholar" ]
Latency-Quality Routing for Functionally Equivalent Tools in LLM Agents
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 l...
[ "Kexin Chu", "Dawei Xiang", "Wei Zhang" ]
[ "cs.LG" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14241
https://arxiv.org/pdf/2605.14241v2
2605.14241
null
0
0
false
null
null
0.35
40038dc22a839241eea9fc0e1f4a884c0d4aeacd0d35034e7aaee8b4d644c83d
[ "arxiv", "semantic_scholar" ]
GroupMemBench: Benchmarking LLM Agent Memory in Multi-Party Conversations
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...
[ "Jingbo Yang", "Kwei-Herng Lai", "Xiaowen Wang", "Shiyu Chang", "Yaar Harari", "Evgeniy Gabrilovich" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14498
https://arxiv.org/pdf/2605.14498v2
2605.14498
null
2
0
false
null
null
0.35
a71c0b15a0e23ab9070dbb9610ddc06c438e402963bafafb7422c98ff16cc147
[ "arxiv", "semantic_scholar" ]
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
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 h...
[ "Vidya K Sudarshan", "Anushka Sisodia", "Reshma A Ramachandra", "Sia Batra", "Josephine Chong Leng Leng" ]
[ "cs.AI", "cs.CY" ]
[ "Computer Science" ]
2026-05-14T00:00:00
https://arxiv.org/abs/2605.14266
https://arxiv.org/pdf/2605.14266v1
2605.14266
null
0
0
false
null
null
0.35
70a89ae4fff390538074a3884ab586529ef725c8300b2179cdbef8d8700f97c2
[ "arxiv", "semantic_scholar" ]
Speculative Interaction Agents: Building Real-Time Agents with Asynchronous I/O and Speculative Tool Calling
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 ...
[ "Coleman Hooper", "Minwoo Kang", "Suhong Moon", "Nicholas Lee", "Eric Wen", "John Wawrzynek", "Michael W. Mahoney", "Yakun Sophia Shao", "Amir Gholami", "Kurt Keutzer" ]
[ "cs.LG" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13360
https://arxiv.org/pdf/2605.13360v2
2605.13360
null
0
0
false
null
null
0.35
271a062541240609a96c3c31a77d6c3f8cfcf8d59cc7376dabf2253de902e3a8
[ "arxiv", "semantic_scholar" ]
Hierarchical Attacks for Multi-Modal Multi-Agent Reasoning
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, e...
[ "Hao Zhou", "Tiru Wu", "Yan Jiang", "Wanqi Zhou", "Junxing Hu", "Ai Han" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13213
https://arxiv.org/pdf/2605.13213v1
2605.13213
null
0
0
false
null
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
0.55
576236b235ac67e87fabfc03c3974b67e62badf3a38a122e9efabec2b73d9a51
[ "arxiv", "semantic_scholar" ]
MARLIN: Multi-Agent Game-Theoretic Reinforcement Learning for Sustainable LLM Inference in Cloud Datacenters
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 inf...
[ "H. Moore", "S. Qi", "D. Milojicic", "C. Bash", "S. Pasricha" ]
[ "cs.DC", "cs.LG" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13496
https://arxiv.org/pdf/2605.13496v1
2605.13496
null
0
0
false
null
null
0.35
deacc5441d8c0e0106e68ae992cca8c4728f848fa3f51aae33bce8688394458c
[ "arxiv", "semantic_scholar" ]
Reinforced Collaboration in Multi-Agent Flow Networks
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 propaga...
[ "Zheng Wang", "Yuang Liu", "Yangkai Ding" ]
[ "cs.LG" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.12943
https://arxiv.org/pdf/2605.12943v1
2605.12943
null
0
0
true
https://github.com/openJiuwen-ai/agent-store/tree/main/community/mango
null
0.65
3039edee21038e7a4199d7dc0a8ec7d4566194277171a50ce555bfc9dbef85a3
[ "arxiv", "semantic_scholar" ]
Reinforcement Learning for Tool-Calling Agents in Fast Healthcare Interoperability Resources (FHIR)
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...
[ "Marius S. Knorr", "Robert Müller", "Jan P. Bremer", "Nils Schweingruber" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.14126
https://arxiv.org/pdf/2605.14126v1
2605.14126
null
0
0
false
null
null
0.35
4627974c6c6f0f12ce52a9f283a15c50c8ad8544bbc75ca04aec56edc2b0b8a0
[ "arxiv", "semantic_scholar" ]
Collaborating in Multi-Armed Bandits with Strategic Agents
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 agent...
