Duplicate from Reja1/jee-neet-benchmark

Co-authored-by: Md Rejaullah <Reja1@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,3 @@
+images/** filter=lfs diff=lfs merge=lfs -text
+images/**/*.png filter=lfs diff=lfs merge=lfs -text
+*.tar.gz filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,89 @@
+# Python virtual environment
+hf-env/
+venv/
+env/
+*.venv
+
+# Python cache files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+# Usually these files are written by a python script from a template
+# before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IDE / Editor specific files
+.idea/
+.vscode/
+*.sublime-project
+*.sublime-workspace
+
+# OS generated files
+.DS_Store
+Thumbs.db
+
+# Results (optional - uncomment if you don't want to track results in Git)
+# results/
+
+# Configuration files containing secrets (uncomment if you use these names)
+# configs/openrouter_config.yaml
+.env*
+.env
+
+# Logs
+logs/
+*.log
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+pnpm-debug.log*
+lerna-debug.log*
+
+# Other
+*.swp
+*~

CLAUDE.md

@@ -0,0 +1,87 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+A benchmark for evaluating vision-capable LLMs on Indian competitive exam questions (JEE Main, JEE Advanced, NEET). Questions are images sent to models via the OpenRouter API; responses are parsed from `<answer>...</answer>` tags and scored using exam-specific marking schemes.
+
+## Running the Benchmark
+
+```bash
+# Setup
+uv sync
+echo "OPENROUTER_API_KEY=your_key" > .env
+
+# Must run from project root (paths are resolved relative to cwd)
+uv run python src/benchmark_runner.py --model "google/gemini-2.5-pro-preview-03-25" --exam_name JEE_ADVANCED --exam_year 2025
+
+# Filter by question IDs
+uv run python src/benchmark_runner.py --model "openai/o3" --question_ids "N24T3001,N24T3002"
+```
+
+CLI args: `--model` (required), `--exam_name` (all/NEET/JEE_ADVANCED/JEE_MAIN), `--exam_year` (all/2024/2025), `--question_ids`, `--output_dir`, `--config`, `--resume`, `--temperature` (override config), `--num_runs` (default 1, use 3+ for variance analysis).
+
+### Analysis Scripts
+
+```bash
+# Generate cross-model leaderboard from all results
+uv run python scripts/generate_leaderboard.py
+
+# Aggregate multiple runs of the same model for variance analysis
+uv run python scripts/aggregate_runs.py --pattern "openai_o3_JEE_ADVANCED_2025"
+```
+
+## Testing
+
+```bash
+# Run the full pytest suite (67 tests)
+uv run python -m pytest tests/ -v
+
+# Run individual module self-tests
+uv run python src/utils.py        # answer parsing logic
+uv run python src/evaluation.py   # scoring logic
+uv run python src/llm_interface.py  # API calls (requires .env and network)
+```
+
+## Architecture
+
+```
+benchmark_runner.py  ─── orchestrator / entry point
+  ├── loads config from configs/benchmark_config.yaml
+  ├── loads dataset directly from metadata.jsonl (JSONL → HuggingFace Dataset)
+  │     ├── metadata.jsonl (question metadata, 578 questions)
+  │     └── images/ (question PNGs, stored in Git LFS)
+  ├── calls llm_interface.py for each question
+  │     ├── prompts.py (prompt templates)
+  │     └── utils.py (parse_llm_answer extracts from <answer> tags)
+  ├── scores via evaluation.py (exam-specific marking schemes)
+  └── writes results incrementally to results/{model}_{exam}_{year}_{timestamp}/
+        ├── predictions.jsonl  (raw API responses)
+        ├── summary.jsonl      (scored per-question results)
+        └── summary.md         (human-readable report)
+```
+
+### Key data flow
+
+1. Dataset loaded directly from `metadata.jsonl` into a HuggingFace `Dataset` object, filtered by exam/year
+2. Each question image is base64-encoded and sent to OpenRouter with a structured prompt
+3. If the response can't be parsed, a re-prompt is sent (text-only, with the bad response)
+4. If the API call fails, the question is queued for retry (up to 3 attempts, exponential backoff via `tenacity`)
+5. Answers are parsed from `<answer>...</answer>` tags by `utils.parse_llm_answer()`
+6. `evaluation.py` scores using JEE/NEET marking schemes (partial credit for MCQ_MULTIPLE_CORRECT in JEE Advanced)
+
+## Answer Format Conventions
+
+- `MCQ_SINGLE_CORRECT`: `<answer>A</answer>` → `["A"]`
+- `MCQ_MULTIPLE_CORRECT`: `<answer>A,C</answer>` → `["A", "C"]` (sorted, deduplicated)
+- `INTEGER`: `<answer>42</answer>` → `["42"]`
+- `SKIP`: `<answer>SKIP</answer>` → no penalty
+
+## Important Notes
+
+- **Git LFS**: Images and `metadata.jsonl` are in LFS. Run `git lfs pull` after cloning.
+- **Working directory**: Scripts must be run from project root — config, data, and image paths are resolved relative to cwd.
+- **Python 3.10+**: Uses union type syntax (`list[str] | str | None`).
+- **Models**: Configured in `configs/benchmark_config.yaml` under `openrouter_models`. All must support vision input.
+- **Result directory naming**: `results/{provider}_{model}_{exam}_{year}_{YYYYMMDD_HHMMSS}/` (slashes in model IDs replaced with underscores).

README.md

@@ -0,0 +1,373 @@
+---
+license: mit
+language:
+  - en
+task_categories:
+  - visual-question-answering
+  - image-text-to-text
+  - question-answering
+pretty_name: Indian Competitive Exams (JEE/NEET) LLM Benchmark
+size_categories:
+  - n<1K
+tags:
+  - education
+  - science
+  - india
+  - competitive-exams
+  - llm-benchmark
+configs:
+  - config_name: default
+    data_files:
+      - split: test
+        path: "images/**"
+    drop_labels: true
+dataset_info:
+  features:
+    - name: image
+      dtype: image
+    - name: question_id
+      dtype: string
+    - name: exam_name
+      dtype: string
+    - name: exam_year
+      dtype: int32
+    - name: subject
+      dtype: string
+    - name: question_type
+      dtype: string
+    - name: correct_answer
+      dtype: string
+    - name: paper_id
+      dtype: int64
+  splits:
+    - name: test
+      num_examples: 578
+---
+# JEE/NEET LLM Benchmark Dataset
+
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+
+## Dataset Description
+
+This repository contains a benchmark dataset designed for evaluating the capabilities of Large Language Models (LLMs) on questions from major Indian competitive examinations:
+*   **JEE (Main & Advanced):** Joint Entrance Examination for engineering.
+*   **NEET:** National Eligibility cum Entrance Test for medical fields.
+
+The questions are presented in image format (`.png`) as they appear in the original papers. The dataset includes metadata linking each image to its corresponding exam details (name, year, subject, question type), and correct answer(s). The benchmark framework supports various question types including Single Correct MCQs, Multiple Correct MCQs (with partial marking for JEE Advanced), and Integer type questions.
+
+**Current Data:**
+*   **NEET 2024** (Code T3): 200 questions across Physics, Chemistry, Botany, and Zoology
+*   **NEET 2025** (Code 45): 180 questions across Physics, Chemistry, Botany, and Zoology  
+*   **JEE Advanced 2024** (Paper 1 & 2): 102 questions across Physics, Chemistry, and Mathematics
+*   **JEE Advanced 2025** (Paper 1 & 2): 96 questions across Physics, Chemistry, and Mathematics
+*   **Total:** 578 questions with comprehensive metadata
+
+## Key Features
+
+*   **🖼️ Multimodal Reasoning:** Uses images of questions directly, testing the multimodal reasoning capability of models
+*   **📊 Exam-Specific Scoring:** Implements authentic scoring rules for different exams and question types, including partial marking for JEE Advanced
+*   **🔄 Robust API Handling:** Built-in retry mechanism and re-prompting for failed API calls or parsing errors
+*   **🎯 Flexible Filtering:** Filter by exam name, year, or specific question IDs for targeted evaluation
+*   **📈 Comprehensive Results:** Generates detailed JSON and human-readable Markdown summaries with section-wise breakdowns
+*   **🔧 Easy Configuration:** Simple YAML-based configuration for models and parameters
+
+## Leaderboard
+
+Generate an up-to-date leaderboard from your local results:
+
+```bash
+uv run python scripts/generate_leaderboard.py
+```
+
+See `scripts/generate_leaderboard.py --help` for options including `--min-questions` and `--output`.
+
+## How to Use
+
+### Using `datasets` Library
+
+The dataset is hosted on the Hugging Face Hub and can be loaded directly:
+
+```python
+from datasets import load_dataset
+import json
+
+# Load the evaluation split
+dataset = load_dataset("Reja1/jee-neet-benchmark", split='test')
+
+# Example: Access the first question
+example = dataset[0]
+image = example["image"]
+question_id = example["question_id"]
+subject = example["subject"]
+correct_answers = json.loads(example["correct_answer"])  # Parse JSON string
+
+print(f"Question ID: {question_id}")
+print(f"Subject: {subject}")
+print(f"Correct Answer(s): {correct_answers}")
+# Display the image (requires Pillow)
+# image.show()
+```
+
+### Manual Usage (Benchmark Scripts)
+
+This repository contains scripts to run the benchmark evaluation directly:
+
+1.  **Clone the repository:**
+    ```bash
+    git clone https://huggingface.co/datasets/Reja1/jee-neet-benchmark
+    cd jee-neet-benchmark
+    # Ensure Git LFS is installed and pull large files
+    git lfs pull
+    ```
+
+2.  **Install dependencies:**
+    ```bash
+    uv sync
+    ```
+
+3.  **Configure API Key:**
+    *   Create a file named `.env` in the root directory of the project.
+    *   Add your OpenRouter API key to this file:
+        ```dotenv
+        OPENROUTER_API_KEY=your_actual_openrouter_api_key_here
+        ```
+    *   **Important:** The `.gitignore` file is already configured to prevent committing the `.env` file. Never commit your API keys directly.
+
+4.  **Configure Models:**
+    *   Edit the `configs/benchmark_config.yaml` file.
+    *   Modify the `openrouter_models` list to include the specific model identifiers you want to evaluate:
+        ```yaml
+        openrouter_models:
+          - "google/gemini-2.5-pro-preview-03-25"
+          - "anthropic/claude-sonnet-4"
+          - "openai/o3"
+        ```
+    *   Ensure these models support vision input on OpenRouter.
+    *   You can also adjust other parameters like `max_tokens` and `request_timeout` if needed.
+
+5.  **Run the benchmark:**
+
+    **Basic usage (run all available models on all questions):**
+    ```bash
+    uv run python src/benchmark_runner.py --config configs/benchmark_config.yaml --model "google/gemini-2.5-pro-preview-03-25"
+    ```
+
+    **Filter by exam and year:**
+    ```bash
+    # Run only NEET 2024 questions
+    uv run python src/benchmark_runner.py --config configs/benchmark_config.yaml --model "openai/o3" --exam_name NEET --exam_year 2024
+
+    # Run only JEE Advanced 2025 questions
+    uv run python src/benchmark_runner.py --config configs/benchmark_config.yaml --model "anthropic/claude-sonnet-4" --exam_name JEE_ADVANCED --exam_year 2025
+    ```
+
+    **Run specific questions:**
+    ```bash
+    # Run specific question IDs
+    uv run python src/benchmark_runner.py --config configs/benchmark_config.yaml --model "google/gemini-2.5-pro-preview-03-25" --question_ids "N24T3001,N24T3002,JA24P1M01"
+    ```
+
+    **Resume an interrupted run:**
+    ```bash
+    # Resume from an existing results directory (skips already-completed questions)
+    uv run python src/benchmark_runner.py --model "google/gemini-2.5-pro-preview-03-25" --resume results/google_gemini-2.5-pro-preview-03-25_NEET_2024_20250524_141230
+    ```
+
+    **Custom output directory:**
+    ```bash
+    uv run python src/benchmark_runner.py --config configs/benchmark_config.yaml --model "openai/gpt-4o" --output_dir my_custom_results
+    ```
+
+    **Available options:**
+    - `--exam_name`: Choose from `NEET`, `JEE_MAIN`, `JEE_ADVANCED`, or `all` (default)
+    - `--exam_year`: Choose from available years (`2024`, `2025`, etc.) or `all` (default)
+    - `--question_ids`: Comma-separated list of specific question IDs to evaluate (e.g., "N24T3001,JA24P1M01")
+    - `--resume`: Path to an existing results directory to resume an interrupted run
+
+6.  **Check Results:**
+    *   Results for each model run will be saved in timestamped subdirectories within the `results/` folder.
+    *   Each run's folder (e.g., `results/google_gemini-2.5-pro-preview-03-25_NEET_2024_20250524_141230/`) contains:
+        *   **`predictions.jsonl`**: Raw API responses for each question including:
+            - Raw LLM responses
+            - API call success/failure information
+            - Parse success status and errors
+        *   **`summary.jsonl`**: Per-question scored results including:
+            - Predicted answers and ground truth
+            - Evaluation status and marks awarded
+        *   **`summary.md`**: Human-readable Markdown summary with:
+            - Overall exam scores
+            - Question type breakdown
+            - Section-wise breakdown (by subject)
+            - Detailed statistics on correct/incorrect/skipped questions
+
+## Scoring System
+
+The benchmark implements authentic scoring systems for each exam type:
+
+### NEET Scoring
+- **Single Correct MCQ**: +4 for correct, -1 for incorrect, 0 for skipped/API failure
+
+### JEE Main Scoring
+- **Single Correct MCQ**: +4 for correct, -1 for incorrect, 0 for skipped/API failure
+- **Integer Type**: +4 for correct, 0 for incorrect, 0 for skipped/API failure
+
+### JEE Advanced Scoring
+- **Single Correct MCQ**: +3 for correct, -1 for incorrect, 0 for skipped/API failure
+- **Multiple Correct MCQ**: Partial marking system:
+  - +4 for all correct options selected
+  - +3 for 3 out of 4 correct options (when 4 are correct)
+  - +2 for 2 out of 3+ correct options
+  - +1 for 1 out of 2+ correct options
+  - -2 for any incorrect option selected
+  - 0 for skipped/API failure
+- **Integer Type**: +4 for correct, 0 for incorrect, 0 for skipped/API failure
+
+> **Note:** API failures and parse failures are scored as 0 (no penalty) since they do not represent a deliberate wrong choice.
+
+## Advanced Features
+
+### Retry Mechanism
+- Automatic retry for failed API calls (up to 3 attempts with exponential backoff)
+- Retries on HTTP 429 (rate limit), 500, 502, 503, 504 status codes
+- Separate retry pass for questions that failed initially
+- Comprehensive error tracking and reporting
+
+### Resume Capability
+- Resume interrupted benchmark runs with `--resume <results_dir>`
+- Reads existing `summary.jsonl` to identify completed questions and skips them
+- Appends new results to the same output files
+
+### Re-prompting System
+- If initial response parsing fails, the system automatically re-prompts the model
+- Uses the previous response to ask for properly formatted answers
+- Shows only relevant format examples based on question type (MCQ single, MCQ multiple, or integer)
+
+### Comprehensive Evaluation
+- Tracks multiple metrics: correct answers, partial credit, skipped questions, API failures
+- Section-wise breakdown by subject
+- Color-coded progress indicators in terminal output
+
+## Dataset Structure
+
+*   **`metadata.jsonl`**: Contains metadata for each question image with fields:
+    - `file_name`: Path to the question image (relative to repo root)
+    - `question_id`: Unique identifier (e.g., "N24T3001")
+    - `exam_name`: Exam type ("NEET", "JEE_MAIN", "JEE_ADVANCED")
+    - `exam_year`: Year of the exam (integer)
+    - `subject`: Subject name (e.g., "Physics", "Chemistry", "Mathematics")
+    - `question_type`: Question format ("MCQ_SINGLE_CORRECT", "MCQ_MULTIPLE_CORRECT", "INTEGER")
+    - `correct_answer`: JSON-serialized string of correct answers (e.g., `'["A"]'`, `'["B", "C"]'`, `'["42"]'`)
+
+*   **`images/`**: Contains subdirectories for each exam set:
+    - `images/NEET_2024_T3/`: NEET 2024 question images
+    - `images/NEET_2025_45/`: NEET 2025 question images
+    - `images/JEE_ADVANCED_2024/`: JEE Advanced 2024 question images
+    - `images/JEE_ADVANCED_2025/`: JEE Advanced 2025 question images
+
+
+*   **`src/`**: Python source code for the benchmark system:
+    - `benchmark_runner.py`: Main benchmark execution script
+    - `llm_interface.py`: OpenRouter API interface with retry logic
+    - `evaluation.py`: Scoring and evaluation functions
+    - `prompts.py`: LLM prompts for different question types
+    - `utils.py`: Utility functions for parsing and configuration
+
+*   **`configs/`**: Configuration files:
+    - `benchmark_config.yaml`: Model selection and API parameters
+
+*   **`results/`**: Directory where benchmark results are stored (timestamped subdirectories)
+
+## Data Fields
+
+The dataset contains the following fields (accessible via `datasets`):
+
+*   `image`: The question image (`datasets.Image`)
+*   `question_id`: Unique identifier for the question (string)
+*   `exam_name`: Name of the exam (e.g., "NEET", "JEE_ADVANCED") (string)
+*   `exam_year`: Year of the exam (int)
+*   `subject`: Subject (e.g., "Physics", "Chemistry", "Mathematics") (string)
+*   `question_type`: Type of question (e.g., "MCQ_SINGLE_CORRECT", "INTEGER") (string)
+*   `correct_answer`: JSON-serialized string containing the correct answer(s). Use `json.loads()` to parse.
+    - For MCQs, these are option identifiers (e.g., `'["1"]'`, `'["A"]'`, `'["B", "C"]'`). The LLM should output the identifier as it appears in the question.
+    - For INTEGER type, this is the numerical answer as a string (e.g., `'["42"]'`, `'["12.75"]'`). The LLM should output the number.
+    - For some `MCQ_SINGLE_CORRECT` questions, multiple answers in the list are considered correct if the LLM prediction matches any one of them.
+
+## LLM Answer Format
+
+The LLM is expected to return its answer enclosed in `<answer>` tags. For example:
+- MCQ Single Correct (Option A): `<answer>A</answer>`
+- MCQ Single Correct (Option 2): `<answer>2</answer>`
+- MCQ Multiple Correct (Options B and D): `<answer>B,D</answer>`
+- Integer Answer: `<answer>42</answer>`
+- Decimal Answer: `<answer>12.75</answer>`
+- Skipped Question: `<answer>SKIP</answer>
+
+The system parses these formats. Prompts are designed to guide the LLM accordingly.
+
+## Troubleshooting
+
+### Common Issues
+
+**API Key Issues:**
+- Ensure your `.env` file is in the root directory
+- Verify your OpenRouter API key is valid and has sufficient credits
+- Check that the key has access to vision-capable models
+
+**Model Not Found:**
+- Verify the model identifier exists on OpenRouter
+- Ensure the model supports vision input
+- Check your OpenRouter account has access to the specific model
+
+**Memory Issues:**
+- Reduce `max_tokens` in the config file
+- Process smaller subsets using `--question_ids` filter
+- Use models with smaller context windows
+
+**Parsing Failures:**
+- The system automatically attempts re-prompting for parsing failures
+- Check the raw responses in `predictions.jsonl` to debug prompt issues
+- Consider adjusting prompts in `src/prompts.py` for specific models
+
+## Limitations & Data Contamination
+
+### Contamination Risk
+This benchmark uses questions from publicly administered exams (JEE Advanced and NEET). These questions are widely published online after each exam and may appear in the training data of evaluated models, particularly for older exam years (e.g., 2024). High scores on this benchmark may therefore partially reflect memorization rather than genuine reasoning ability.
+
+To help assess contamination effects:
+- **Compare across years**: Models may score higher on older exams (2024) whose questions had more time to enter training data, compared to newer exams (2025).
+- **Cross-reference with novel benchmarks**: Compare performance on this benchmark with contamination-resistant benchmarks like GPQA or Humanity's Last Exam.
+
+This benchmark is best understood as an evaluation on **publicly available exam questions** rather than a contamination-free assessment of reasoning capability.
+
+### JEE Main Support
+The benchmark framework fully supports JEE Main scoring rules in code, but the current dataset does not include JEE Main questions. JEE Main support is available for users who wish to add their own JEE Main question sets.
+
+### Other Limitations
+- **Single prompt template**: Results may vary with different prompt formulations. The benchmark currently uses one prompt template per question type.
+- **No multi-run variance**: Each model is evaluated once per exam. Results may vary slightly across runs due to non-deterministic model behavior.
+- **Image quality dependence**: Performance may be affected by image resolution, scan quality, or the presence of artifacts in question images.
+- **Language Support**: Currently only supports English questions.
+- **Model Dependencies**: Requires models with vision capabilities available through OpenRouter.
+
+## Citation
+
+If you use this dataset or benchmark code, please cite:
+
+```bibtex
+@misc{rejaullah_2025_jeeneetbenchmark,
+  title={JEE/NEET LLM Benchmark},
+  author={Md Rejaullah},
+  year={2025},
+  howpublished={\url{https://huggingface.co/datasets/Reja1/jee-neet-benchmark}},
+}
+```
+
+## Contact
+
+For questions, suggestions, or collaboration, feel free to reach out:
+
+*   **X (Twitter):** [https://x.com/RejaullahmdMd](https://x.com/RejaullahmdMd)
+
+## License
+
+This dataset and associated code are licensed under the [MIT License](https://opensource.org/licenses/MIT).