[ "Idan Barnea", "Ofir Schlisselberg", "Yishay Mansour" ]
[ "cs.LG" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13145
https://arxiv.org/pdf/2605.13145v1
2605.13145
null
0
0
false
null
null
0.35
8b6ed9395f084e06396a9ee82d752cc62472b282b2acd83cf88e624d6c210de1
[ "arxiv", "semantic_scholar" ]
Submodular Multi-Agent Policy Learning for Online Distributed Task Allocation in Open Multi-Agent Systems
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 in...
[ "Jing Liu", "Yangyang Yang", "Luca Ballotta", "Fangfei Li", "Yang Tang", "Ruggero Carli" ]
[ "eess.SY" ]
[ "Engineering", "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13269
https://arxiv.org/pdf/2605.13269v1
2605.13269
null
0
0
false
null
null
0.35
4022879579a0f5f894d699fd213321925bc0a394816d75e172e552618923b4ee
[ "arxiv", "semantic_scholar" ]
TERMS-Bench: Diagnosing LLM Negotiation Agents Beyond Deal Rate
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...
[ "Erica Zhang", "Fangzhao Zhang", "Aneesh Pappu", "Batu El", "Jose Blanchet", "Susan Athey", "Jiashuo Liu", "James Zou" ]
[ "cs.GT", "cs.AI" ]
[ "Computer Science" ]
2026-05-13T00:00:00
https://arxiv.org/abs/2605.13909
https://arxiv.org/pdf/2605.13909v1
2605.13909
null
0
0
false
null
null
0.35
82eb7e6fcce27ec07b72905f149571ae672b0d49474d3324e7e217cbc8895e0c
[ "arxiv", "semantic_scholar" ]
LLM-X: A Scalable Negotiation-Oriented Exchange for Communication Among Personal LLM Agents
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...
[ "Giuliano Lorenzoni", "Paulo Alencar", "Donald Cowan" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11376
https://arxiv.org/pdf/2605.11376v1
2605.11376
10.1145/3786167.3788429
0
0
false
null
null
0.35
cbac687f7ad0f6f12cae28b4df6620b643eb2f3af4b01794c9175a3ab6f426fd
[ "arxiv", "semantic_scholar" ]
Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations
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, ...
[ "Junjue Wang", "Weihao Xuan", "Heli Qi", "Pengyu Dai", "Kunyi Liu", "Hongruixuan Chen", "Zhuo Zheng", "Junshi Xia", "Stefano Ermon", "Naoto Yokoya" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11633
https://arxiv.org/pdf/2605.11633v1
2605.11633
null
0
0
false
null
null
0.35
d2e7284e3e15c93303ad4effeb4c0695ce0da65b310f2aab23d9a5d2ecedd63e
[ "arxiv", "semantic_scholar" ]
FlowSteer: Prompt-Only Workflow Steering Exposes Planning-Time Vulnerabilities in Multi-Agent LLM Systems
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 sha...
[ "Fanxiao Li", "Jiaying Wu", "Tingchao Fu", "Natasha Jaques", "Wei Zhou", "Min-Yen Kan" ]
[ "cs.CR" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11514
https://arxiv.org/pdf/2605.11514v1
2605.11514
null
2
0
false
null
null
0.35
02d238ce4bcdc8b7ec6b67af51f5eb81187d1e2214fcf32741eee6e89cef6cda
[ "arxiv", "semantic_scholar" ]
Predictive Maps of Multi-Agent Reasoning: A Successor-Representation Spectrum for LLM Communication Topologies
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 ev...
[ "Ethan Parks", "Dalal Alharthi" ]
[ "cs.MA", "cs.AI", "cs.LG", "cs.SI", "math.SP" ]
[ "Computer Science", "Mathematics" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11453
https://arxiv.org/pdf/2605.11453v2
2605.11453
null
0
0
false
null
null
0.35
cfde58fb27f5fd2fc739c3796209fca567a5d181c0beebdec8183ac47cb52025
[ "arxiv", "semantic_scholar" ]
Coordinated Diffusion: Generating Multi-Agent Behavior Without Multi-Agent Demonstrations
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 exponentia...
[ "Lasse Peters", "Laura Ferranti", "Andrea Bajcsy", "Javier Alonso-Mora" ]
[ "cs.RO" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11485
https://arxiv.org/pdf/2605.11485v2
2605.11485
null
0
0
false
null
null
0.35
fd223853382fed2eb9a07980f3b2f3e60c0c80bf4ac153dd2da3f12e23e8eebe
[ "arxiv", "semantic_scholar" ]
No Action Without a NOD: A Heterogeneous Multi-Agent Architecture for Reliable Service Agents
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 d...