configs/benchmark_config.yaml

@@ -0,0 +1,31 @@
+# List of OpenRouter model identifiers to evaluate.
+# Find identifiers at: https://openrouter.ai/docs#models
+# Ensure the models support vision input.
+openrouter_models:
+  - "google/gemini-2.5-pro-preview-03-25"
+  - "anthropic/claude-sonnet-4"
+  - "google/gemini-2.5-flash-preview-05-20:thinking"
+  - "openai/o3"
+  - "openai/gpt-5"
+  - "x-ai/grok-4-fast:free"
+  - "google/gemini-3-pro-preview"
+  # - "google/gemini-pro-vision" # Example - uncomment or add others
+  # - "anthropic/claude-3-opus" # Example - check vision support and access
+  # - "anthropic/claude-3-sonnet"
+  # - "anthropic/claude-3-haiku"
+
+# Path to the metadata JSONL file and base directory for image paths.
+metadata_path: "images/metadata.jsonl"
+images_base_dir: "images"
+
+# Base directory where results for each model will be saved.
+results_base_dir: "results"
+
+# Optional: API request parameters
+# Maximum tokens to generate in the response. Keep it low as we expect only numbers.
+max_tokens: 10000
+# Timeout for each API request in seconds.
+# Set high enough for thinking models (o3, Gemini Flash Thinking) which can take longer.
+request_timeout: 200
+# Sampling temperature. 0 = deterministic (greedy decoding) for reproducibility.
+temperature: 0

pyproject.toml

@@ -0,0 +1,17 @@
+[project]
+name = "jee-neet-benchmark"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.12"
+dependencies = [
+    "datasets==3.5.1",
+    "huggingface-hub==0.30.2",
+    "pillow==11.2.1",
+    "pytest==9.0.2",
+    "python-dotenv==1.1.0",
+    "pyyaml==6.0.2",
+    "requests==2.32.3",
+    "tenacity==9.1.2",
+    "tqdm==4.67.1",
+]

results/anthropic_claude-sonnet-4_JEE_ADVANCED_2025_20250603_220354/summary.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["A"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P05", "marks_awarded": 1, "evaluation_status": "partial_1_of_2_plus", "predicted_answer": ["B"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P1P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P07", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["B", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1P08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1P09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["7"], "ground_truth": [["21", "25"]], "attempt": 1}
+{"question_id": "JA25P1P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P1P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.5"], "ground_truth": ["0.5", "0.75"], "attempt": 1}
+{"question_id": "JA25P1P12", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["72"], "ground_truth": [["75", "79"]], "attempt": 1}
+{"question_id": "JA25P1P13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["27"], "ground_truth": ["72"], "attempt": 1}
+{"question_id": "JA25P1P14", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P15", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P1C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P1C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["100"], "ground_truth": ["100"], "attempt": 1}
+{"question_id": "JA25P1C09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": [["2.2", "2.3"]], "attempt": 1}
+{"question_id": "JA25P1C10", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["1.01"], "ground_truth": [["-7.2", "-7"]], "attempt": 1}
+{"question_id": "JA25P1C11", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["30"], "ground_truth": [["-29.95", "-29.8"], ["29.8", "29.95"]], "attempt": 1}
+{"question_id": "JA25P1C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["280"], "ground_truth": ["280"], "attempt": 1}
+{"question_id": "JA25P1C13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["138"], "ground_truth": ["175"], "attempt": 1}
+{"question_id": "JA25P1C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C15", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C16", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["A"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["16"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M05", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "D"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P1M06", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "C", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105"], "ground_truth": ["105"], "attempt": 1}
+{"question_id": "JA25P1M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.2"], "ground_truth": [["1.15", "1.25"]], "attempt": 1}
+{"question_id": "JA25P1M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["762"], "ground_truth": ["762"], "attempt": 1}
+{"question_id": "JA25P1M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.4"], "ground_truth": [["2.35", "2.45"]], "attempt": 1}
+{"question_id": "JA25P1M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["96"], "ground_truth": ["96"], "attempt": 1}
+{"question_id": "JA25P1M13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["0"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1M14", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["A"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M16", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2P02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["A", "B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2P05", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B", "C", "D"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P06", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["C"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2P07", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P08", "marks_awarded": 2, "evaluation_status": "partial_2_of_3_plus", "predicted_answer": ["B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["0.5"], "ground_truth": [["1.65", "1.67"]], "attempt": 1}
+{"question_id": "JA25P2P10", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["12"], "ground_truth": [["11.7", "11.9"]], "attempt": 1}
+{"question_id": "JA25P2P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.6"], "ground_truth": ["1.6"], "attempt": 1}
+{"question_id": "JA25P2P12", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["3"], "ground_truth": [["2.3", "2.4"]], "attempt": 1}
+{"question_id": "JA25P2P13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["1"], "ground_truth": ["0.2"], "attempt": 1}
+{"question_id": "JA25P2P14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2.4"], "ground_truth": ["1.2"], "attempt": 1}
+{"question_id": "JA25P2P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["170"], "ground_truth": [["167", "171"]], "attempt": 1}
+{"question_id": "JA25P2P16", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["16"], "ground_truth": [["26", "33"]], "attempt": 1}
+{"question_id": "JA25P2C01", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P2C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["C", "D"], "ground_truth": ["C", "D"], "attempt": 1}
+{"question_id": "JA25P2C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P2C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2C08", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P2C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11.0"], "ground_truth": [["10.85", "11.1"]], "attempt": 1}
+{"question_id": "JA25P2C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3.95"], "ground_truth": [["3.85", "4.15"]], "attempt": 1}
+{"question_id": "JA25P2C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["16"], "ground_truth": [["15.5", "16.5"]], "attempt": 1}
+{"question_id": "JA25P2C12", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["67"], "ground_truth": [["4", "4.25"]], "attempt": 1}
+{"question_id": "JA25P2C13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.5"], "ground_truth": [["2.4", "2.55"]], "attempt": 1}
+{"question_id": "JA25P2C14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["106"], "ground_truth": [["105.4", "105.6"]], "attempt": 1}
+{"question_id": "JA25P2C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["7.73"], "ground_truth": [["7.5", "7.8"]], "attempt": 1}
+{"question_id": "JA25P2C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P2M01", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2M03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["3"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2M05", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["D"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M08", "marks_awarded": 1, "evaluation_status": "partial_1_of_2_plus", "predicted_answer": ["B"], "ground_truth": ["B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2M09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": [["0.7", "0.8"]], "attempt": 1}
+{"question_id": "JA25P2M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["6"], "ground_truth": ["6"], "attempt": 1}
+{"question_id": "JA25P2M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.3"], "ground_truth": [["0.27", "0.33"]], "attempt": 1}
+{"question_id": "JA25P2M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["4"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.25"], "ground_truth": [["0.2", "0.3"]], "attempt": 1}
+{"question_id": "JA25P2M15", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["1"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P2M16", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["7"], "ground_truth": ["21"], "attempt": 1}

results/anthropic_claude-sonnet-4_JEE_ADVANCED_2025_20250603_220354/summary.md

@@ -0,0 +1,33 @@
+# Benchmark Results: anthropic/claude-sonnet-4
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2025
+**Timestamp:** 20250603_220354
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 482
+**Total Questions Processed in this Run:** 96
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **151** / **360**
+- **Fully Correct Answers:** 48
+- **Partially Correct Answers:** 3
+- **Incorrectly Answered (Choice Made):** 45
+- **Skipped Questions:** 0
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 96
+
+### Detailed Score Calculation by Question Type
+**Integer (42 questions):** 88 marks
+  *Calculation:* 22 Correct (+4) + 20 Incorrect (0) = 88
+**Mcq Multiple Correct (21 questions):** 28 marks
+  *Calculation:* 10 Correct (+4) + 3 Partial + 8 Incorrect (-2) = 28
+**Mcq Single Correct (33 questions):** 27 marks
+  *Calculation:* 16 Correct (+3) + 17 Incorrect (-1) + 0 API/Parse Fail (-1) = 31 +8(There are some 4 marks mcq questions)
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 68    | 20            | 0                 | 12               | 0       | 0                  |
+| Math          | 45    | 16            | 1                 | 15               | 0       | 0                  |
+| Physics       | 33    | 12            | 2                 | 18               | 0       | 0                  |

results/google_gemini-2.5-flash-preview-05-20:thinking_JEE_ADVANCED_2025_20250607_154346/summary.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 2}
+{"question_id": "JA25P1P03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1P04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 2}
+{"question_id": "JA25P1P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 2}
+{"question_id": "JA25P1P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 2}
+{"question_id": "JA25P1P08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "JA25P1P09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["23"], "ground_truth": [["21", "25"]], "attempt": 2}
+{"question_id": "JA25P1P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "JA25P1P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.5"], "ground_truth": ["0.5", "0.75"], "attempt": 2}
+{"question_id": "JA25P1P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["76"], "ground_truth": [["75", "79"]], "attempt": 2}
+{"question_id": "JA25P1P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["72"], "ground_truth": ["72"], "attempt": 2}
+{"question_id": "JA25P1P14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 2}
+{"question_id": "JA25P1C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P1C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["100"], "ground_truth": ["100"], "attempt": 1}
+{"question_id": "JA25P1C09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": [["2.2", "2.3"]], "attempt": 2}
+{"question_id": "JA25P1C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-7.1"], "ground_truth": [["-7.2", "-7"]], "attempt": 2}
+{"question_id": "JA25P1C11", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["-30"], "ground_truth": [["-29.95", "-29.8"], ["29.8", "29.95"]], "attempt": 2}
+{"question_id": "JA25P1C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["280"], "ground_truth": ["280"], "attempt": 1}
+{"question_id": "JA25P1C13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["170"], "ground_truth": ["175"], "attempt": 2}
+{"question_id": "JA25P1C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["16"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 2}
+{"question_id": "JA25P1M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 2}
+{"question_id": "JA25P1M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 2}
+{"question_id": "JA25P1M08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105"], "ground_truth": ["105"], "attempt": 2}
+{"question_id": "JA25P1M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.2"], "ground_truth": [["1.15", "1.25"]], "attempt": 2}
+{"question_id": "JA25P1M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["762"], "ground_truth": ["762"], "attempt": 2}
+{"question_id": "JA25P1M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.4"], "ground_truth": [["2.35", "2.45"]], "attempt": 2}
+{"question_id": "JA25P1M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["96"], "ground_truth": ["96"], "attempt": 2}
+{"question_id": "JA25P1M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "JA25P1M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P2P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2P04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["A", "B", "C", "D"], "attempt": 2}
+{"question_id": "JA25P2P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 2}
+{"question_id": "JA25P2P06", "marks_awarded": 1, "evaluation_status": "partial_1_of_2_plus", "predicted_answer": ["A"], "ground_truth": ["A", "B"], "attempt": 2}
+{"question_id": "JA25P2P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2P08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["1"], "ground_truth": [["1.65", "1.67"]], "attempt": 2}
+{"question_id": "JA25P2P10", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["12"], "ground_truth": [["11.7", "11.9"]], "attempt": 2}
+{"question_id": "JA25P2P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.6"], "ground_truth": ["1.6"], "attempt": 2}
+{"question_id": "JA25P2P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.331"], "ground_truth": [["2.3", "2.4"]], "attempt": 2}
+{"question_id": "JA25P2P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.2"], "ground_truth": ["0.2"], "attempt": 2}
+{"question_id": "JA25P2P14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2.4"], "ground_truth": ["1.2"], "attempt": 2}
+{"question_id": "JA25P2P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["170"], "ground_truth": [["167", "171"]], "attempt": 2}
+{"question_id": "JA25P2P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["32"], "ground_truth": [["26", "33"]], "attempt": 2}
+{"question_id": "JA25P2C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2C02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2C03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["D"], "attempt": 2}
+{"question_id": "JA25P2C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["C", "D"], "ground_truth": ["C", "D"], "attempt": 1}
+{"question_id": "JA25P2C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P2C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 2}
+{"question_id": "JA25P2C08", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B", "C"], "ground_truth": ["B", "C"], "attempt": 2}
+{"question_id": "JA25P2C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11"], "ground_truth": [["10.85", "11.1"]], "attempt": 2}
+{"question_id": "JA25P2C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": [["3.85", "4.15"]], "attempt": 2}
+{"question_id": "JA25P2C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["16"], "ground_truth": [["15.5", "16.5"]], "attempt": 2}
+{"question_id": "JA25P2C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": [["4", "4.25"]], "attempt": 2}
+{"question_id": "JA25P2C13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": [["2.4", "2.55"]], "attempt": 2}
+{"question_id": "JA25P2C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105.5"], "ground_truth": [["105.4", "105.6"]], "attempt": 2}
+{"question_id": "JA25P2C15", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["8"], "ground_truth": [["7.5", "7.8"]], "attempt": 2}
+{"question_id": "JA25P2C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "JA25P2M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P2M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 2}
+{"question_id": "JA25P2M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 2}
+{"question_id": "JA25P2M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 2}
+{"question_id": "JA25P2M08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C", "D"], "ground_truth": ["B", "C", "D"], "attempt": 2}
+{"question_id": "JA25P2M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.75"], "ground_truth": [["0.7", "0.8"]], "attempt": 2}
+{"question_id": "JA25P2M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["6"], "ground_truth": ["6"], "attempt": 2}
+{"question_id": "JA25P2M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.30"], "ground_truth": [["0.27", "0.33"]], "attempt": 2}
+{"question_id": "JA25P2M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 2}
+{"question_id": "JA25P2M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 2}
+{"question_id": "JA25P2M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.25"], "ground_truth": [["0.2", "0.3"]], "attempt": 2}
+{"question_id": "JA25P2M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "JA25P2M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["21"], "ground_truth": ["21"], "attempt": 2}

results/google_gemini-2.5-flash-preview-05-20:thinking_JEE_ADVANCED_2025_20250607_154346/summary.md