[ "Zixu Yang", "Hang Zheng", "Nan Jiang", "Zhiyang Tang", "Situo Zhang", "Xiaobao Wu", "Lu Chen", "Kai Yu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.12240
https://arxiv.org/pdf/2605.12240v1
2605.12240
null
0
0
false
null
null
0.35
577e461f8790899ce88458cd0ef2d2ebbb0a20b72d39614dfe711cb8da73d245
[ "arxiv", "semantic_scholar" ]
CTFusion: A CTF-based Benchmark for LLM Agent Evaluation
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 ...
[ "Dongjun Lee", "Ga-eun Bae", "Insu Yun" ]
[ "cs.LG", "cs.CR" ]
[ "Computer Science" ]
2026-05-12T00:00:00
https://arxiv.org/abs/2605.11504
https://arxiv.org/pdf/2605.11504v1
2605.11504
null
1
0
true
null
null
0.65
01ed7cb89353aab6eae6be4038ee642394bef0132974dc076a28c6841a52a4d8
[ "arxiv", "semantic_scholar" ]
AnomalyClaw: A Universal Visual Anomaly Detection Agent via Tool-Grounded Refutation
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 tr...
[ "Xi Jiang", "Yinjie Zhao", "Zesheng Yang", "Feng Zheng" ]
[ "cs.CV", "cs.AI" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10397
https://arxiv.org/pdf/2605.10397v1
2605.10397
null
0
0
true
https://github.com/jam-cc/AnomalyClaw
null
0.65
deb049020bdf018c6efc575012cff59013a0f8a9b412f684cb967f0bcd9ab7f8
[ "arxiv", "semantic_scholar" ]
OLIVIA: Online Learning via Inference-time Action Adaptation for Decision Making in LLM ReAct Agents
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...
[ "Sheldon Yu", "Junda Wu", "Xintong Li", "Nikki Lijing Kuang", "Sizhe Zhou", "Tong Yu", "Jiawei Han", "Jingbo Shang", "Julian McAuley" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.11169
https://arxiv.org/pdf/2605.11169v1
2605.11169
null
0
0
false
null
null
0.35
b5e7edd8fd8732c151b6a618bb97b8dcc16346b08e9c702a8e16fdf0699ccdb9
[ "arxiv", "semantic_scholar" ]
Safe Multi-Agent Behavior Must Be Maintained, Not Merely Asserted: Constraint Drift in LLM-Based Multi-Agent Systems
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 privat...
[ "Tianxiao Li", "Yixing Ma", "Haiquan Wen", "Zhenglin Huang", "Qianyu Zhou", "Zeyu Fu", "Guangliang Cheng" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10481
https://arxiv.org/pdf/2605.10481v1
2605.10481
null
0
0
false
null
null
0.35
24b643bc218338eb499555efe3b153632be2498cae7379795f5c2dd7c9931ec7
[ "arxiv", "semantic_scholar" ]
RecoAtlas: From Semantic Plausibility to Set-Level Utility in LLM Recommendation Agents
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 A...
[ "Imad Aouali", "Flavian Vasile", "Otmane Sakhi", "Alexandre Gilotte", "Benjamin Heymann" ]
[ "cs.IR", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.18805
https://arxiv.org/pdf/2605.18805v1
2605.18805
null
0
0
false
null
null
0.35
81552708381cc2572b9140dc5244c9b85e6ea83c153011b5c9d60bd9bcb0c4db
[ "arxiv", "semantic_scholar" ]
AgentRx: A Benchmark Study of LLM Agents for Multimodal Clinical Prediction Tasks
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 vario...
[ "Baraa Al Jorf", "Farah E. Shamout" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10286
https://arxiv.org/pdf/2605.10286v1
2605.10286
null
0
0
false
null
null
0.35
9343e5bf88ac1848c0470ca375df1914687790e5679e483674f76bec05f3803b
[ "arxiv", "semantic_scholar" ]
Collective Alignment in LLM Multi-Agent Systems: Disentangling Bias from Cooperation via Statistical Physics
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-agen...
[ "Cristiano De Nobili" ]
[ "cond-mat.stat-mech", "cs.CL", "cs.MA", "physics.soc-ph" ]
[ "Physics", "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10528
https://arxiv.org/pdf/2605.10528v1
2605.10528
null
0
0
false
null
null
0.35
c4ce8d2d143644c4667df938fc9028c48848551d88c883ec6e725a3e80de4ca7
[ "arxiv", "semantic_scholar" ]
Control Charts for Multi-agent Systems
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...
[ "Hayden Helm", "Carey Priebe", "Brandon Duderstadt" ]
[ "cs.MA", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.11135
https://arxiv.org/pdf/2605.11135v1
2605.11135
null
1
0
false
null
null
0.35
99e1d26655b759cda4d62b2e3bdcf6efddae87338230ac3c3c03591eb0b3900b
[ "arxiv", "semantic_scholar" ]
TRACER: Verifiable Generative Provenance for Multimodal Tool-Using Agents
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 supp...
[ "Bihui Yu", "Caijun Jia", "Jing Chi", "Xiaohan Liu", "Yining Wang", "He Bai", "Yuchen Liu", "Jingxuan Wei", "Junnan Zhu" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.09934
https://arxiv.org/pdf/2605.09934v1
2605.09934
null
0
0
false
null
null
0.35
a2a46ddccd7f4791969aa3c599d76ed541b12581c6cb1048267bba54b9028dfb
[ "arxiv", "semantic_scholar" ]
PRISM: Generation-Time Detection and Mitigation of Secret Leakage in Multi-Agent LLM Pipelines
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 b...
[ "Riya Tapwal", "Abhishek Kumar", "Carsten Maple" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10614
https://arxiv.org/pdf/2605.10614v1
2605.10614
null
0
0
false
null
null
0.35
3f6dac0296b9761823c1e70d7524c0e9ab0cd55d4d00f7691c95cce392ab5f3c
[ "arxiv", "semantic_scholar" ]
Agent-First Tool API: A Semantic Interface Paradigm for Enterprise AI Agent Systems
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, re...
[ "Kai Pan" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10555
https://arxiv.org/pdf/2605.10555v1
2605.10555
null
0
0
false
null
null
0.35
c6895095b32772656457af695ee443d710d744382abfe87788e766b26b9dd727
[ "arxiv", "semantic_scholar" ]
LITMUS: Benchmarking Behavioral Jailbreaks of LLM Agents in Real OS Environments
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 evaluat...
[ "Chiyu Zhang", "Huiqin Yang", "Bendong Jiang", "Xiaolei Zhang", "Yiran Zhao", "Ruyi Chen", "Lu Zhou", "Xiaogang Xu", "Jiafei Wu", "Liming Fang", "Zhe Liu" ]
[ "cs.CR", "cs.CL" ]
[ "Computer Science" ]
2026-05-11T00:00:00
https://arxiv.org/abs/2605.10779
https://arxiv.org/pdf/2605.10779v1
2605.10779
null
0
0
false
null
null
0.35
dae0121a44690a3903617e80bf7b7cc778c6e5eb1fc8154a21a69ceadd22e36d
[ "arxiv", "semantic_scholar" ]
LLM Agents Already Know When to Call Tools -- Even Without Reasoning
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-ho...
[ "Chung-En Sun", "Linbo Liu", "Ge Yan", "Zimo Wang", "Tsui-Wei Weng" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-10T00:00:00
https://arxiv.org/abs/2605.09252
https://arxiv.org/pdf/2605.09252v2
2605.09252
null
5
0
true
https://github.com/Trustworthy-ML-Lab/when2tool
null
0.65
1c6654509e569f2015d971b216cac222f9327a5b11447305594ab32a57a15e4b
[ "arxiv", "semantic_scholar" ]
SkillMAS: Skill Co-Evolution with LLM-based Multi-Agent System
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 ...
[ "Shuai Pan", "Yixiang Liu", "Jiaye Gao", "Te Gao", "Weiwen Liu", "Jianghao Lin", "Zhihui Fu", "Jun Wang", "Weinan Zhang", "Yong Yu" ]
[ "cs.MA", "cs.CL" ]
[ "Computer Science" ]
2026-05-10T00:00:00
https://arxiv.org/abs/2605.09341
https://arxiv.org/pdf/2605.09341v2
2605.09341
null
1
0
false
null
null
0.35
39902d887ac8f9b0ac3c3bcb84508fcb788f66b432ba59685c1b8bfda5361d8a
[ "arxiv", "semantic_scholar" ]
TacoMAS: Test-Time Co-Evolution of Topology and Capability in LLM-based Multi-Agent Systems
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 topolo...
[ "Chen Xu", "Yicheng Hu", "Ruizi Wang", "Xinyu Lin", "Wenjie Wang", "Dongrui Liu", "Fuli Feng" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-10T00:00:00
https://arxiv.org/abs/2605.09539
https://arxiv.org/pdf/2605.09539v1
2605.09539
null
0
0
true
https://github.com/chenxu2-gif/TacoMAS-MultiAgent
null
0.65
0fb6e3494c8ffcdb0061b65bc1df6d173a398edd7d9d5d3f9b2216593f027eb4
[ "arxiv", "semantic_scholar" ]
AgentShield: Deception-based Compromise Detection for Tool-using LLM Agents
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...