@@ -0,0 +1,33 @@
+# Benchmark Results: google/gemini-2.5-flash-preview-05-20:thinking
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2025
+**Timestamp:** 20250607_154346
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 482
+**Total Questions Processed in this Run:** 96
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **295** / **360**
+- **Fully Correct Answers:** 80
+- **Partially Correct Answers:** 1
+- **Incorrectly Answered (Choice Made):** 15
+- **Skipped Questions:** 0
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 96
+
+### Detailed Score Calculation by Question Type
+**Integer (42 questions):** 136 marks
+  *Calculation:* 34 Correct (+4) + 8 Incorrect (0) = 136
+**Mcq Multiple Correct (21 questions):** 75 marks
+  *Calculation:* 19 Correct (+4) + 1 Partial + 1 Incorrect (-2) = 75
+**Mcq Single Correct (33 questions):** 71 marks
+  *Calculation:* 27 Correct (+3) + 6 Incorrect (-1) + 0 API/Parse Fail (-1) = 75 + 9(some 4 marks mcq is there)
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 82    | 23            | 0                 | 9                | 0       | 0                  |
+| Math          | 116   | 31            | 0                 | 1                | 0       | 0                  |
+| Physics       | 97    | 26            | 1                 | 5                | 0       | 0                  |

results/google_gemini-2.5-flash-preview-05-20:thinking_NEET_2025_20250614_154446/summary.jsonl

@@ -0,0 +1,180 @@
+{"question_id": "N2545001", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545002", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545003", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545004", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545005", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545006", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545007", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545008", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545009", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545010", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545011", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545012", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545013", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545014", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545015", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545016", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545017", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545018", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545019", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545020", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545021", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545022", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545023", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545024", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545025", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545026", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545027", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545028", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545029", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545030", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545031", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545032", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545033", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545034", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545035", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545036", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545037", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545038", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["3"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545039", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545040", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["1", "2"], "attempt": 2}
+{"question_id": "N2545041", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545042", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545043", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545044", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545045", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545046", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545047", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545048", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545049", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545050", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545051", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545052", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545053", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545054", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545055", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545056", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545057", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545058", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545059", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545060", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545061", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545062", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545063", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1", "2"], "attempt": 2}
+{"question_id": "N2545064", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545065", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545066", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545067", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545068", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545069", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545070", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545071", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545072", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545073", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545074", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545075", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545076", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545077", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545078", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["3"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545079", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545080", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545081", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545082", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545083", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545084", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545085", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545086", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 2}
+{"question_id": "N2545087", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545088", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545089", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545090", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545091", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545092", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545093", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545094", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545095", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545096", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545097", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545098", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545099", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545100", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545101", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545102", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545103", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545104", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545105", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545106", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545107", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545108", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545109", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545110", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545111", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545112", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545113", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545114", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545115", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545116", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545117", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545118", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545119", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545120", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545121", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545122", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545123", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 2}
+{"question_id": "N2545124", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545125", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545126", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545127", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545128", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545129", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545130", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545131", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545132", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545133", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545134", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545135", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545136", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545137", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545138", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545139", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545140", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545141", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545142", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545143", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545144", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545145", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545146", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545147", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545148", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545149", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545150", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545151", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545152", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545153", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545154", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545155", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545156", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545157", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["1"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545158", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545159", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545160", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545161", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545162", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545163", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545164", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545165", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545166", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 2}
+{"question_id": "N2545167", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545168", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545169", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545170", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545171", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545172", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 2}
+{"question_id": "N2545173", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545174", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545175", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["2"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "N2545176", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545177", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "N2545178", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1"], "ground_truth": ["1"], "attempt": 1}
+{"question_id": "N2545179", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": ["4"], "attempt": 1}
+{"question_id": "N2545180", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}

results/google_gemini-2.5-flash-preview-05-20:thinking_NEET_2025_20250614_154446/summary.md

@@ -0,0 +1,28 @@
+# Benchmark Results: google/gemini-2.5-flash-preview-05-20:thinking
+**Exam Name:** NEET
+**Exam Year:** 2025
+**Timestamp:** 20250614_154446
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 398
+**Total Questions Processed in this Run:** 180
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **685** / **720**
+- **Fully Correct Answers:** 173
+- **Incorrectly Answered (Choice Made):** 7
+- **Skipped Questions:** 0
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 180
+
+### Detailed Score Calculation by Question Type
+**Mcq Single Correct (180 questions):** 685 marks
+  *Calculation:* 173 Correct (+4) + 7 Incorrect (-1) = 685
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Biology       | 340   | 86            | 0                 | 4                | 0       | 0                  |
+| Chemistry     | 175   | 44            | 0                 | 1                | 0       | 0                  |
+| Physics       | 170   | 43            | 0                 | 2                | 0       | 0                  |

results/google_gemini-2.5-pro-preview-03-25_JEE_ADVANCED_2024_20250527_100743/summary.json

@@ -0,0 +1,43 @@
+{
+    "model_name": "google/gemini-2.5-pro-preview-03-25",
+    "exam_name": "JEE_ADVANCED",
+    "exam_year": "2024",
+    "question_ids_filter": "None",
+    "timestamp": "20250527_100743",
+    "total_questions_in_dataset": 482,
+    "total_questions_processed_in_run": 102,
+    "overall_score": 322,
+    "overall_correct_full": 91,
+    "overall_partial_correct": 0,
+    "overall_incorrect_choice": 7,
+    "overall_skipped": 2,
+    "overall_api_parse_failures": 2,
+    "total_questions_processed": 102,
+    "unmapped_section_questions": 0,
+    "section_breakdown": {
+        "Chemistry": {
+            "score": 102,
+            "correct": 29,
+            "incorrect": 4,
+            "skipped": 1,
+            "api_parse_failures": 0,
+            "partial_correct": 0
+        },
+        "Math": {
+            "score": 113,
+            "correct": 32,
+            "incorrect": 1,
+            "skipped": 1,
+            "api_parse_failures": 0,
+            "partial_correct": 0
+        },
+        "Physics": {
+            "score": 107,
+            "correct": 31,
+            "incorrect": 3,
+            "skipped": 0,
+            "api_parse_failures": 0,
+            "partial_correct": 0
+        }
+    }
+}

results/google_gemini-2.5-pro-preview-03-25_JEE_ADVANCED_2024_20250527_100743/summary.md

@@ -0,0 +1,24 @@
+# Benchmark Results: google/gemini-2.5-pro-preview-03-25
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2024
+**Timestamp:** 20250527_100743
+**Total Questions in Dataset:** 482
+**Questions Filtered Out:** 380
+**Total Questions Processed in this Run:** 102
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **322** (Max score is 360)
+- **Fully Correct Answers:** 92
+- **Incorrectly Answered (Choice Made):** 8
+- **Skipped Questions:** 2
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 102
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 102   | 29            | 0                 | 4                | 1       | 0                  |
+| Math          | 113   | 32            | 0                 | 1                | 1       | 0                  |
+| Physics       | 107    | 31            | 0                 | 3                | 0       | 0                  |

results/google_gemini-2.5-pro-preview-03-25_JEE_ADVANCED_2025_20250601_180855/summary.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["A"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 2}
+{"question_id": "JA25P1P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 2}
+{"question_id": "JA25P1P08", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["1"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1P09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["23"], "ground_truth": [["21", "25"]], "attempt": 1}
+{"question_id": "JA25P1P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P1P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.5"], "ground_truth": ["0.5", "0.75"], "attempt": 2}
+{"question_id": "JA25P1P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["75.6"], "ground_truth": [["75", "79"]], "attempt": 1}
+{"question_id": "JA25P1P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["72"], "ground_truth": ["72"], "attempt": 1}
+{"question_id": "JA25P1P14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P1C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P1C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["100"], "ground_truth": ["100"], "attempt": 1}
+{"question_id": "JA25P1C09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["5"], "ground_truth": [["2.2", "2.3"]], "attempt": 1}
+{"question_id": "JA25P1C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-7.1"], "ground_truth": [["-7.2", "-7"]], "attempt": 1}
+{"question_id": "JA25P1C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["29.88"], "ground_truth": [["-29.95", "-29.8"], ["29.8", "29.95"]], "attempt": 1}
+{"question_id": "JA25P1C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["280"], "ground_truth": ["280"], "attempt": 1}
+{"question_id": "JA25P1C13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["116.67"], "ground_truth": ["175"], "attempt": 1}
+{"question_id": "JA25P1C14", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P1C15", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1C16", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 2}
+{"question_id": "JA25P1M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P1M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105"], "ground_truth": ["105"], "attempt": 1}
+{"question_id": "JA25P1M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.2"], "ground_truth": [["1.15", "1.25"]], "attempt": 1}
+{"question_id": "JA25P1M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["762"], "ground_truth": ["762"], "attempt": 1}
+{"question_id": "JA25P1M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.4"], "ground_truth": [["2.35", "2.45"]], "attempt": 1}
+{"question_id": "JA25P1M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["96"], "ground_truth": ["96"], "attempt": 1}
+{"question_id": "JA25P1M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P1M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["A", "B", "C", "D"], "attempt": 2}
+{"question_id": "JA25P2P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 2}
+{"question_id": "JA25P2P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.67"], "ground_truth": [["1.65", "1.67"]], "attempt": 1}
+{"question_id": "JA25P2P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11.8"], "ground_truth": [["11.7", "11.9"]], "attempt": 1}
+{"question_id": "JA25P2P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.6"], "ground_truth": ["1.6"], "attempt": 1}
+{"question_id": "JA25P2P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.331"], "ground_truth": [["2.3", "2.4"]], "attempt": 1}
+{"question_id": "JA25P2P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.2"], "ground_truth": ["0.2"], "attempt": 1}
+{"question_id": "JA25P2P14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2.4"], "ground_truth": ["1.2"], "attempt": 1}
+{"question_id": "JA25P2P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["170"], "ground_truth": [["167", "171"]], "attempt": 1}
+{"question_id": "JA25P2P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["32"], "ground_truth": [["26", "33"]], "attempt": 2}
+{"question_id": "JA25P2C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 2}
+{"question_id": "JA25P2C02", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["B"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P2C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["C", "D"], "ground_truth": ["C", "D"], "attempt": 2}
+{"question_id": "JA25P2C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 2}
+{"question_id": "JA25P2C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2C08", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P2C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11"], "ground_truth": [["10.85", "11.1"]], "attempt": 1}
+{"question_id": "JA25P2C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3.95"], "ground_truth": [["3.85", "4.15"]], "attempt": 1}
+{"question_id": "JA25P2C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["16"], "ground_truth": [["15.5", "16.5"]], "attempt": 1}
+{"question_id": "JA25P2C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": [["4", "4.25"]], "attempt": 1}
+{"question_id": "JA25P2C13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.49"], "ground_truth": [["2.4", "2.55"]], "attempt": 1}
+{"question_id": "JA25P2C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105.5"], "ground_truth": [["105.4", "105.6"]], "attempt": 1}
+{"question_id": "JA25P2C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["7.73"], "ground_truth": [["7.5", "7.8"]], "attempt": 1}
+{"question_id": "JA25P2C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P2M01", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": ["C"], "attempt": 2}
+{"question_id": "JA25P2M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 2}
+{"question_id": "JA25P2M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 2}
+{"question_id": "JA25P2M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C", "D"], "ground_truth": ["B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.75"], "ground_truth": [["0.7", "0.8"]], "attempt": 1}
+{"question_id": "JA25P2M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["6"], "ground_truth": ["6"], "attempt": 1}
+{"question_id": "JA25P2M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.3"], "ground_truth": [["0.27", "0.33"]], "attempt": 1}
+{"question_id": "JA25P2M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 2}
+{"question_id": "JA25P2M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.25"], "ground_truth": [["0.2", "0.3"]], "attempt": 1}
+{"question_id": "JA25P2M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P2M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["21"], "ground_truth": ["21"], "attempt": 1}

results/google_gemini-2.5-pro-preview-03-25_JEE_ADVANCED_2025_20250601_180855/summary.md

@@ -0,0 +1,32 @@
+# Benchmark Results: google/gemini-2.5-pro-preview-05-06
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2025
+**Timestamp:** 20250601_180855
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 482
+**Total Questions Processed in this Run:** 96
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **323** / **360**
+- **Fully Correct Answers:** 87
+- **Incorrectly Answered (Choice Made):** 8
+- **Skipped Questions:** 1
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 96
+
+### Detailed Score Calculation by Question Type
+**Integer (42 questions):** 152 marks
+  *Calculation:* 38 Correct (+4) + 4 Incorrect (0) = 152
+**Mcq Multiple Correct (21 questions):** 78 marks
+  *Calculation:* 20 Correct (+4) + 1 Incorrect (-2) = 78
+**Mcq Single Correct (33 questions):** 93 marks
+  *Calculation:* 29 Correct (+3) + 2 Incorrect (-1) + 1 Skipped (0) = 85 (+8 as there are some mcq single correct questions which have full marks 4)
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 102    | 28            | 0                 | 4                | 0       | 0                  |
+| Math          | 113   | 30            | 0                 | 1                | 1       | 0                  |
+| Physics       | 108    | 29            | 0                 | 3                | 0       | 0                  |

results/google_gemini-2.5-pro-preview-03-25_NEET_2024_20250505_091752/predictions.jsonl