[ "Yassin H. Rassul", "Tarik A. Rashid" ]
[ "cs.CR", "cs.CL" ]
[ "Computer Science" ]
2026-05-10T00:00:00
https://arxiv.org/abs/2605.11026
https://arxiv.org/pdf/2605.11026v1
2605.11026
null
0
0
true
https://github.com/Yassin-H-Rassul/AgentShield
null
0.65
6d27f0b76e7e77d7e1d16936a5fc5e00982fa4c6476bb536fe6315dac2f1e10a
[ "arxiv", "semantic_scholar" ]
CalBench: Evaluating Coordination-Privacy Trade-offs in Multi-Agent LLMs
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 ben...
[ "Chelsea Zou", "Yiheng Yao", "Selena She", "Noah Goodman", "Robert D. Hawkins" ]
[ "cs.MA", "cs.AI" ]
[ "Computer Science" ]
2026-05-10T00:00:00
https://arxiv.org/abs/2605.09823
https://arxiv.org/pdf/2605.09823v3
2605.09823
null
0
0
false
null
null
0.35
f67bc9d13d94478275248363cc2a72f91d451b8c323d773daf65df6a6f063b97
[ "arxiv", "semantic_scholar" ]
Learning the Preferences of a Learning Agent
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 a...
[ "Karim Abdel Sadek", "Mark Bedaywi", "Rhys Gould", "Stuart Russell" ]
[ "cs.AI", "cs.LG", "cs.MA" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.09217
https://arxiv.org/pdf/2605.09217v1
2605.09217
null
0
0
false
null
null
0.35
2d867ea02e2a5e5d89e973f704c7e3bd7e84622ed27449c19286ea074345b937
[ "arxiv", "semantic_scholar" ]
Robust Multi-Agent LLMs under Byzantine Faults
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 performanc...
[ "Haejoon Lee", "Vincent-Daniel Yun", "Hyeonho Oh", "Dimitra Panagou", "Sai Praneeth Karimireddy" ]
[ "cs.MA", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.09076
https://arxiv.org/pdf/2605.09076v1
2605.09076
null
0
0
false
null
null
0.35
91734bcd64b2c71d71b636b424dd40b3357b8d94aa4855bf2ea7fd5167a9526f
[ "arxiv", "semantic_scholar" ]
Iterative Critique-and-Routing Controller for Multi-Agent Systems with Heterogeneous LLMs
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 dra...
[ "Wenzhi Fang", "Liangqi Yuan", "Guangchen Lan", "Dong-Jun Han", "Christopher G. Brinton" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.08686
https://arxiv.org/pdf/2605.08686v1
2605.08686
null
0
0
false
null
null
0.35
3f49fe069af76a509b67d5e3e2faeeeceae2da2e54f0873735552e56236cc0ee
[ "arxiv", "semantic_scholar" ]
Beyond the All-in-One Agent: Benchmarking Role-Specialized Multi-Agent Collaboration in Enterprise Workflows
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 ex...
[ "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" ]
[ "cs.MA", "cs.LG" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.08761
https://arxiv.org/pdf/2605.08761v1
2605.08761
null
0
0
false
null
null
0.35
088ea5b6b4d83fb5696366f33443d3a40347cc377eddd82e22f87f1e74998962
[ "arxiv", "semantic_scholar" ]
EvoMAS: Learning Execution-Time Workflows for Multi-Agent Systems
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...
[ "Chengdong Xu", "Kaiqiang Ke", "Ziheng Liu", "Jiaqi Wei", "Zibo Shao", "Weile Guo", "Chao Yu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.08769
https://arxiv.org/pdf/2605.08769v1
2605.08769
null
0
0
false
null
null
0.35
e001b789103572eaa0f0290e444163e6bd5218d0c3cedb556103089c05273d29
[ "arxiv", "semantic_scholar" ]
AgentPSO: Evolving Agent Reasoning Skill via Multi-agent Particle Swarm Optimization
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 bias...
[ "Hyunmin Hwang", "Jaemin Kim", "Choonghan Kim", "Hangeol Chang", "Jong Chul Ye" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-09T00:00:00
https://arxiv.org/abs/2605.08704
https://arxiv.org/pdf/2605.08704v1
2605.08704
null
0
0
true
https://github.com/HYUNMIN-HWANG/AgentPSO/
null
0.65
858b20f7ca653ee08ed9fff712e2b3a636fd60be41f2bf0ad8ec274c3ed23958
[ "arxiv", "semantic_scholar" ]
Insider Attacks in Multi-Agent LLM Consensus Systems
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 mu...
[ "Xiaolin Sun", "Zixuan Liu", "Yibin Hu", "Zizhan Zheng" ]
[ "cs.MA", "cs.AI" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.08268
https://arxiv.org/pdf/2605.08268v1
2605.08268
null
0
0
false
null
null
0.35
c27c914eeff902f22ef61407be43d6180b29dce257679415cee2d0ac2ae1c0f5
[ "arxiv", "semantic_scholar" ]
OrchJail: Jailbreaking Tool-Calling Text-to-Image Agents by Orchestration-Guided Fuzzing
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, makin...
[ "Jianming Chen", "Yawen Wang", "Junjie Wang", "Zhe Liu", "Qing Wang", "Fanjiang Xu" ]
[ "cs.MA", "cs.AI", "cs.CR" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.07414
https://arxiv.org/pdf/2605.07414v1
2605.07414
null
0
0
false
null
ICML 2026
0.55
c9b59a8a5e974825379d58bb4cc14231554f7c8bbcd2c535e8e0ac0e59e59a4a
[ "arxiv", "semantic_scholar" ]
Do Agents Need to Plan Step-by-Step? Rethinking Planning Horizon in Data-Centric Tool Calling
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) singl...
[ "Naoki Otani", "Nikita Bhutani", "Hannah Kim", "Dan Zhang", "Estevam Hruschka" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.08477
https://arxiv.org/pdf/2605.08477v1
2605.08477
10.1145/3786335.3813129
0
0
false
null
null
0.35
f8e3d7957d1cc653d7faef7523626ce06ebd6f84b596294792501b978e5cc843
[ "arxiv", "semantic_scholar" ]
Switchcraft: AI Model Router for Agentic Tool Calling
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 knowled...
[ "Sharad Agarwal", "Pooria Namyar", "Alec Wolman", "Rahul Ambavat", "Ankur Gupta", "Qizheng Zhang" ]
[ "cs.AI", "cs.MA" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.07112
https://arxiv.org/pdf/2605.07112v1
2605.07112
null
0
0
false
null
null
0.35
aedab8da480a11ae5a9d5537d8e589fa4220ed038a555eda35a4797cfd185d80
[ "arxiv", "semantic_scholar" ]
When Child Inherits: Modeling and Exploiting Subagent Spawn in Multi-Agent Networks
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-ag...
[ "Ziwen Cai", "Yihe Zhang", "Xiali Hei" ]
[ "cs.CR", "cs.AI" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.08460
https://arxiv.org/pdf/2605.08460v1
2605.08460
null
0
0
false
null
null
0.35
1f9a6a078a41bec8f3ac17d52c99ef2577b0a6acfb8c72570dbba35cbcd7ffb4
[ "arxiv", "semantic_scholar" ]
MASPrism: Lightweight Failure Attribution for Multi-Agent Systems Using Prefill-Stage Signals
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 ...
[ "Yang Liu", "Hongjiang Feng", "Junsong Pu", "Zhuangbin Chen" ]
[ "cs.SE" ]
[ "Computer Science" ]
2026-05-08T00:00:00
https://arxiv.org/abs/2605.07509
https://arxiv.org/pdf/2605.07509v2
2605.07509
null
1
0
false
null
null
0.35
af66b865ac4131498f97bf55624e1777aa5d1dd13c733f754243a94c5d1178a1
[ "arxiv", "semantic_scholar" ]
Designing Intelligent Enterprise Agents: A Capability-Aligned Multi-Agent Architecture
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 enterpr...
[ "John deVadoss" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.08258
https://arxiv.org/pdf/2605.08258v1
2605.08258
null
0
0
false
null
null
0.35
736f53aa911ba078da69934e4d1441ed69763aac9b91488ced5774b4926827c4
[ "arxiv", "semantic_scholar" ]
Beyond the Black Box: Interpretability of Agentic AI Tool Use
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 obser...
[ "Hariom Tatsat", "Ariye Shater" ]
[ "cs.AI", "cs.MA" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06890
https://arxiv.org/pdf/2605.06890v3
2605.06890
null
0
0
false
null
null
0.35
3c530d338dc537427335918fb544198e21f2a7fa96fe6bea5216fd0a3564a2ed
[ "arxiv", "semantic_scholar" ]
AgenticPrecoding: LLM-Empowered Multi-Agent System for Precoding Optimization
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 exp...