@@ -0,0 +1,200 @@
+{"question_id": "NEET_2024_T3_001", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_002", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_003", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` P\n\u2191\nd\nC\n300 kPa\n100 kPa\nb\na\n3\n3\n100 cm\n400 cm\nV-\n```\n\nThe question asks for the work done by the gas along the path *bc*.\nWork done by the gas is given by the integral W = \u222b P dV.\nAlong the path *bc*, the process goes from state *b* to state *c*.\nFrom the diagram:\nState *b*: V_b = 400 cm\u00b3, P_b = 100 kPa\nState *c*: V_c = 400 cm\u00b3, P_c = 300 kPa\n\nSince the volume is constant along the path *bc* (V_b = V_c = 400 cm\u00b3), the change in volume dV is 0.\nTherefore, the work done W_bc = \u222b P dV = P \u0394V = P * (V_c - V_b) = P * (400 cm\u00b3 - 400 cm\u00b3) = 0.\nThe work done by the gas along the path *bc* is zero.\n\nComparing this with the options:\n(1) -90 J\n(2) -60 J\n(3) zero\n(4) 30 J\n\nThe correct option is (3).\n\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_004", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_005", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_006", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_007", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_008", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_009", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_010", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "``` Mnemonic:\n1. Analyze the question: Replace monochromatic light with white light in YDSE.\n2. Apply physics concepts:\n    - Central fringe (path difference = 0): Constructive interference for ALL wavelengths. Result: White light.\n    - Other fringes (path difference \u2260 0): Position depends on wavelength (y = m\u03bbD/d). Different colors form fringes at different positions. Result: Colored fringes.\n    - Fringe width (\u03b2 = \u03bbD/d): Depends on wavelength. Fringes are not of equal width for different colors.\n3. Evaluate options:\n    - (1) Central bright white fringe, surrounded by colored fringes. -> Matches explanation.\n    - (2) All bright fringes equal width. -> Incorrect, width depends on \u03bb.\n    - (3) Pattern disappears. -> Incorrect, pattern forms but is colored.\n    - (4) Central \"dear\" fringe (typo). Central fringe is white.\n4. Select the best option: (1) is the most accurate description. ```\n\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_011", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` C\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_012", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_013", "subject": "Physics", "ground_truth": [3], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_014", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_015", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "``` SiliBocon\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_016", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_017", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n```json\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_018", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_019", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_020", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_021", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_022", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_023", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_024", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_025", "subject": "Physics", "ground_truth": [1, 3], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_026", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_027", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_028", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_029", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` T = 2\u03c0 * sqrt(I / (M * B)) ```\nWhere:\n*   T is the time period of oscillation\n*   I is the moment of inertia\n*   M is the magnetic moment\n*   B is the magnetic field strength\n\nFirst, calculate the time period T:\nThe needle performs 20 complete oscillations in 5 seconds.\n``` T = Time / Number of oscillations = 5 s / 20 = 1/4 s = 0.25 s ```\n\nGiven values:\n*   B = 0.049 T\n*   I = 9.8 x 10\u207b\u2076 kg m\u00b2\n*   T = 0.25 s\n*   M = x * 10\u207b\u2075 Am\u00b2\n\nRearrange the formula to solve for M:\n``` T\u00b2 = (2\u03c0)\u00b2 * (I / (M * B)) ```\n``` T\u00b2 = 4\u03c0\u00b2 * I / (M * B) ```\n``` M = 4\u03c0\u00b2 * I / (T\u00b2 * B) ```\n\nSubstitute the values:\n``` M = 4\u03c0\u00b2 * (9.8 * 10\u207b\u2076) / ((0.25)\u00b2 * 0.049) ```\n``` M = 4\u03c0\u00b2 * (9.8 * 10\u207b\u2076) / (0.0625 * 0.049) ```\n``` M = 4\u03c0\u00b2 * (9.8 * 10\u207b\u2076) / 0.0030625 ```\n\nLet's simplify the calculation: 0.049 = 49 * 10\u207b\u00b3, 9.8 = 98 * 10\u207b\u00b9, 0.0625 = 1/16.\n``` M = 4\u03c0\u00b2 * (9.8 * 10\u207b\u2076) / ((1/16) * 0.049) ```\n``` M = 4\u03c0\u00b2 * 16 * (9.8 * 10\u207b\u2076) / 0.049 ```\n``` M = 64\u03c0\u00b2 * (9.8 / 0.049) * 10\u207b\u2076 ```\n``` M = 64\u03c0\u00b2 * (980 / 49) * 10\u207b\u2076 ```\n``` M = 64\u03c0\u00b2 * 20 * 10\u207b\u2076 ```\n``` M = 1280\u03c0\u00b2 * 10\u207b\u2076 ```\n``` M = 12.80\u03c0\u00b2 * 10\u207b\u2074 ```\n``` M = 1.280\u03c0\u00b2 * 10\u207b\u00b3 ```\n``` M = 0.1280\u03c0\u00b2 * 10\u207b\u00b2 ```\n``` M = 1280\u03c0\u00b2 * 10\u207b\u2076 Am\u00b2 ```\n\nWe are given that M = x * 10\u207b\u2075 Am\u00b2. Let's convert our result to this form:\n``` M = 1280 * 10\u207b\u00b9 * \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2 ```\n``` M = 128.0 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2 ```\n\nWait, let me recalculate 64 * 20.\n64 * 20 = 1280.\nSo, M = 1280 \u03c0\u00b2 * 10\u207b\u2076 Am\u00b2\nConvert to 10\u207b\u2075:\nM = (1280 / 10) \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2\nM = 128 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2\n\nComparing M = 128 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2 with M = x * 10\u207b\u2075 Am\u00b2:\nx = 128 \u03c0\u00b2\n\nLooking at the options:\n(1) 50 \u03c0\u00b2\n(2) 1280 \u03c0\u00b2\n(3) 5 \u03c0\u00b2\n(4) 128 \u03c0\u00b2\n\nThe value of x is 128 \u03c0\u00b2, which corresponds to option (4).\n\nLet me recheck the calculation:\nM = 4\u03c0\u00b2 * I / (T\u00b2 * B)\nM = 4\u03c0\u00b2 * (9.8 x 10\u207b\u2076) / ( (1/4)\u00b2 * 0.049 )\nM = 4\u03c0\u00b2 * (9.8 x 10\u207b\u2076) / ( (1/16) * 0.049 )\nM = 4\u03c0\u00b2 * 16 * (9.8 x 10\u207b\u2076) / 0.049\nM = 64\u03c0\u00b2 * (9.8 / 0.049) * 10\u207b\u2076\nM = 64\u03c0\u00b2 * 200 * 10\u207b\u2076\nM = 12800 \u03c0\u00b2 * 10\u207b\u2076\nM = 12800 * 10\u207b\u00b9 \u03c0\u00b2 * 10\u207b\u2075\nM = 1280 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2\n\nComparing M = 1280 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2 with M = x * 10\u207b\u2075 Am\u00b2:\nx = 1280 \u03c0\u00b2\n\nSo the correct option is (2).\n\nLet's review the calculations once more to be sure.\nT = 5/20 = 1/4 s\nB = 0.049 T\nI = 9.8 x 10\u207b\u2076 kg m\u00b2\nM = x * 10\u207b\u2075 Am\u00b2\n\nT = 2\u03c0 sqrt(I / MB)\nT\u00b2 = 4\u03c0\u00b2 I / (MB)\nM = 4\u03c0\u00b2 I / (T\u00b2 B)\nM = 4\u03c0\u00b2 (9.8 x 10\u207b\u2076) / ((1/4)\u00b2 * 0.049)\nM = 4\u03c0\u00b2 (9.8 x 10\u207b\u2076) / ((1/16) * 0.049)\nM = 4\u03c0\u00b2 * 16 * (9.8 x 10\u207b\u2076) / 0.049\nM = 64\u03c0\u00b2 * (9.8/0.049) * 10\u207b\u2076\nM = 64\u03c0\u00b2 * 200 * 10\u207b\u2076\nM = 12800 \u03c0\u00b2 * 10\u207b\u2076\nM = (12800 / 10) \u03c0\u00b2 * 10\u207b\u2075\nM = 1280 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2\n\nComparing M = x * 10\u207b\u2075 Am\u00b2\nx = 1280 \u03c0\u00b2\n\nThe calculation seems correct. The value of x is 1280 \u03c0\u00b2. This corresponds to option (2).\n\nFinal check:\nT = 0.25 s\nT\u00b2 = 0.0625 s\u00b2\nB = 0.049 T\nI = 9.8 x 10\u207b\u2076 kg m\u00b2\nM = 4\u03c0\u00b2 * (9.8e-6) / (0.0625 * 0.049)\nM = 4\u03c0\u00b2 * (9.8e-6) / 0.0030625\nM = 4\u03c0\u00b2 * 3200 * 10\u207b\u2076\nM = 12800 \u03c0\u00b2 * 10\u207b\u2076\nM = 1.28 \u03c0\u00b2 * 10\u207b\u00b2\nM = 1280 \u03c0\u00b2 * 10\u207b\u2075 Am\u00b2\nx = 1280 \u03c0\u00b2\nOption (2) is correct.<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_030", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_031", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_032", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` M\u2070L\u2070T\u2070 ```\n\n*   **Strain:** Strain is defined as the ratio of change in dimension (length, area, volume) to the original dimension. For example, longitudinal strain = \u0394L/L. Since both numerator and denominator have dimensions of length [L], strain is dimensionless: ``` [L]/[L] = [M\u2070L\u2070T\u2070] ```.\n*   **Arc:** Arc length is a measure of distance along a curve, so it has dimensions of length: ``` [L] ```.\n*   **Angular speed (\u03c9):** Angular speed is the rate of change of angle (\u03b8) with respect to time (t), \u03c9 = d\u03b8/dt. Angle (plane angle) is dimensionless, and time has dimensions [T]. So, angular speed has dimensions: ``` [M\u2070L\u2070T\u2070] / [T] = [T\u207b\u00b9] ```.\n*   **Stress:** Stress is defined as force (F) per unit area (A). Force has dimensions [MLT\u207b\u00b2] and area has dimensions [L\u00b2]. So, stress has dimensions: ``` [MLT\u207b\u00b2] / [L\u00b2] = [ML\u207b\u00b9T\u207b\u00b2] ```.\n*   **Angle (plane angle, \u03b8):** Plane angle is defined as the ratio of the arc length (s) subtended to the radius (r), \u03b8 = s/r. Arc length has dimensions [L] and radius has dimensions [L]. So, angle is dimensionless: ``` [L]/[L] = [M\u2070L\u2070T\u2070] ```.\n\nThe question asks for quantities that have the same dimensions as a solid angle. A solid angle (\u03a9) is defined as the ratio of the subtended area (A) on a sphere to the square of the radius (r\u00b2), \u03a9 = A/r\u00b2. Area has dimensions [L\u00b2] and radius squared has dimensions [L\u00b2]. Thus, solid angle is dimensionless: ``` [L\u00b2] / [L\u00b2] = [M\u2070L\u2070T\u2070] ```.\n\nWe need to find the option where both quantities are dimensionless.\n\n*   Option (1): Strain (dimensionless), Arc ([L]). Not both dimensionless.\n*   Option (2): Angular speed ([T\u207b\u00b9]), Stress ([ML\u207b\u00b9T\u207b\u00b2]). Neither is dimensionless.\n*   Option (3): Strain (dimensionless), Angle (dimensionless). Both are dimensionless, matching the dimensions of a solid angle.\n*   Option (4): Stress ([ML\u207b\u00b9T\u207b\u00b2]), Angle (dimensionless). Not both dimensionless.\n\nTherefore, the correct option is (3) because both strain and angle are dimensionless, just like a solid angle.\n\n```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_033", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_034", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_035", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_036", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_037", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_038", "subject": "Physics", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_039", "subject": "Physics", "ground_truth": [3], "predicted_answer": [4], "raw_response": "``` C\n1.  Analyze the Question: The question asks to identify the circuit diagram that represents a Wheatstone bridge configuration capable of achieving balance.\n2.  Analyze the Options: Four circuit diagrams labeled (1), (2), (3), and (4) are provided. All four diagrams depict a Wheatstone bridge circuit with four resistors (10\u03a9, 10\u03a9, 15\u03a9, 5\u03a9), a galvanometer (G), a component labeled 'D', and a voltage source (E).\n3.  Check Balance Condition: The condition for a balanced Wheatstone bridge is R1/R2 = R3/R4, where R1, R2, R3, R4 are the resistors in the arms. Using the values given (labeling TL=10, TR=10, BL=15, BR=5):\n    *   TL/BL = 10/15 = 2/3\n    *   TR/BR = 10/5 = 2\n    *   Since 2/3 \u2260 2, the bridge with these specific *fixed* resistor values is not balanced.\n4.  Interpret \"Can Achieve Balance\": The question asks which circuit *can* achieve balance. This might imply either:\n    *   One of the components is variable.\n    *   The diagram represents the correct *setup* for a potentially balanceable bridge, even if the values shown don't balance it.\n5.  Analyze Component Placement:\n    *   Source E: Connected correctly across two opposite vertices in all options.\n    *   Galvanometer G: Connected correctly between the other two opposite vertices (the central branch) in all options.\n    *   Component D:\n        *   In options 1, 2, and 3, 'D' is placed in series with the 5\u03a9 resistor in the bottom-right arm.\n        *   In option 4, 'D' is placed in series with the galvanometer 'G' in the central branch.\n6.  Identify Structural Correctness for Detection: The central branch (where G is located) is where the balance detector (typically a galvanometer) is placed to sense zero current. Placing an additional component 'D', possibly related to detection, in this same central branch (as in option 4) maintains the correct structural location for the detection part of the circuit. Placing 'D' in series with one of the main bridge arms (as in options 1, 2, 3) is non-standard for a *detector*. If 'D' were a variable resistor *for balancing*, then placement in an arm (like 1, 2, 3) would be logical. However, typically, a component labeled 'D' alongside 'G' would imply Detector.\n7.  Consider Identical Options: Options (1), (2), and (3) are electrically and visually identical diagrams. In a multiple-choice question with a single correct answer, if one of these were correct, they all would be. This suggests that either the question/options are flawed or that option (4) must be the intended answer based on some principle. The principle most likely relates to the correct *placement* of detection components (G and D) in the central bridge arm.\n8.  Conclusion: Although the specific values don't lead to balance, option 4 represents the most standard configuration regarding the placement of detection components (G and implicitly D if it's also for detection) within the central branch of the Wheatstone bridge circuit.\n\n```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_040", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_041", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "``` C\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_042", "subject": "Physics", "ground_truth": [4], "predicted_answer": "SKIP", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "skipped", "marks_awarded": 0}
+{"question_id": "NEET_2024_T3_043", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_044", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_045", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_046", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_047", "subject": "Physics", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_048", "subject": "Physics", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_049", "subject": "Physics", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```cpp\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_050", "subject": "Physics", "ground_truth": [4], "predicted_answer": "SKIP", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "skipped", "marks_awarded": 0}
+{"question_id": "NEET_2024_T3_051", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_052", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```xml\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_053", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_054", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_055", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_056", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_057", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_058", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_059", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_060", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_061", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_062", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_063", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_064", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```c\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_065", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_066", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_067", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_068", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_069", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n```json\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_070", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_071", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```xml\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_072", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_073", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_074", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n[\n  {\"box_2d\": [94, 118, 117, 149], \"text\": \"74.\"},\n  {\"box_2d\": [117, 118, 202, 956], \"text\": \"In which of the following equilibria, Kp and Ke are\"},\n  {\"box_2d\": [217, 116, 255, 298], \"text\": \"NOT equal?\"},\n  {\"box_2d\": [266, 116, 316, 701], \"text\": \"(1) CO(g) + H2O(g)CO2(g) + H2(g)\"},\n  {\"box_2d\": [396, 115, 447, 671], \"text\": \"(2) 2 BrCl(g) Br2(g) + Cl2(g)\"},\n  {\"box_2d\": [529, 116, 571, 148], \"text\": \"(3)\"},\n  {\"box_2d\": [531, 189, 572, 280], \"text\": \"PCl5(g)\"},\n  {\"box_2d\": [554, 400, 596, 650], \"text\": \"PCl3(g) + Cl2(g)\"},\n  {\"box_2d\": [656, 116, 699, 147], \"text\": \"(4)\"},\n  {\"box_2d\": [658, 188, 712, 598], \"text\": \"H2(g) + I2(g) = 2 HI(g)\"}\n]\n```\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_075", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_076", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` C6H11+\n<answer>2</answer>```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_077", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_078", "subject": "Chemistry", "ground_truth": [1], "predicted_answer": [1], "raw_response": "``` D. **Cl2(g) \u2192 2Cl(g):** One mole of gas (Cl2) forms two moles of gas (2Cl). The number of moles of gas increases (\u0394n_g = 2 - 1 = 1 > 0), leading to an increase in entropy. Entropy increases.\n\nTherefore, entropy increases in processes A, C, and D.\n\nLooking at the options:\n(1) A, C and D\n(2) C and D\n(3) A and C\n(4) A, B and D\n\nThe correct option includes A, C, and D. This corresponds to option (1).\n\n```<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_079", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_080", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_081", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_082", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "``` C\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_083", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_084", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` DRAFT\n1.  **Analyze the question:** The question asks to identify the most stable carbocation among the four given chemical structures. Carbocation stability increases with the number of alkyl groups attached to the positively charged carbon atom due to inductive effects and hyperconjugation. The order of stability is generally: tertiary (3\u00b0) > secondary (2\u00b0) > primary (1\u00b0) > methyl. Resonance can further enhance stability.\n\n2.  **Analyze the options:**\n    *   **(1)** This is a primary carbocation. The positively charged carbon (-CH\u2082\u207a) is attached to only one other carbon atom (the cyclopentyl ring karbon).\n    *   **(2)** This is a tertiary carbocation. The positively charged carbon (within the cyclohexane ring) is attached to three other carbon atoms (two within the ring and one from the methyl group).\n    *   **(3)** This is a secondary carbocation. The positively charged carbon (-CH\u207a-) is attached to two other carbon atoms (one from the ethyl group -CH\u2082- and one from the isopropyl group -CH-). Specifically, it is 3-methylpentan-3-yl cation based on the drawing: CH3-CH2-CH(+)-CH(CH3)2. This is incorrect reading of the structure. Actually it is CH3-CH2-CH(+)-CH(CH3)2 - yes, this is secondary.\n    *   **(4)** This is a secondary carbocation. The positively charged carbon (-CH\u207a-) is attached to two other carbon atoms (one from the methyl group -CH\u2083 and one from the isobutyl group -CH\u2082-). Specifically, it is 4-methylpentan-2-yl cation: CH3-CH(+)-CH2-CH(CH3)2.\n\n3.  **Compare stability:** Based on the order of stability (tertiary > secondary > primary), the tertiary carbocation in option (2) is the most stable. Options (1) is primary, and options (3) and (4) are secondary.\n\n4.  **Determine the correct integer:** The correct option is (2).\n\n5.  **Format the answer:** <answer>2</answer>\n```\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_085", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_086", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n```json\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_087", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_088", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` Memento ```python\nimport math\n\n# Given information\nslope = 25.73  # L bar mol^-1\nR = 0.083  # L bar mol^-1 K^-1\n\n# Formula relating osmotic pressure (\u03a0), concentration (C), and temperature (T)\n# \u03a0 = i * C * R * T\n# Plot of \u03a0 vs C gives a straight line with slope = i * R * T\n# Assuming i=1 (van't Hoff factor for non-electrolyte or ideal solution)\n# Slope = R * T\n\n# Calculate Temperature in Kelvin (T_K)\nT_K = slope / R\nT_K = 25.73 / 0.083\nT_K = 310.0  # Kelvin\n\n# Convert Temperature to Celsius (T_C)\n# T_C = T_K - 273.15\n# More commonly in these types of problems, 273 is used for conversion.\nT_C_approx = T_K - 273\nT_C_approx = 310 - 273\nT_C_approx = 37  # degrees Celsius\n\nT_C_exact = T_K - 273.15\nT_C_exact = 310.0 - 273.15\nT_C_exact = 36.85 # degrees Celsius\n\n# Check the options\noption1 = 25.73  # \u00b0C\noption2 = 12.05  # \u00b0C\noption3 = 37     # \u00b0C\noption4 = 310    # \u00b0C\n\n# The calculated temperature (36.85\u00b0C or approximately 37\u00b0C) matches option (3).\n\n# Select the option number corresponding to 37\u00b0C.\ncorrect_option_number = 3\n```\n\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_089", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "``` C6H11\n|\nC=CH + KMnO4/H+ -> 'P' (major product)\n| |\nH C6H11\n```\n\n1.  **Analyze the reaction:** The reaction involves an alkene (1,2-dicyclohexylethylene) reacting with potassium permanganate (KMnO4) in acidic conditions (H+). Hot, acidic KMnO4 is a strong oxidizing agent that causes oxidative cleavage of the carbon-carbon double bond.\n2.  **Determine the products of oxidative cleavage:** The double bond (C=C) is broken. Each carbon atom that was part of the double bond gets oxidized.\n    *   If a carbon atom of the double bond has one hydrogen attached (R-CH=), it gets oxidized to a carboxylic acid group (-COOH).\n    *   If a carbon atom of the double bond has two hydrogens attached (=CH2), it gets oxidized to carbon dioxide (CO2).\n    *   If a carbon atom of the double bond has no hydrogens attached (R-C(R')=), it gets oxidized to a ketone group (=O).\n3.  **Apply to the specific substrate:** In the given alkene, both carbon atoms of the double bond are identical in terms of substitution. Each carbon is attached to one hydrogen atom and one cyclohexyl group (C6H11).\n    ```\n    C6H11-CH=CH-C6H11\n    ```\n    Cleavage occurs at the double bond. The left `C6H11-CH=` fragment will be oxidized to `C6H11-COOH`. The right `=CH-C6H11` fragment will also be oxidized to `C6H11-COOH`.\n4.  **Identify the major product 'P':** The reaction breaks the molecule into two identical fragments, each being cyclohexanecarboxylic acid. Therefore, the major product 'P' is cyclohexanecarboxylic acid.\n5.  **Match with the options:**\n    *   Option (1) is 1,2-dicyclohexyl-1,2-ethanediol. This is the product of syn-dihydroxylation (e.g., cold, dilute, alkaline KMnO4 or OsO4). Not correct for hot/acidic KMnO4.\n    *   Option (2) is 1,2-dicyclohexylethanedione. This would imply oxidation without cleavage, or cleavage of a different starting material. Not correct.\n    *   Option (3) is cyclohexanecarbaldehyde. Aldehydes are intermediates but are readily oxidized to carboxylic acids by strong oxidizing agents like hot/acidic KMnO4. Not the final major product.\n    *   Option (4) is cyclohexanecarboxylic acid. This matches the expected product of oxidative cleavage.\n\nTherefore, the correct option is (4).\n\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_090", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n{\n \"step_by_step_thinking\": \"Step 1: Analyze Statement I. Statement I says '[Co(NH3)6]3+ is a homoleptic complex whereas [Co(NH3)4Cl2]+ is a heteroleptic complex.' A homoleptic complex has only one type of ligand attached to the central metal ion. In [Co(NH3)6]3+, only NH3 ligands are attached to Co. Thus, it is homoleptic. A heteroleptic complex has more than one type of ligand attached to the central metal ion. in [Co(NH3)4Cl2]+, both NH3 and Cl ligands are attached to Co. Thus, it is heteroleptic. Therefore, Statement I is true.\\nStep 2: Analyze Statement II. Statement II says 'Complex [Co(NH3)6]3+ has only one kind of ligands but [Co(NH3)4Cl2]+ has more than one kind of ligands.' As established in Step 1, [Co(NH3)6]3+ has only NH3 ligands (one kind). Also, [Co(NH3)4Cl2]+ has NH3 and Cl ligands (two kinds, which is more than one kind). Therefore, Statement II is true.\\nStep 3: Evaluate the options based on the truthfulness of the statements. Since both Statement I and Statement II are true, we need to find the option that reflects this.\\nOption (1) says Statement I is true but Statement II is false. This is incorrect.\\nOption (2) says Statement I is false but Statement II is true. This is incorrect.\\nOption (3) says Both Statement I and Statement II are true. This is correct.\\nOption (4) says Both Statement I and Statement II are false. This is incorrect.\\nStep 4: Identify the correct option number. The correct option is (3).\\nStep 5: Format the final answer as requested.\",\n \"final_answer\": \"<answer>3</answer>\"\n}\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_091", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_092", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_093", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_094", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` ELECTION ANALYSIS\n1.  **Step 1:** 1-Iodopropane (CH\u2083-CH\u2082-CH\u2082-I) reacts with sodium cyanide (NaCN). This is an SN2 reaction where the cyanide ion (CN\u207b) replaces the iodide ion (I\u207b). The product A is butanenitrile (CH\u2083-CH\u2082-CH\u2082-CN).\n2.  **Step 2:** Butanenitrile (A) undergoes partial hydrolysis with hydroxide ions (OH\u207b). Partial hydrolysis of a nitrile yields an amide. The product B is butanamide (CH\u2083-CH\u2082-CH\u2082-CONH\u2082).\n3.  **Step 3:** Butanamide (B) reacts with sodium hydroxide (NaOH) and bromine (Br\u2082). This is the Hofmann rearrangement (or Hofmann degradation) reaction. In this reaction, an amide is converted to a primary amine with one less carbon atom. Butanamide (4 carbons) is converted to propylamine (CH\u2083-CH\u2082-CH\u2082-NH\u2082), which has 3 carbons. Product C is propylamine.\n\n**Checking the Options:**\n*   (1) butanamide: This is intermediate B.\n*   (2) \u03b1-bromobutanoic acid: Incorrect product.\n*   (3) propylamine: This is the final product C (CH\u2083-CH\u2082-CH\u2082-NH\u2082).\n*   (4) butylamine: Incorrect product (same number of carbons as B, Hofmann degradation removes one carbon).\n\nTherefore, the correct option is (3).\n```\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_095", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_096", "subject": "Chemistry", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```xml\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_097", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [1], "raw_response": "```php\n```\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_098", "subject": "Chemistry", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_099", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` MSON\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_100", "subject": "Chemistry", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_101", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_102", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "1.  **Analyze the question:** The question describes a conservation method where threatened species are removed from their natural habitat and kept in a controlled, protected environment (like zoos, botanical gardens, seed banks). This method is known as *ex-situ* conservation (off-site conservation).\n\n2.  **Analyze the options:**\n    *   (1) Semi-conservative method: Refers to DNA replication. Incorrect.\n    *   (2) Sustainable development: A broad economic and environmental policy concept. Incorrect.\n    *   (3) *in-situ* conservation: Means conservation within the natural habitat (e.g., national parks, wildlife sanctuaries). This is the opposite of the method described in the question. Incorrect.\n    *   (4) Biodiversity conservation: This is the overall effort to preserve the variety of life on Earth. It encompasses both *in-situ* and *ex-situ* strategies.\n\n3.  **Determine the correct option:** The specific type of conservation described in the question is *ex-situ* conservation. This option is not provided. Among the given choices, *in-situ* conservation (3) is the opposite approach. Biodiversity conservation (4) is a broad term that includes *ex-situ* conservation as one of its strategies. Options (1) and (2) are irrelevant. Since the specific term (*ex-situ*) is missing, and *in-situ* is incorrect, the most appropriate answer among the given choices is the broader category under which *ex-situ* conservation falls, which is Biodiversity conservation (4). Although imprecise, it's the best fit given the options.\n\n4.  **Format the answer:** The correct option number is 4.\n\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_103", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_104", "subject": "Botany", "ground_truth": [1], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_105", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_106", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_107", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_108", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_109", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_110", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_111", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n```xml\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_112", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```xml\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_113", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_114", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_115", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_116", "subject": "Botany", "ground_truth": [2], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_117", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_118", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_119", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_120", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_121", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_122", "subject": "Botany", "ground_truth": [2], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_123", "subject": "Botany", "ground_truth": [3], "predicted_answer": [4], "raw_response": "``` MSON ```\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_124", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_125", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_126", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_127", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_128", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_129", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_130", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_131", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_132", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` MuiTypography::variant=\"body1\" ``````json\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_133", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_134", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` MMATCH FAIL: '\u0410-\u0406\u041f' vs 'A-III' ```\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_135", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_136", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_137", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_138", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_139", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_140", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_141", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n[\n  {\"box_2d\": [11, 108, 53, 496], \"label\": \"question\"},\n  {\"box_2d\": [65, 110, 649, 933], \"label\": \"table\"},\n  {\"box_2d\": [670, 110, 752, 972], \"label\": \"question\"},\n  {\"box_2d\": [765, 110, 971, 543], \"label\": \"options\"}\n]\n```\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_142", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_143", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_144", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n[\n    {\"point\": [94, 216], \"label\": \"list_item\"},\n    {\"point\": [96, 305], \"label\": \"list_item\"},\n    {\"point\": [101, 427], \"label\": \"list_item\"},\n    {\"point\": [94, 535], \"label\": \"list_item\"},\n    {\"point\": [424, 216], \"label\": \"list_item\"},\n    {\"point\": [424, 305], \"label\": \"list_item\"},\n    {\"point\": [424, 427], \"label\": \"list_item\"},\n    {\"point\": [424, 535], \"label\": \"list_item\"}\n]\n```\n```json\n[\n    {\"box_2d\": [410, 472, 530, 905], \"label\": \"list_item\"},\n    {\"box_2d\": [396, 133, 551, 421], \"label\": \"list_item\"},\n    {\"box_2d\": [811, 35, 849, 395], \"label\": \"list_item\"},\n    {\"box_2d\": [56, 110, 682, 960], \"label\": \"table\"},\n    {\"box_2d\": [58, 210, 151, 453], \"label\": \"list_item\"},\n    {\"box_2d\": [71, 504, 158, 806], \"label\": \"list_item\"},\n    {\"box_2d\": [151, 112, 266, 449], \"label\": \"list_item\"},\n    {\"box_2d\": [151, 453, 266, 962], \"label\": \"list_item\"},\n    {\"box_2d\": [153, 476, 208, 776], \"label\": \"list_item\"},\n    {\"box_2d\": [167, 229, 252, 438], \"label\": \"list_item\"},\n    {\"box_2d\": [307, 472, 373, 834], \"label\": \"list_item\"},\n    {\"box_2d\": [561, 112, 682, 449], \"label\": \"list_item\"},\n    {\"box_2d\": [274, 112, 392, 449], \"label\": \"list_item\"},\n    {\"box_2d\": [939, 35, 983, 482], \"label\": \"list_item\"},\n    {\"box_2d\": [722, 31, 799, 922], \"label\": \"list\"},\n    {\"box_2d\": [274, 453, 392, 958], \"label\": \"list_item\"},\n    {\"box_2d\": [896, 35, 939, 478], \"label\": \"list_item\"},\n    {\"box_2d\": [561, 451, 682, 958], \"label\": \"list_item\"},\n    {\"box_2d\": [290, 131, 330, 410], \"label\": \"list_item\"},\n    {\"box_2d\": [854, 35, 892, 399], \"label\": \"list_item\"},\n    {\"box_2d\": [578, 468, 660, 937], \"label\": \"list_item\"},\n    {\"box_2d\": [64, 453, 151, 960], \"label\": \"list_item\"},\n    {\"box_2d\": [396, 451, 554, 958], \"label\": \"list_item\"},\n    {\"box_2d\": [575, 131, 665, 371], \"label\": \"list_item\"}\n]\n```\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_145", "subject": "Botany", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_146", "subject": "Botany", "ground_truth": [1], "predicted_answer": [4], "raw_response": "*   **Analyze the question:** The question provides the Net Primary Productivity (NPP) of the first trophic level as 100x (kcal m\u207b\u00b2) yr\u207b\u00b9 and asks for the Gross Primary Productivity (GPP) of the third trophic level.\n*   **Key concepts:**\n    *   GPP (Gross Primary Productivity) is the total rate of photosynthesis by producers (first trophic level).\n    *   NPP (Net Primary Productivity) is GPP minus the energy used by producers for respiration (NPP = GPP - R). NPP is the energy available to the next trophic level (primary consumers).\n    *   Energy transfer between trophic levels is typically around 10% (Lindeman's 10% law).\n    *   GPP and NPP are terms specifically defined for producers (the first trophic level). Higher trophic levels (consumers) acquire energy by consuming other organisms; their energy assimilation and biomass production are referred to as secondary productivity, not GPP or NPP.\n*   **Evaluate the question's premise:** Asking for the GPP of the third trophic level is technically incorrect according to standard ecological definitions, as GPP applies only to producers. Consumers (like those in the third trophic level) do not photosynthesize.\n*   **Interpret the likely intent:** It is highly probable that the question, despite using incorrect terminology (\"GPP of the third trophic level\"), intends to ask for the amount of energy available or assimilated at the third trophic level, assuming the standard 10% energy transfer efficiency between levels, starting from the NPP of the first level.\n*   **Calculate energy transfer:**\n    *   Energy available at the first trophic level (Producers) as NPP = 100x.\n    *   Energy transferred to the second trophic level (Primary Consumers) \u2248 10% of 100x = 0.1 * 100x = 10x.\n    *   Energy transferred to the third trophic level (Secondary Consumers) \u2248 10% of 10x = 0.1 * 10x = x.\n*   **Compare with options:**\n    *   (1) 10x: Energy at the second trophic level.\n    *   (2) 100x/3x = 100/3: Incorrect calculation.\n    *   (3) x/10: Energy potentially at the fourth trophic level.\n    *   (4) x: Energy at the third trophic level based on the 10% rule.\n*   **Conclusion:** Based on the likely intent of the question using the 10% energy transfer rule, the energy available at the third trophic level is approximately x.\n\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_147", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_148", "subject": "Botany", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_149", "subject": "Botany", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n{\n\"A\": \"Monoadelphous\",\n\"B\": \"Diadelphous\",\n\"C\": \"Polyadelphous\",\n\"D\": \"Epiphyllous\",\n\"I\": \"Citrus\",\n\"II\": \"Pea\",\n\"III\": \"Lily\",\n\"IV\": \"China-rose\"\n}\n```\n*   **(A) Monoadelphous:** Filaments united into one bundle. Example: China-rose. So, A matches IV.\n*   **(B) Diadelphous:** Filaments united into two bundles (often 9+1). Example: Pea. So, B matches II.\n*   **(C) Polyadelphous:** Filaments united into more than two bundles. Example: Citrus. So, C matches I.\n*   **(D) Epiphyllous:** Stamens attached to the perianth (tepals). Example: Lily. So, D matches III.\n\nThe correct matching is A-IV, B-II, C-I, D-III.\n\nLet's check the options:\n(1) A-I, B-II, C-IV, D-III\n(2) A-III, B-I, C-IV, D-II\n(3) A-IV, B-II, C-I, D-III\n(4) A-IV, B-I, C-II, D-III\n\nOption (3) matches our derived pairing.\n\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_150", "subject": "Botany", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_151", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_152", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "``` Mismatch\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_153", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```xml\n<answer>1</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_154", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_155", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_156", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_157", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_158", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_159", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_160", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "``` HBox(children=(FloatProgress(value=0.0, bar_style='info', description='Analyzing Image...', max=1.0, style=Pro\u2026\n[\n  {\"box_2d\": [46, 123, 94, 694], \"label\": \"the question\"},\n  {\"box_2d\": [130, 208, 181, 528], \"label\": \"the text A. Myasthenia gravis\"},\n  {\"box_2d\": [189, 208, 249, 507], \"label\": \"the text B. Rheumatoid arthritis\"},\n  {\"box_2d\": [263, 208, 326, 849], \"label\": \"the text C. Gout D. Muscular dystrophy E. Systemic Lupus Erythematosus (SLE) Choose the most appropriate answer from the options given below: (1) B, C & E only (2) C, D & E only (3) A, B & D only (4) A, B & E only\"},\n  {\"box_2d\": [360, 208, 411, 537], \"label\": \"the text D. Muscular dystrophy\"},\n  {\"box_2d\": [440, 208, 491, 772], \"label\": \"the text E. Systemic Lupus Erythematosus (SLE)\"},\n  {\"box_2d\": [520, 129, 654, 990], \"label\": \"the text Choose the most appropriate answer from the options given below:\"},\n  {\"box_2d\": [684, 129, 735, 412], \"label\": \"the text (1) B, C & E only\"},\n  {\"box_2d\": [758, 129, 809, 410], \"label\": \"the text (2) C, D & E only\"},\n  {\"box_2d\": [846, 129, 897, 410], \"label\": \"the text (3) A, B & D only\"},\n  {\"box_2d\": [923, 129, 975, 410], \"label\": \"the text (4) A, B & E only\"}\n]\n```\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_161", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_162", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_163", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "``` Mnemonic:\nEC = TV + ERV\nFRC = ERV + RV\nVC = IRV + TV + ERV\nIC = TV + IRV\n\nMatching:\nA. Expiratory capacity (EC) = TV + ERV -> II\nB. Functional residual capacity (FRC) = ERV + RV -> IV\nC. Vital capacity (VC) = IRV + TV + ERV -> I\nD. Inspiratory capacity (IC) = TV + IRV -> III\n\nCombined: A-II, B-IV, C-I, D-III\n\nChecking options:\n(1) A-II, B-I, C-IV, D-III -> Incorrect\n(2) A-I, B-III, C-II, D-IV -> Incorrect\n(3) A-II, B-IV, C-I, D-III -> Correct\n(4) A-III, B-II, C-IV, D-I -> Incorrect\n\nThe correct option number is 3.\n```\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_164", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_165", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_166", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n```json\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_167", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```json\n<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_168", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_169", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_170", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_171", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_172", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_173", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_174", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n{\n \"A\": \"III\",\n \"B\": \"I\",\n \"C\": \"IV\",\n \"D\": \"II\"\n}\n```\n\n```\nA. Fibrous joints -> III. Skull, don't allow any movement\nB. Cartilaginous joints -> I. Adjacent vertebrae, limited movement\nC. Hinge joints -> IV. Knee, help in locomotion (Hinge joints allow movement in one plane, like the knee or elbow)\nD. Ball and socket joints -> II. Humerus and pectoral girdle, rotational movement (Shoulder and hip joints are examples)\n\nMatching:\nA - III\nB - I\nC - IV\nD - II\n\nCheck options:\n(1) A-II, B-III, C-I, D-IV -> Incorrect\n(2) A-III, B-I, C-IV, D-II -> Correct\n(3) A-IV, B-II, C-III, D-I -> Incorrect\n(4) A-I, B-III, C-II, D-IV -> Incorrect\n\nThe correct option is (2).\n```\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_175", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_176", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_177", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_178", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_179", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_180", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_181", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_182", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_183", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_184", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_185", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n{\n  \"A\": \"Down's syndrome\",\n  \"B\": \"\u03b1 - Thalassemia\",\n  \"C\": \"\u03b2-Thalassemia\",\n  \"D\": \"Klinefelter's\",\n  \"I\": \"11th chromosome\",\n  \"II\": \"'X' chromosome\",\n  \"III\": \"21st chromosome\",\n  \"IV\": \"16th chromosome\"\n}\n```\n*   **A. Down's syndrome:** This is caused by trisomy of the 21st chromosome. So, **A matches III**.\n*   **B. \u03b1 - Thalassemia:** This is typically caused by mutations or deletions in the genes located on the 16th chromosome. So, **B matches IV**.\n*   **C. \u03b2-Thalassemia:** This is typically caused by mutations in the HBB gene on the 11th chromosome. So, **C matches I**.\n*   **D. Klinefelter's syndrome:** This is a sex chromosome aneuploidy in males, characterized by the presence of an extra X chromosome (XXY). So, **D matches II**.\n\nMatching combination: A-III, B-IV, C-I, D-II.\n\nNow check the options:\n(1) A-III, B-IV, C-I, D-II - This matches our findings.\n(2) A-IV, B-I, C-II, D-III - Incorrect.\n(3) A-I, B-II, C-III, D-IV - Incorrect.\n(4) A-II, B-III, C-IV, D-I - Incorrect.\n\nThe correct option is (1).\n\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_186", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_187", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```json\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_188", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n{\n \"reasoning\": \"Statement I states that the cerebral hemispheres are connected by the corpus callosum. This is a well-established fact in neuroanatomy. The corpus callosum is a large, C-shaped nerve fiber bundle beneath the cerebral cortex that stretches across the midline of the brain, connecting the left and right cerebral hemispheres and enabling communication between them. Therefore, Statement I is correct. Statement II states that the brain stem consists of the medulla oblongata, pons, and cerebrum. The brain stem is typically defined as consisting of the midbrain, pons, and medulla oblongata. The cerebrum is the largest part of the brain, responsible for higher cognitive functions, and it is structurally distinct from the brain stem. Therefore, Statement II is incorrect as it incorrectly includes the cerebrum as part of the brain stem. Based on this analysis, Statement I is correct and Statement II is incorrect. Option (1) reflects this conclusion.\",\n \"correct_option_number\": 1\n}\n```\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_189", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_190", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "```cpp\n<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_191", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_192", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n{\n \"A\": \"Unicellular glandular epithelium\",\n \"B\": \"Compound epithelium\",\n \"C\": \"Multicellular glandular epithelium\",\n \"D\": \"Endocrine glandular epithelium\",\n \"I\": \"Salivary glands\",\n \"II\": \"Pancreas\",\n \"III\": \"Goblet cells of alimentary canal\",\n \"IV\": \"Moist surface of buccal cavity\"\n}\n```\n\n1.  **A. Unicellular glandular epithelium:** Goblet cells are unicellular glands found in the epithelium of the alimentary canal and respiratory tract. They secrete mucus. Therefore, A matches with III.\n2.  **B. Compound epithelium:** This epithelium consists of more than one layer of cells. It provides protection against mechanical and chemical stresses. The moist surface of the buccal cavity (lining of the mouth) is lined by stratified squamous non-keratinized epithelium, which is a type of compound epithelium. Therefore, B matches with IV.\n3.  **C. Multicellular glandular epithelium:** These glands are formed by clusters of cells. Salivary glands are examples of multicellular exocrine glands. Therefore, C matches with I.\n4.  **D. Endocrine glandular epithelium:** Endocrine glands are ductless glands that secrete hormones directly into the bloodstream. The pancreas has both endocrine (Islets of Langerhans secreting insulin, glucagon) and exocrine functions. Thus, the endocrine glandular epithelium is represented by the pancreas (in its endocrine capacity). Therefore, D matches with II.\n\nMatching pairs:\nA - III\nB - IV\nC - I\nD - II\n\nNow let's check the options:\n(1) A-III, B-IV, C-I, D-II - This matches our findings.\n(2) A-II, B-I, C-IV, D-III\n(3) A-II, B-I, C-III, D-IV\n(4) A-IV, B-III, C-I, D-II\n\nThe correct option is (1).\n\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_193", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [1], "raw_response": "<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}
+{"question_id": "NEET_2024_T3_194", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_195", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [1], "raw_response": "```json\n<answer>1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_196", "subject": "Zoology", "ground_truth": [4], "predicted_answer": [4], "raw_response": "```xml\n<answer>4</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_197", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_198", "subject": "Zoology", "ground_truth": [2], "predicted_answer": [2], "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_199", "subject": "Zoology", "ground_truth": [3], "predicted_answer": [3], "raw_response": "```json\n<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "correct", "marks_awarded": 4}
+{"question_id": "NEET_2024_T3_200", "subject": "Zoology", "ground_truth": [1], "predicted_answer": [3], "raw_response": "```xml\n<answer>3</answer>\n```", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "evaluation_status": "incorrect", "marks_awarded": -1}