[ "Zijiu Yang", "Zixiang Zhang", "Shunpu Tang", "Qianqian Yang", "Zhiguo Shi" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06443
https://arxiv.org/pdf/2605.06443v1
2605.06443
null
0
0
false
null
null
0.35
f20beb4bb6552c43b6995905216b6aee7054d79a4ff11de1ebe39a9b4d8f79ec
[ "arxiv", "semantic_scholar" ]
MANTRA: Synthesizing SMT-Validated Compliance Benchmarks for Tool-Using LLM Agents
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 ...
[ "Ashwani Anand", "Ivi Chatzi", "Ritam Raha", "Anne-Kathrin Schmuck" ]
[ "cs.CL", "cs.LG", "cs.LO" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06334
https://arxiv.org/pdf/2605.06334v1
2605.06334
null
0
0
false
null
null
0.35
f1a866f378e5eb661586d3ddae2c156b7b17b83f107ecf4eef0ee149db16185f
[ "arxiv", "semantic_scholar" ]
MASPO: Joint Prompt Optimization for LLM-based Multi-Agent Systems
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, prim...
[ "Zhexuan Wang", "Xuebo Liu", "Li Wang", "Zifei Shan", "Yutong Wang", "Zhenxi Song", "Min Zhang" ]
[ "cs.AI", "cs.CL", "cs.LG", "cs.MA" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06623
https://arxiv.org/pdf/2605.06623v1
2605.06623
null
1
0
true
https://github.com/wangzx1219/MASPO
null
0.65
20ac76aeeb3dae4e3307d84a6128e547a81decd1a0684ffd54ef5d654b8d78d5
[ "arxiv", "semantic_scholar" ]
Towards Security-Auditable LLM Agents: A Unified Graph Representation
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...
[ "Chaofan Li", "Lyuye Zhang", "Jintao Zhai", "Siyue Feng", "Xichun Yang", "Huahao Wang", "Shihan Dou", "Yu Ji", "Yutao Hu", "Yueming Wu", "Yang Liu", "Deqing Zou" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06812
https://arxiv.org/pdf/2605.06812v1
2605.06812
null
0
0
false
null
null
0.35
a9f973d14351bc6d65e95c3e0767d59066d0f9168aba8b0daa950076a19cb7dd
[ "arxiv", "semantic_scholar" ]
A Self-Healing Framework for Reliable LLM-Based Autonomous Agents
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-heali...
[ "Cheonsu Jeong", "Younggun Shin" ]
[ "cs.SE", "cs.AI" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.06737
https://arxiv.org/pdf/2605.06737v1
2605.06737
null
1
0
false
null
null
0.35
3d54e3b6ed1ae3d00206a55ee7284e7c5d2647d52142f389377c3f67cb0abaac
[ "arxiv", "semantic_scholar" ]
Active Learning for Communication Structure Optimization in LLM-Based Multi-Agent Systems
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 ...
[ "Huchen Yang", "Xinghao Dong", "Dan Negrut", "Jin-Long Wu" ]
[ "cs.MA", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-05-07T00:00:00
https://arxiv.org/abs/2605.05703
https://arxiv.org/pdf/2605.05703v2
2605.05703
null
0
0
false
null
null
0.35
9a4cde77ec3dd676758cdd0d50b62bbb05d28df120647d91081b91c81773855a
[ "arxiv", "semantic_scholar" ]
GEM: Graph-Enhanced Mixture-of-Experts with ReAct Agents for Dialogue State Tracking
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 g...
[ "Ziqi Zhu", "Adithya Suresh", "Tomal Deb", "Iman Abbasnejad" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2026-05-06T00:00:00
https://arxiv.org/abs/2605.04449
https://arxiv.org/pdf/2605.04449v1
2605.04449
null
0
0
false
null
null
0.35
4ebb27d505e6809ee41dd63af486c8ffe20378ea93840108d8b40380cc5b8272
[ "arxiv", "semantic_scholar" ]
Strat-Reasoner: Reinforcing Strategic Reasoning of LLMs in Multi-Agent Games
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 credi...
[ "Yidong He", "Yutao Lai", "Pengxu Yang", "Jiarui Gan", "Jiexin Wang", "Yi Cai", "Mengchen Zhao" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-06T00:00:00
https://arxiv.org/abs/2605.04906
https://arxiv.org/pdf/2605.04906v2
2605.04906
null
0
0
true
https://github.com/ydhe1012/Strat-Reasoner
null
0.65
5e3eb0748d8569f7c85916537b112d358e08cf73a5c9775c1734ef8a61dca37b
[ "arxiv", "semantic_scholar" ]
SensingAgents: A Multi-Agent Collaborative Framework for Robust IMU Activity Recognition
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...