results/google_gemini-2.5-pro-preview-03-25_NEET_2024_20250505_091752/summary.json

@@ -0,0 +1,62 @@
+{
+    "model_name": "google/gemini-2.5-pro-preview-03-25",
+    "timestamp": "20250505_091752",
+    "total_questions_in_dataset": 200,
+    "overall_score": 727,
+    "overall_correct": 185,
+    "overall_incorrect": 13,
+    "overall_skipped": 2,
+    "overall_api_parse_failures": 0,
+    "total_questions_processed": 200,
+    "unmapped_section_questions": 0,
+    "section_breakdown": {
+        "Physics_A": {
+            "score": 130,
+            "correct": 33,
+            "incorrect": 2,
+            "skipped": 0
+        },
+        "Physics_B": {
+            "score": 47,
+            "correct": 12,
+            "incorrect": 1,
+            "skipped": 2
+        },
+        "Chemistry_A": {
+            "score": 140,
+            "correct": 35,
+            "incorrect": 0,
+            "skipped": 0
+        },
+        "Chemistry_B": {
+            "score": 55,
+            "correct": 14,
+            "incorrect": 1,
+            "skipped": 0
+        },
+        "Botany_A": {
+            "score": 120,
+            "correct": 31,
+            "incorrect": 4,
+            "skipped": 0
+        },
+        "Botany_B": {
+            "score": 55,
+            "correct": 14,
+            "incorrect": 1,
+            "skipped": 0
+        },
+        "Zoology_A": {
+            "score": 130,
+            "correct": 33,
+            "incorrect": 2,
+            "skipped": 0
+        },
+        "Zoology_B": {
+            "score": 50,
+            "correct": 13,
+            "incorrect": 2,
+            "skipped": 0
+        }
+    }
+}

results/google_gemini-2.5-pro-preview-03-25_NEET_2024_20250505_091752/summary.md

@@ -0,0 +1,25 @@
+# Benchmark Results: google/gemini-2.5-pro-preview-03-25
+**Timestamp:** 20250505_091752
+**Total Questions in Dataset:** 200
+
+---
+
+## NEET Scoring Results
+**Overall Score:** **727 / 800**
+- **Correct Answers (+4):** 185
+- **Incorrect Answers (-1):** 13
+- **Skipped Questions (0):** 2
+- **API/Parse Failures (-1):** 0
+- **Total Questions Processed:** 200
+
+### Section Breakdown
+| Section       | Score | Correct | Incorrect | Skipped |
+|---------------|-------|---------|-----------|---------|
+| Physics A     | 130   | 33      | 2         | 0       |
+| Physics B     | 47    | 12      | 1         | 2       |
+| Chemistry A   | 140   | 35      | 0         | 0       |
+| Chemistry B   | 55    | 14      | 1         | 0       |
+| Botany A      | 120   | 31      | 4         | 0       |
+| Botany B      | 55    | 14      | 1         | 0       |
+| Zoology A     | 130   | 33      | 2         | 0       |
+| Zoology B     | 50    | 13      | 2         | 0       |

results/google_gemini-2.5-pro-preview-03-25_NEET_2025_20250506_195634/summary.json

@@ -0,0 +1,34 @@
+{
+    "model_name": "google/gemini-2.5-pro-preview-03-25",
+    "exam_name": "NEET",
+    "exam_year": "2025",
+    "timestamp": "20250506_195634",
+    "total_questions_in_dataset": 180,
+    "overall_score": 685,
+    "overall_correct": 173,
+    "overall_incorrect": 7,
+    "overall_skipped": 0,
+    "overall_api_parse_failures": 6,
+    "total_questions_processed": 180,
+    "unmapped_section_questions": 0,
+    "section_breakdown": {
+        "Biology": {
+            "score": 340,
+            "correct": 87,
+            "incorrect": 3,
+            "skipped": 0
+        },
+        "Chemistry": {
+            "score": 170,
+            "correct": 43,
+            "incorrect": 2,
+            "skipped": 0
+        },
+        "Physics": {
+            "score": 170,
+            "correct": 43,
+            "incorrect": 2,
+            "skipped": 0
+        }
+    }
+}

results/google_gemini-2.5-pro-preview-03-25_NEET_2025_20250506_195634/summary.md

@@ -0,0 +1,22 @@
+# Benchmark Results: google/gemini-2.5-pro-preview-03-25
+**Exam Name:** NEET
+**Exam Year:** 2025
+**Timestamp:** 20250506_195634
+**Total Questions in Dataset:** 180
+
+---
+
+## NEET Scoring Results
+**Overall Score:** **685 / 720**
+- **Correct Answers (+4):** 173
+- **Incorrect Answers (-1):** 7
+- **Skipped Questions (0):** 0
+- **API/Parse Failures (-1):** 6
+- **Total Questions Processed:** 180
+
+### Section Breakdown
+| Section       | Score | Correct | Incorrect | Skipped |
+|---------------|-------|---------|-----------|---------|
+| Biology       | 340   | 87      | 3         | 0       |
+| Chemistry     | 170   | 43      | 2         | 0       |
+| Physics       | 170   | 43      | 2         | 0       |

results/openai_gpt-5_JEE_ADVANCED_2025_20250923_174647/predictions.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P02", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P03", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P04", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P05", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P06", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P07", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P08", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P09", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>23</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P10", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P11", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.5</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P12", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>38</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P13", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>72</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P14", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P15", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P16", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C01", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C02", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C03", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C04", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C05", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C06", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C07", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C08", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>100</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C09", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.236</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C10", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-7.1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C11", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>30</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C12", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>280</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C13", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C14", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C15", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C16", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M01", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M02", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M03", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M04", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M05", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M06", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M07", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M08", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>105</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M09", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "previous_raw_response_on_reprompt": "<answer>6/5</answer>"}
+{"question_id": "JA25P1M10", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>762</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M11", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M12", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>96</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M13", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M14", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M15", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M16", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P01", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P02", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P03", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P04", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": null, "parse_successful": false, "api_call_successful": true, "error": "Initial API call returned empty content. Re-prompt skipped.", "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P05", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P06", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P07", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P08", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P09", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>5/3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "previous_raw_response_on_reprompt": "<answer>5/3</answer>"}
+{"question_id": "JA25P2P10", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>11.8</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P11", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>1.6</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P12", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.331</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P13", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P14", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>4.8</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P15", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>170</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P16", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>32</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C01", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C02", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C03", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C04", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C05", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>C,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C06", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C07", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C08", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C09", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>11</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C10", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C11", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>16</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C12", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>4.17</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C13", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.49</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C14", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>105.5</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C15", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>7.73</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C16", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M01", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M02", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M03", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M04", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M05", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M06", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M07", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M08", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,C,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M09", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.75</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M10", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>6</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M11", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M12", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M13", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M14", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.25</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M15", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M16", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>21</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}

results/openai_gpt-5_JEE_ADVANCED_2025_20250923_174647/summary.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P1P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1P08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1P09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["23"], "ground_truth": [["21", "25"]], "attempt": 1}
+{"question_id": "JA25P1P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P1P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.5"], "ground_truth": ["0.5", "0.75"], "attempt": 1}
+{"question_id": "JA25P1P12", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["38"], "ground_truth": [["75", "79"]], "attempt": 1}
+{"question_id": "JA25P1P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["72"], "ground_truth": ["72"], "attempt": 1}
+{"question_id": "JA25P1P14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P15", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P1C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P1C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["100"], "ground_truth": ["100"], "attempt": 1}
+{"question_id": "JA25P1C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.236"], "ground_truth": [["2.2", "2.3"]], "attempt": 1}
+{"question_id": "JA25P1C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-7.1"], "ground_truth": [["-7.2", "-7"]], "attempt": 1}
+{"question_id": "JA25P1C11", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["30"], "ground_truth": [["-29.95", "-29.8"], ["29.8", "29.95"]], "attempt": 1}
+{"question_id": "JA25P1C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["280"], "ground_truth": ["280"], "attempt": 1}
+{"question_id": "JA25P1C13", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": ["175"], "attempt": 1}
+{"question_id": "JA25P1C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P1M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105"], "ground_truth": ["105"], "attempt": 1}
+{"question_id": "JA25P1M09", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": [["1.15", "1.25"]], "attempt": 2}
+{"question_id": "JA25P1M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["762"], "ground_truth": ["762"], "attempt": 1}
+{"question_id": "JA25P1M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.4"], "ground_truth": [["2.35", "2.45"]], "attempt": 1}
+{"question_id": "JA25P1M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["96"], "ground_truth": ["96"], "attempt": 1}
+{"question_id": "JA25P1M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P04", "marks_awarded": 0, "evaluation_status": "failure_api_or_parse", "predicted_answer": null, "ground_truth": ["A", "B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": "5/3", "ground_truth": [["1.65", "1.67"]], "attempt": 2}
+{"question_id": "JA25P2P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11.8"], "ground_truth": [["11.7", "11.9"]], "attempt": 1}
+{"question_id": "JA25P2P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.6"], "ground_truth": ["1.6"], "attempt": 1}
+{"question_id": "JA25P2P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.331"], "ground_truth": [["2.3", "2.4"]], "attempt": 1}
+{"question_id": "JA25P2P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.2"], "ground_truth": ["0.2"], "attempt": 1}
+{"question_id": "JA25P2P14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["4.8"], "ground_truth": ["1.2"], "attempt": 1}
+{"question_id": "JA25P2P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["170"], "ground_truth": [["167", "171"]], "attempt": 1}
+{"question_id": "JA25P2P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["32"], "ground_truth": [["26", "33"]], "attempt": 1}
+{"question_id": "JA25P2C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P2C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["C", "D"], "ground_truth": ["C", "D"], "attempt": 1}
+{"question_id": "JA25P2C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P2C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2C08", "marks_awarded": -2, "evaluation_status": "incorrect_negative", "predicted_answer": ["A", "B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P2C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11"], "ground_truth": [["10.85", "11.1"]], "attempt": 1}
+{"question_id": "JA25P2C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": [["3.85", "4.15"]], "attempt": 1}
+{"question_id": "JA25P2C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["16"], "ground_truth": [["15.5", "16.5"]], "attempt": 1}
+{"question_id": "JA25P2C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4.17"], "ground_truth": [["4", "4.25"]], "attempt": 1}
+{"question_id": "JA25P2C13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.49"], "ground_truth": [["2.4", "2.55"]], "attempt": 1}
+{"question_id": "JA25P2C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105.5"], "ground_truth": [["105.4", "105.6"]], "attempt": 1}
+{"question_id": "JA25P2C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["7.73"], "ground_truth": [["7.5", "7.8"]], "attempt": 1}
+{"question_id": "JA25P2C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P2M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C", "D"], "ground_truth": ["B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.75"], "ground_truth": [["0.7", "0.8"]], "attempt": 1}
+{"question_id": "JA25P2M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["6"], "ground_truth": ["6"], "attempt": 1}
+{"question_id": "JA25P2M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.3"], "ground_truth": [["0.27", "0.33"]], "attempt": 1}
+{"question_id": "JA25P2M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.25"], "ground_truth": [["0.2", "0.3"]], "attempt": 1}
+{"question_id": "JA25P2M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P2M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["21"], "ground_truth": ["21"], "attempt": 1}

results/openai_gpt-5_JEE_ADVANCED_2025_20250923_174647/summary.md

@@ -0,0 +1,32 @@
+# Benchmark Results: openai/gpt-5
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2025
+**Timestamp:** 20250923_174647
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 482
+**Total Questions Processed in this Run:** 96
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **318** / **360**
+- **Fully Correct Answers:** 85
+- **Incorrectly Answered (Choice Made):** 7
+- **Skipped Questions:** 3
+- **API/Parse Failures:** 1
+- **Total Questions Processed:** 96
+
+### Detailed Score Calculation by Question Type
+**Integer (42 questions):** 144 marks
+  *Calculation:* 38 Correct (+4) + 2 Incorrect (0) + 2 Skipped (0) = 148
+**Mcq Multiple Correct (21 questions):** 73 marks
+  *Calculation:* 20 Correct (+4) + 1 Incorrect (-2) + 0 API/Parse Fail (-1) = 78
+**Mcq Single Correct (33 questions):** 83 marks
+  *Calculation:* 29 Correct (+3) + 3 Incorrect (-1) + 1 API/Parse Fail (-1) = 83 (+8 due to some special questions)
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 102    | 28            | 0                 | 3                | 1       | 0                  |
+| Math          | 116   | 31            | 0                 | 0                | 1       | 0                  |
+| Physics       | 100    | 26            | 0                 | 4                | 1       | 1                  |

results/openai_o3_JEE_ADVANCED_2025_20250630_175202/predictions.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P02", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P03", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P04", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P05", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P06", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P07", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P08", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P09", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>16</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P10", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P11", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.5</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P12", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>76</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P13", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>72</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P14", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P15", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1P16", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C01", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C02", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C03", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C04", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C05", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C06", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C07", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C08", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>100</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C09", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.24</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C10", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-7.1</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C11", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>30</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C12", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>280</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C13", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>332</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C14", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C15", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1C16", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M01", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M02", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M03", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M04", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M05", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M06", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M07", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M08", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>105</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M09", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>1.2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M10", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>762</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M11", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M12", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>96</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M13", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M14", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M15", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P1M16", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P01", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P02", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P03", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P04", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P05", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P06", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P07", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P08", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P09", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>SKIP</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 2, "previous_raw_response_on_reprompt": "<answer>5/3</answer>"}
+{"question_id": "JA25P2P10", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>11.8</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P11", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>1.6</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P12", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.331</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P13", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P14", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P15", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>170</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2P16", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>32</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C01", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C02", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C03", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C04", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C05", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>C,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C06", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C07", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C08", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C09", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>11</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C10", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>4</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C11", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>16</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C12", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": null, "parse_successful": false, "api_call_successful": true, "error": "Initial API call returned empty content. Re-prompt skipped.", "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C13", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2.49</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C14", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>105.5</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C15", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>7.73</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2C16", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M01", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M02", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M03", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M04", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "raw_response": "<answer>A</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M05", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,B</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M06", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M07", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>A,C</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M08", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "raw_response": "<answer>B,C,D</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M09", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.75</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M10", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>6</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M11", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M12", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M13", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>-2</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M14", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>0.25</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M15", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>3</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}
+{"question_id": "JA25P2M16", "subject": "Math", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "raw_response": "<answer>21</answer>", "parse_successful": true, "api_call_successful": true, "error": null, "attempt": 1, "previous_raw_response_on_reprompt": null}

results/openai_o3_JEE_ADVANCED_2025_20250630_175202/summary.jsonl

@@ -0,0 +1,96 @@
+{"question_id": "JA25P1P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P03", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["D"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["A"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P1P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P1P07", "marks_awarded": 1, "evaluation_status": "partial_1_of_2_plus", "predicted_answer": ["D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1P08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1P09", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["16"], "ground_truth": [["21", "25"]], "attempt": 1}
+{"question_id": "JA25P1P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P1P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.5"], "ground_truth": ["0.5", "0.75"], "attempt": 1}
+{"question_id": "JA25P1P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["76"], "ground_truth": [["75", "79"]], "attempt": 1}
+{"question_id": "JA25P1P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["72"], "ground_truth": ["72"], "attempt": 1}
+{"question_id": "JA25P1P14", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1P15", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C04", "marks_awarded": -1, "evaluation_status": "incorrect", "predicted_answer": ["C"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P1C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P1C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["100"], "ground_truth": ["100"], "attempt": 1}
+{"question_id": "JA25P1C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.24"], "ground_truth": [["2.2", "2.3"]], "attempt": 1}
+{"question_id": "JA25P1C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-7.1"], "ground_truth": [["-7.2", "-7"]], "attempt": 1}
+{"question_id": "JA25P1C11", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["30"], "ground_truth": [["-29.95", "-29.8"], ["29.8", "29.95"]], "attempt": 1}
+{"question_id": "JA25P1C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["280"], "ground_truth": ["280"], "attempt": 1}
+{"question_id": "JA25P1C13", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["332"], "ground_truth": ["175"], "attempt": 1}
+{"question_id": "JA25P1C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P1M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P1M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "D"], "ground_truth": ["A", "D"], "attempt": 1}
+{"question_id": "JA25P1M08", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105"], "ground_truth": ["105"], "attempt": 1}
+{"question_id": "JA25P1M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.2"], "ground_truth": [["1.15", "1.25"]], "attempt": 1}
+{"question_id": "JA25P1M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["762"], "ground_truth": ["762"], "attempt": 1}
+{"question_id": "JA25P1M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.4"], "ground_truth": [["2.35", "2.45"]], "attempt": 1}
+{"question_id": "JA25P1M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["96"], "ground_truth": ["96"], "attempt": 1}
+{"question_id": "JA25P1M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P1M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P1M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P1M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2P02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2P04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["A", "B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2P05", "marks_awarded": 2, "evaluation_status": "partial_2_of_3_plus", "predicted_answer": ["A", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2P07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2P08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B", "C"], "ground_truth": ["A", "B", "C"], "attempt": 1}
+{"question_id": "JA25P2P09", "marks_awarded": 0, "evaluation_status": "skipped", "predicted_answer": "SKIP", "ground_truth": [["1.65", "1.67"]], "attempt": 2}
+{"question_id": "JA25P2P10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11.8"], "ground_truth": [["11.7", "11.9"]], "attempt": 1}
+{"question_id": "JA25P2P11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["1.6"], "ground_truth": ["1.6"], "attempt": 1}
+{"question_id": "JA25P2P12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.331"], "ground_truth": [["2.3", "2.4"]], "attempt": 1}
+{"question_id": "JA25P2P13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.2"], "ground_truth": ["0.2"], "attempt": 1}
+{"question_id": "JA25P2P14", "marks_awarded": 0, "evaluation_status": "incorrect", "predicted_answer": ["2.4"], "ground_truth": ["1.2"], "attempt": 1}
+{"question_id": "JA25P2P15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["170"], "ground_truth": [["167", "171"]], "attempt": 1}
+{"question_id": "JA25P2P16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["32"], "ground_truth": [["26", "33"]], "attempt": 1}
+{"question_id": "JA25P2C01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2C03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["D"], "ground_truth": ["D"], "attempt": 1}
+{"question_id": "JA25P2C04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2C05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["C", "D"], "ground_truth": ["C", "D"], "attempt": 1}
+{"question_id": "JA25P2C06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "D"], "ground_truth": ["B", "D"], "attempt": 1}
+{"question_id": "JA25P2C07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2C08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C"], "ground_truth": ["B", "C"], "attempt": 1}
+{"question_id": "JA25P2C09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["11"], "ground_truth": [["10.85", "11.1"]], "attempt": 1}
+{"question_id": "JA25P2C10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4"], "ground_truth": [["3.85", "4.15"]], "attempt": 1}
+{"question_id": "JA25P2C11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["16"], "ground_truth": [["15.5", "16.5"]], "attempt": 1}
+{"question_id": "JA25P2C12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["4.2"], "ground_truth": [["4", "4.25"]], "attempt": 1}
+{"question_id": "JA25P2C13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2.49"], "ground_truth": [["2.4", "2.55"]], "attempt": 1}
+{"question_id": "JA25P2C14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["105.5"], "ground_truth": [["105.4", "105.6"]], "attempt": 1}
+{"question_id": "JA25P2C15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["7.73"], "ground_truth": [["7.5", "7.8"]], "attempt": 1}
+{"question_id": "JA25P2C16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["2"], "ground_truth": ["2"], "attempt": 1}
+{"question_id": "JA25P2M01", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M02", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["B"], "ground_truth": ["B"], "attempt": 1}
+{"question_id": "JA25P2M03", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["C"], "ground_truth": ["C"], "attempt": 1}
+{"question_id": "JA25P2M04", "marks_awarded": 3, "evaluation_status": "correct", "predicted_answer": ["A"], "ground_truth": ["A"], "attempt": 1}
+{"question_id": "JA25P2M05", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "B"], "ground_truth": ["A", "B"], "attempt": 1}
+{"question_id": "JA25P2M06", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M07", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["A", "C"], "ground_truth": ["A", "C"], "attempt": 1}
+{"question_id": "JA25P2M08", "marks_awarded": 4, "evaluation_status": "correct_full", "predicted_answer": ["B", "C", "D"], "ground_truth": ["B", "C", "D"], "attempt": 1}
+{"question_id": "JA25P2M09", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.75"], "ground_truth": [["0.7", "0.8"]], "attempt": 1}
+{"question_id": "JA25P2M10", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["6"], "ground_truth": ["6"], "attempt": 1}
+{"question_id": "JA25P2M11", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.3"], "ground_truth": [["0.27", "0.33"]], "attempt": 1}
+{"question_id": "JA25P2M12", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M13", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["-2"], "ground_truth": ["-2"], "attempt": 1}
+{"question_id": "JA25P2M14", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["0.25"], "ground_truth": [["0.2", "0.3"]], "attempt": 1}
+{"question_id": "JA25P2M15", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["3"], "ground_truth": ["3"], "attempt": 1}
+{"question_id": "JA25P2M16", "marks_awarded": 4, "evaluation_status": "correct", "predicted_answer": ["21"], "ground_truth": ["21"], "attempt": 1}

results/openai_o3_JEE_ADVANCED_2025_20250630_175202/summary.md

@@ -0,0 +1,33 @@
+# Benchmark Results: openai/o3
+**Exam Name:** JEE_ADVANCED
+**Exam Year:** 2025
+**Timestamp:** 20250630_175202
+**Total Questions in Dataset:** 578
+**Questions Filtered Out:** 482
+**Total Questions Processed in this Run:** 96
+
+---
+
+## Exam Scoring Results
+**Overall Score:** **315** / **360**
+- **Fully Correct Answers:** 84
+- **Partially Correct Answers:** 2
+- **Incorrectly Answered (Choice Made):** 7
+- **Skipped Questions:** 3
+- **API/Parse Failures:** 0
+- **Total Questions Processed:** 96
+
+### Detailed Score Calculation by Question Type
+**Integer (42 questions):** 144 marks
+  *Calculation:* 37 Correct (+4) + 4 Incorrect (0) + 1 Skipped (0) + 0 API/Parse Fail (0) = 148
+**Mcq Multiple Correct (21 questions):** 79 marks
+  *Calculation:* 19 Correct (+4) + 2 Partial = 79
+**Mcq Single Correct (33 questions):** 88 marks
+  *Calculation:* 28 Correct (+3) + 3 Incorrect (-1) + 2 Skipped (0) = 81(+7 some questions have full marks 4)
+
+### Section Breakdown
+| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |
+|---------------|-------|---------------|-------------------|------------------|---------|--------------------|
+| Chemistry     | 108   | 29            | 0                 | 3                | 0       | 0                  |
+| Math          | 120   | 32            | 0                 | 0                | 0       | 0                  |
+| Physics       | 87    | 23            | 2                 | 4                | 3       | 0                  |

scripts/aggregate_runs.py

@@ -0,0 +1,210 @@
+"""Aggregate results from multiple benchmark runs of the same model/exam/year.
+
+Computes mean, std dev, 95% CI, min/max scores, and per-question stability.
+
+Usage:
+    uv run python scripts/aggregate_runs.py results/model_EXAM_YEAR_*/
+    uv run python scripts/aggregate_runs.py --pattern "openai_o3_JEE_ADVANCED_2025"
+"""
+
+import argparse
+import glob
+import json
+import math
+import os
+import sys
+from collections import defaultdict
+
+
+def load_summary_jsonl(path: str) -> list[dict]:
+    """Load all records from a summary.jsonl file."""
+    records = []
+    with open(path, "r") as f:
+        for line in f:
+            line = line.strip()
+            if line:
+                records.append(json.loads(line))
+    return records
+
+
+def compute_run_stats(records: list[dict]) -> dict:
+    """Compute aggregate stats for a single run."""
+    total_score = sum(r.get("marks_awarded", 0) for r in records)
+    correct = sum(1 for r in records if r.get("evaluation_status") in ("correct", "correct_full"))
+    incorrect = sum(1 for r in records if r.get("evaluation_status") in ("incorrect", "incorrect_negative"))
+    skipped = sum(1 for r in records if r.get("evaluation_status") == "skipped")
+    failures = sum(1 for r in records if r.get("evaluation_status") in ("failure_api_or_parse", "failure_unexpected_type"))
+    partial = sum(1 for r in records if r.get("evaluation_status", "").startswith("partial_"))
+    return {
+        "score": total_score,
+        "correct": correct,
+        "partial": partial,
+        "incorrect": incorrect,
+        "skipped": skipped,
+        "failures": failures,
+        "num_questions": len(records),
+    }
+
+
+def aggregate_runs(run_dirs: list[str]) -> dict:
+    """Aggregate stats across multiple runs."""
+    if not run_dirs:
+        return {"error": "No run directories provided."}
+
+    all_run_stats = []
+    per_question_answers: dict[str, list] = defaultdict(list)
+
+    for run_dir in run_dirs:
+        summary_path = os.path.join(run_dir, "summary.jsonl")
+        if not os.path.exists(summary_path):
+            print(f"Warning: {summary_path} not found, skipping.", file=sys.stderr)
+            continue
+
+        records = load_summary_jsonl(summary_path)
+        stats = compute_run_stats(records)
+        stats["run_dir"] = run_dir
+        all_run_stats.append(stats)
+
+        for r in records:
+            qid = r.get("question_id")
+            if qid:
+                per_question_answers[qid].append({
+                    "predicted_answer": r.get("predicted_answer"),
+                    "evaluation_status": r.get("evaluation_status"),
+                    "marks_awarded": r.get("marks_awarded", 0),
+                })
+
+    if not all_run_stats:
+        return {"error": "No valid runs found."}
+
+    scores = [s["score"] for s in all_run_stats]
+    n = len(scores)
+    mean_score = sum(scores) / n
+    if n > 1:
+        variance = sum((s - mean_score) ** 2 for s in scores) / (n - 1)
+        std_dev = math.sqrt(variance)
+        # 95% CI using t-distribution approximation (for small n, use 2.0 as rough multiplier)
+        t_multiplier = 2.0 if n < 30 else 1.96
+        ci_half = t_multiplier * std_dev / math.sqrt(n)
+    else:
+        std_dev = 0.0
+        ci_half = 0.0
+
+    # Per-question stability
+    question_stability = {}
+    for qid, answers in per_question_answers.items():
+        statuses = [a["evaluation_status"] for a in answers]
+        most_common = max(set(statuses), key=statuses.count)
+        agreement_rate = statuses.count(most_common) / len(statuses)
+        question_stability[qid] = {
+            "agreement_rate": round(agreement_rate, 3),
+            "dominant_status": most_common,
+            "all_statuses": statuses,
+            "scores": [a["marks_awarded"] for a in answers],
+        }
+
+    # Unstable questions (agreement < 100%)
+    unstable = {qid: v for qid, v in question_stability.items() if v["agreement_rate"] < 1.0}
+
+    return {
+        "num_runs": n,
+        "scores": scores,
+        "mean_score": round(mean_score, 2),
+        "std_dev": round(std_dev, 2),
+        "ci_95_lower": round(mean_score - ci_half, 2),
+        "ci_95_upper": round(mean_score + ci_half, 2),
+        "min_score": min(scores),
+        "max_score": max(scores),
+        "per_run": all_run_stats,
+        "num_questions": all_run_stats[0]["num_questions"] if all_run_stats else 0,
+        "unstable_questions": len(unstable),
+        "total_questions": len(question_stability),
+        "question_stability": question_stability,
+    }
+
+
+def print_report(result: dict):
+    """Print a human-readable report."""
+    if "error" in result:
+        print(f"Error: {result['error']}")
+        return
+
+    print(f"# Multi-Run Aggregation Report")
+    print(f"\n**Runs:** {result['num_runs']}")
+    print(f"**Questions per run:** {result['num_questions']}")
+    print(f"\n## Score Summary")
+    print(f"| Metric | Value |")
+    print(f"|--------|-------|")
+    print(f"| Mean | {result['mean_score']} |")
+    print(f"| Std Dev | {result['std_dev']} |")
+    print(f"| 95% CI | [{result['ci_95_lower']}, {result['ci_95_upper']}] |")
+    print(f"| Min | {result['min_score']} |")
+    print(f"| Max | {result['max_score']} |")
+    print(f"| Scores | {result['scores']} |")
+
+    print(f"\n## Per-Question Stability")
+    print(f"- **Stable questions:** {result['total_questions'] - result['unstable_questions']}/{result['total_questions']}")
+    print(f"- **Unstable questions:** {result['unstable_questions']}/{result['total_questions']}")
+
+    if result["unstable_questions"] > 0:
+        print(f"\n### Unstable Questions (different answers across runs)")
+        print(f"| Question ID | Agreement | Dominant Status | Scores |")
+        print(f"|-------------|-----------|-----------------|--------|")
+        for qid, info in sorted(result["question_stability"].items()):
+            if info["agreement_rate"] < 1.0:
+                print(f"| {qid} | {info['agreement_rate']:.0%} | {info['dominant_status']} | {info['scores']} |")
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Aggregate results from multiple benchmark runs."
+    )
+    parser.add_argument(
+        "run_dirs",
+        nargs="*",
+        help="Paths to result directories to aggregate.",
+    )
+    parser.add_argument(
+        "--pattern",
+        type=str,
+        help="Glob pattern to match result directories (e.g., 'openai_o3_JEE_ADVANCED_2025').",
+    )
+    parser.add_argument(
+        "--results-dir",
+        type=str,
+        default="results",
+        help="Base results directory (default: results).",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        help="Output JSON file path for aggregated results.",
+    )
+    args = parser.parse_args()
+
+    run_dirs = list(args.run_dirs) if args.run_dirs else []
+
+    if args.pattern:
+        pattern = os.path.join(args.results_dir, f"*{args.pattern}*")
+        matched = sorted(glob.glob(pattern))
+        run_dirs.extend(d for d in matched if os.path.isdir(d))
+
+    if not run_dirs:
+        parser.error("No run directories specified. Provide paths or use --pattern.")
+
+    run_dirs = sorted(set(run_dirs))
+    print(f"Aggregating {len(run_dirs)} runs...", file=sys.stderr)
+
+    result = aggregate_runs(run_dirs)
+    print_report(result)
+
+    if args.output:
+        # Remove per-question detail for compact output unless needed
+        output_data = {k: v for k, v in result.items() if k != "question_stability"}
+        with open(args.output, "w") as f:
+            json.dump(output_data, f, indent=2)
+        print(f"\nResults saved to {args.output}", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/generate_leaderboard.py