[ "Naiyu Zheng", "Tianlong Yu", "Haochen Yin", "Xiaoyi Fan", "Xiping Hu", "Zhimeng Yin" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-06T00:00:00
https://arxiv.org/abs/2605.04608
https://arxiv.org/pdf/2605.04608v1
2605.04608
null
0
0
false
null
null
0.35
80e92f9ff480a913b8ca1e13d02e84747febb5f65c93c6f9eb67ecaa6924dd7f
[ "arxiv", "semantic_scholar" ]
Tree-based Credit Assignment for Multi-Agent Memory System
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...
[ "Marina Mao", "Alexandr Liu", "Pengbo Li", "Siheng Li", "Bo Zhou", "Xiang Wang" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-06T00:00:00
https://arxiv.org/abs/2605.04811
https://arxiv.org/pdf/2605.04811v1
2605.04811
null
0
0
false
null
null
0.35
baf4e67c0a656e99c144e57cba9a1a14a1706d0322af1c617759f0bac1d12f2b
[ "arxiv", "semantic_scholar" ]
AgentTrust: Runtime Safety Evaluation and Interception for AI Agent Tool Use
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 benc...
[ "Chenglin Yang" ]
[ "cs.AI", "cs.CR" ]
[ "Computer Science" ]
2026-05-06T00:00:00
https://arxiv.org/abs/2605.04785
https://arxiv.org/pdf/2605.04785v1
2605.04785
null
2
0
false
null
null
0.35
a35135d96838c6029392c3141c801349a6a859a7396aeebdc6452c64bc4f30f0
[ "arxiv", "semantic_scholar" ]
When Context Hurts: The Crossover Effect of Knowledge Transfer on Multi-Agent Design Exploration
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 cove...
[ "Saranyan Vigraham" ]
[ "cs.AI", "cs.SE" ]
[ "Computer Science" ]
2026-05-05T00:00:00
https://arxiv.org/abs/2605.04361
https://arxiv.org/pdf/2605.04361v1
2605.04361
null
0
0
false
null
null
0.35
099465807ae3b25a77d39fd504f693d15c499ab325d8072de4b76328e157a19d
[ "arxiv", "semantic_scholar" ]
Coordination as an Architectural Layer for LLM-Based Multi-Agent Systems
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 mapp...
[ "Maksym Nechepurenko", "Pavel Shuvalov" ]
[ "cs.MA", "cs.LG", "q-fin.TR" ]
[ "Computer Science", "Economics" ]
2026-05-05T00:00:00
https://arxiv.org/abs/2605.03310
https://arxiv.org/pdf/2605.03310v1
2605.03310
null
0
0
false
null
null
0.35
b92ccd965fdd943f633455368b8271e65d49230aa730315160994b7c5733fdae
[ "arxiv", "semantic_scholar" ]
Governed Collaborative Memory as Artificial Selection in LLM-Based Multi-Agent Systems
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...
[ "Diego F. Cuadros", "Abdoul-Aziz Maiga", "Helen Meskhidze", "Andre Curtis-Trudel" ]
[ "cs.MA" ]
[ "Computer Science" ]
2026-05-05T00:00:00
https://arxiv.org/abs/2605.04264
https://arxiv.org/pdf/2605.04264v1
2605.04264
null
0
0
false
null
null
0.35
4d380f48858784ab31c76c494d0a51e682fc35a9057f9f41548e25aefd7ba605
[ "arxiv", "semantic_scholar" ]
From Intent to Execution: Composing Agentic Workflows with Agent Recommendation
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. T...
[ "Kishan Athrey", "Ramin Pishehvar", "Brian Riordan", "Mahesh Viswanathan" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-05-05T00:00:00
https://arxiv.org/abs/2605.03986
https://arxiv.org/pdf/2605.03986v1
2605.03986
null
0
0
false
null
null
0.35
99514b0799b7951081a25cf5b5fc1cbd70288ae0c125f59cf5b933ed4c43ae32
[ "arxiv", "semantic_scholar" ]
SkCC: Portable and Secure Skill Compilation for Cross-Framework LLM Agents
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-agno...
[ "Yipeng Ouyang", "Yi Xiao", "Yuhao Gu", "Xianwei Zhang" ]
[ "cs.CR", "cs.AI" ]
[ "Computer Science" ]
2026-05-05T00:00:00
https://arxiv.org/abs/2605.03353
https://arxiv.org/pdf/2605.03353v4
2605.03353
null
1
0
true
https://github.com/Nexa-Language/Skill-Compiler/
null
0.65