@@ -0,0 +1,276 @@
+"""Generate a cross-model leaderboard from benchmark results.
+
+Reads all result directories, computes scores, and outputs a Markdown
+leaderboard table grouped by exam+year.
+
+Usage:
+    uv run python scripts/generate_leaderboard.py
+    uv run python scripts/generate_leaderboard.py --results-dir results --min-questions 10
+    uv run python scripts/generate_leaderboard.py --output leaderboard.json
+"""
+
+import argparse
+import json
+import os
+import re
+import sys
+from collections import defaultdict
+
+
+# Known exam names that may contain underscores
+KNOWN_EXAMS = {"JEE_ADVANCED", "JEE_MAIN", "NEET"}
+
+
+def parse_result_dirname(dirname: str) -> dict | None:
+    """Parse a result directory name into model, exam, year, timestamp.
+
+    Format: {model_id}_{EXAM_NAME}_{YEAR}_{YYYYMMDD}_{HHMMSS}
+    where model_id has slashes replaced with underscores.
+
+    Returns dict with keys: model, exam, year, timestamp, or None if unparseable.
+    """
+    parts = dirname.split("_")
+    if len(parts) < 5:
+        return None
+
+    # Last 2 parts are timestamp
+    timestamp = "_".join(parts[-2:])
+
+    # Validate timestamp format
+    if not (len(parts[-2]) == 8 and parts[-2].isdigit() and
+            len(parts[-1]) == 6 and parts[-1].isdigit()):
+        return None
+
+    # Year is parts[-3]
+    year = parts[-3]
+    if not (len(year) == 4 and year.isdigit()):
+        return None
+
+    # Exam name: check if it's a 2-part exam (JEE_ADVANCED, JEE_MAIN)
+    if len(parts) >= 6 and parts[-4] in ("ADVANCED", "MAIN"):
+        exam = parts[-5] + "_" + parts[-4]
+        model_parts = parts[:-5]
+    else:
+        exam = parts[-4]
+        model_parts = parts[:-4]
+
+    if exam not in KNOWN_EXAMS:
+        return None
+
+    if not model_parts:
+        return None
+
+    # Reconstruct model name: first underscore -> slash (provider/model)
+    model_name = "_".join(model_parts)
+    first_underscore = model_name.find("_")
+    if first_underscore > 0:
+        model_name = model_name[:first_underscore] + "/" + model_name[first_underscore + 1:]
+
+    return {
+        "model": model_name,
+        "exam": exam,
+        "year": year,
+        "timestamp": timestamp,
+    }
+
+
+def load_summary_jsonl(filepath: str) -> list[dict]:
+    """Load records from a summary.jsonl file."""
+    records = []
+    with open(filepath, "r") as f:
+        for line in f:
+            line = line.strip()
+            if line:
+                records.append(json.loads(line))
+    return records
+
+
+def extract_max_score_from_md(filepath: str) -> int | None:
+    """Extract max possible score from summary.md's Overall Score line.
+
+    Handles both formats:
+    - New: **Overall Score:** **315** / **360**
+    - Old: **Overall Score:** **727 / 800**
+    """
+    try:
+        with open(filepath, "r") as f:
+            for line in f:
+                # New format: **315** / **360**
+                match = re.search(r"\*\*Overall Score:\*\*\s+\*\*\d+\*\*\s*/\s*\*\*(\d+)\*\*", line)
+                if match:
+                    return int(match.group(1))
+                # Old format: **727 / 800**
+                match = re.search(r"\*\*Overall Score:\*\*\s+\*\*\d+\s*/\s*(\d+)\*\*", line)
+                if match:
+                    return int(match.group(1))
+                # Oldest format: **322** (Max score is 360)
+                match = re.search(r"\(Max score is (\d+)\)", line)
+                if match:
+                    return int(match.group(1))
+    except (OSError, ValueError):
+        pass
+    return None
+
+
+def compute_stats(records: list[dict]) -> dict:
+    """Compute aggregate stats from summary.jsonl records."""
+    total_score = sum(r.get("marks_awarded", 0) for r in records)
+    correct = sum(1 for r in records if r.get("evaluation_status") in ("correct", "correct_full"))
+    partial = sum(1 for r in records if r.get("evaluation_status", "").startswith("partial_"))
+    incorrect = sum(1 for r in records if r.get("evaluation_status") in ("incorrect", "incorrect_negative"))
+    skipped = sum(1 for r in records if r.get("evaluation_status") == "skipped")
+    failures = sum(1 for r in records if r.get("evaluation_status") in (
+        "failure_api_or_parse", "failure_unexpected_type", "error_bad_ground_truth"))
+
+    return {
+        "score": total_score,
+        "correct": correct,
+        "partial": partial,
+        "incorrect": incorrect,
+        "skipped": skipped,
+        "failures": failures,
+        "num_questions": len(records),
+    }
+
+
+def load_summary_json(filepath: str) -> dict | None:
+    """Load stats from an old-format summary.json file."""
+    try:
+        with open(filepath, "r") as f:
+            data = json.load(f)
+        return {
+            "score": data.get("overall_score", 0),
+            "correct": data.get("overall_correct", data.get("overall_correct_full", 0)),
+            "partial": data.get("overall_partial_correct", 0),
+            "incorrect": data.get("overall_incorrect", data.get("overall_incorrect_choice", 0)),
+            "skipped": data.get("overall_skipped", 0),
+            "failures": data.get("overall_api_parse_failures", 0),
+            "num_questions": data.get("total_questions_processed", 0),
+        }
+    except (OSError, json.JSONDecodeError, KeyError):
+        return None
+
+
+def scan_results(results_dir: str, min_questions: int) -> list[dict]:
+    """Scan result directories and collect stats."""
+    entries = []
+
+    if not os.path.isdir(results_dir):
+        print(f"Results directory not found: {results_dir}", file=sys.stderr)
+        return entries
+
+    for dirname in sorted(os.listdir(results_dir)):
+        dirpath = os.path.join(results_dir, dirname)
+        if not os.path.isdir(dirpath):
+            continue
+
+        parsed = parse_result_dirname(dirname)
+        if not parsed:
+            continue
+
+        # Try summary.jsonl first (new format), then summary.json (old format)
+        stats = None
+        summary_jsonl_path = os.path.join(dirpath, "summary.jsonl")
+        summary_json_path = os.path.join(dirpath, "summary.json")
+
+        if os.path.exists(summary_jsonl_path):
+            records = load_summary_jsonl(summary_jsonl_path)
+            if records:
+                stats = compute_stats(records)
+        elif os.path.exists(summary_json_path):
+            stats = load_summary_json(summary_json_path)
+
+        if not stats or stats.get("num_questions", 0) < min_questions:
+            continue
+
+        # Try to get max score from summary.md
+        md_path = os.path.join(dirpath, "summary.md")
+        max_score = extract_max_score_from_md(md_path)
+
+        entry = {
+            **parsed,
+            **stats,
+            "max_score": max_score,
+            "result_dir": dirname,
+        }
+        entries.append(entry)
+
+    return entries
+
+
+def generate_markdown(entries: list[dict]) -> str:
+    """Generate a Markdown leaderboard table grouped by exam+year."""
+    if not entries:
+        return "# Benchmark Leaderboard\n\nNo results found.\n"
+
+    # Group by exam+year
+    groups: dict[str, list[dict]] = defaultdict(list)
+    for e in entries:
+        key = f"{e['exam']}_{e['year']}"
+        groups[key].append(e)
+
+    lines = ["# Benchmark Leaderboard\n"]
+
+    for group_key in sorted(groups.keys()):
+        group_entries = groups[group_key]
+        # Sort by score descending
+        group_entries.sort(key=lambda x: x["score"], reverse=True)
+
+        exam_display = group_key.replace("_", " ").replace("JEE ADVANCED", "JEE Advanced").replace("JEE MAIN", "JEE Main")
+        lines.append(f"\n## {exam_display}\n")
+        lines.append("| Rank | Model | Score | Max | % | Correct | Partial | Incorrect | Skipped | Failures |")
+        lines.append("|------|-------|-------|-----|---|---------|---------|-----------|---------|----------|")
+
+        for rank, e in enumerate(group_entries, 1):
+            max_score = e.get("max_score") or "?"
+            if isinstance(max_score, int) and max_score > 0:
+                pct = f"{e['score'] / max_score * 100:.1f}%"
+            else:
+                pct = "?"
+            lines.append(
+                f"| {rank} | {e['model']} | {e['score']} | {max_score} | {pct} "
+                f"| {e['correct']} | {e['partial']} | {e['incorrect']} | {e['skipped']} | {e['failures']} |"
+            )
+
+    lines.append("")
+    return "\n".join(lines)
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Generate benchmark leaderboard.")
+    parser.add_argument(
+        "--results-dir",
+        type=str,
+        default="results",
+        help="Path to the results directory (default: results).",
+    )
+    parser.add_argument(
+        "--min-questions",
+        type=int,
+        default=10,
+        help="Minimum questions to include a run (default: 10, filters incomplete runs).",
+    )
+    parser.add_argument(
+        "--output",
+        type=str,
+        help="Output path for leaderboard.json.",
+    )
+    args = parser.parse_args()
+
+    entries = scan_results(args.results_dir, args.min_questions)
+
+    if not entries:
+        print("No valid results found.", file=sys.stderr)
+        sys.exit(1)
+
+    md = generate_markdown(entries)
+    print(md)
+
+    if args.output:
+        with open(args.output, "w") as f:
+            json.dump(entries, f, indent=2)
+        print(f"\nJSON output saved to {args.output}", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

src/benchmark_runner.py

@@ -0,0 +1,923 @@
+import argparse
+import subprocess
+import yaml
+import os
+import json
+import logging
+import datetime # Added for timestamp
+from typing import Dict, Any, List, Set, Tuple # Added for type hinting
+from datasets import Dataset, Features, Value, Image as HFImage
+from tqdm import tqdm
+from PIL import Image as PILImage # Import PIL for type hinting
+
+# ANSI escape codes for colors
+GREEN = '\033[92m'
+RED = '\033[91m'
+RESET = '\033[0m'
+YELLOW = '\033[93m' # For skipped
+CYAN = '\033[96m' # For parse failures
+MAGENTA = '\033[95m' # For API failures
+
+# Import local modules
+from utils import load_api_key
+from llm_interface import get_openrouter_prediction
+# Import evaluation functions
+from evaluation import calculate_exam_scores, calculate_single_question_score_details
+
+# Configure logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+
+
+def get_git_sha() -> str | None:
+    """Returns the current git commit SHA, or None if not in a git repo."""
+    try:
+        result = subprocess.run(
+            ["git", "rev-parse", "HEAD"],
+            capture_output=True, text=True, timeout=5
+        )
+        if result.returncode == 0:
+            return result.stdout.strip()
+    except Exception:
+        pass
+    return None
+
+
+def get_available_models(config_path: str) -> List[str]:
+    """Loads models from the benchmark configuration YAML file."""
+    try:
+        with open(config_path, 'r') as f:
+            config = yaml.safe_load(f)
+        models = config.get("openrouter_models", [])
+        if not models:
+            logging.warning(f"No models found in {config_path} under 'openrouter_models'.")
+        return models
+    except FileNotFoundError:
+        logging.error(f"Configuration file not found at {config_path} for model retrieval.")
+        return []
+    except yaml.YAMLError as e:
+        logging.error(f"Error parsing configuration file {config_path} for model retrieval: {e}")
+        return []
+    except Exception as e:
+        logging.error(f"Unexpected error retrieving models from {config_path}: {e}")
+        return []
+
+def get_available_exam_details(metadata_path: str) -> Tuple[List[str], List[str]]:
+    """Reads metadata.jsonl to get unique exam names and years."""
+    exam_names: Set[str] = set()
+    exam_years: Set[str] = set()
+    try:
+        with open(metadata_path, 'r') as f:
+            for line in f:
+                try:
+                    data = json.loads(line)
+                    if 'exam_name' in data:
+                        exam_names.add(data['exam_name'])
+                    if 'exam_year' in data:
+                        exam_years.add(str(data['exam_year']))
+                except json.JSONDecodeError:
+                    logging.warning(f"Skipping malformed JSON line in {metadata_path}: {line.strip()}")
+        
+        sorted_exam_names = sorted(list(exam_names))
+        sorted_exam_years = sorted(list(exam_years))
+        
+        if not sorted_exam_names:
+            logging.warning(f"No exam names found in {metadata_path}.")
+        if not sorted_exam_years:
+            logging.warning(f"No exam years found in {metadata_path}.")
+            
+        return sorted_exam_names, sorted_exam_years
+    except FileNotFoundError:
+        logging.error(f"Metadata file not found at {metadata_path}.")
+        return [], []
+    except Exception as e:
+        logging.error(f"Unexpected error reading or parsing {metadata_path}: {e}")
+        return [], []
+
+def load_config(config_path: str) -> dict:
+    """Loads the benchmark configuration from a YAML file."""
+    try:
+        with open(config_path, 'r') as f:
+            config = yaml.safe_load(f)
+        logging.info(f"Configuration loaded from {config_path}")
+        return config
+    except FileNotFoundError:
+        logging.error(f"Configuration file not found at {config_path}")
+        raise
+    except yaml.YAMLError as e:
+        logging.error(f"Error parsing configuration file {config_path}: {e}")
+        raise
+
+def append_prediction(result: Dict[str, Any], filepath: str):
+    """Appends a single prediction result to a JSONL file."""
+    # Create a copy to avoid modifying the original dict that might be used elsewhere
+    # and remove evaluation-specific fields before saving to predictions.jsonl
+    prediction_data = result.copy()
+    prediction_data.pop('marks_awarded', None)
+    prediction_data.pop('evaluation_status', None)
+    prediction_data.pop('predicted_answer', None) # Remove predicted_answer
+    prediction_data.pop('ground_truth', None) # Remove ground_truth
+    try:
+        with open(filepath, 'a') as f:
+            json.dump(prediction_data, f)
+            f.write('\n')
+    except IOError as e:
+        logging.error(f"Failed to append prediction to {filepath}: {e}")
+    except Exception as e:
+        logging.error(f"Unexpected error appending prediction to {filepath}: {e}")
+
+def append_summary_detail(result_detail: Dict[str, Any], filepath: str):
+    """Appends a single question's summary details (evaluation status, marks, predicted, truth) to a JSONL file."""
+    try:
+        with open(filepath, 'a') as f:
+            json.dump(result_detail, f)
+            f.write('\n')
+    except IOError as e:
+        logging.error(f"Failed to append summary detail to {filepath}: {e}")
+    except Exception as e:
+        logging.error(f"Unexpected error appending summary detail to {filepath}: {e}")
+
+
+# Removed save_summary function as summary.json is no longer needed.
+
+def generate_markdown_summary(summary: Dict[str, Any], filepath: str):
+    """Generates a human-readable Markdown summary from the results dictionary."""
+    try:
+        md_content = []
+        model_name = summary.get("model_name", "N/A")
+        exam_name = summary.get("exam_name", "N/A")
+        exam_year = summary.get("exam_year", "N/A")
+        timestamp = summary.get("timestamp", "N/A")
+        total_questions_in_dataset = summary.get("total_questions_in_dataset", 0)
+        total_questions_processed_in_run = summary.get("total_questions_processed_in_run", 0)
+
+        filtered_questions_count = 0
+        if total_questions_in_dataset > 0 and total_questions_processed_in_run > 0:
+             filtered_questions_count = total_questions_in_dataset - total_questions_processed_in_run
+
+
+        md_content.append(f"# Benchmark Results: {model_name}")
+        if exam_name and exam_name not in ["N/A", "All_Exams"]: # Only display if a specific exam was targeted
+            md_content.append(f"**Exam Name:** {exam_name}")
+        if exam_year and exam_year not in ["N/A", "All_Years"]: # Only display if a specific year was targeted
+            md_content.append(f"**Exam Year:** {exam_year}")
+        md_content.append(f"**Timestamp:** {timestamp}")
+        md_content.append(f"**Total Questions in Dataset:** {total_questions_in_dataset if total_questions_in_dataset > 0 else 'N/A'}")
+        if filtered_questions_count > 0:
+            md_content.append(f"**Questions Filtered Out:** {filtered_questions_count}")
+        md_content.append(f"**Total Questions Processed in this Run:** {total_questions_processed_in_run}")
+
+        # API usage stats
+        total_cost_usd = summary.get("total_cost_usd")
+        total_tokens = summary.get("total_tokens")
+        avg_latency = summary.get("avg_response_latency_ms")
+        median_latency = summary.get("median_response_latency_ms")
+        if total_tokens or total_cost_usd or avg_latency:
+            md_content.append("")
+            md_content.append("### API Usage")
+            if summary.get("total_prompt_tokens"):
+                md_content.append(f"- **Prompt Tokens:** {summary['total_prompt_tokens']:,}")
+            if summary.get("total_completion_tokens"):
+                md_content.append(f"- **Completion Tokens:** {summary['total_completion_tokens']:,}")
+            if total_tokens:
+                md_content.append(f"- **Total Tokens:** {total_tokens:,}")
+            if total_cost_usd:
+                md_content.append(f"- **Total Cost:** ${total_cost_usd:.4f}")
+            if avg_latency:
+                md_content.append(f"- **Avg Response Latency:** {avg_latency:,} ms")
+            if median_latency:
+                md_content.append(f"- **Median Response Latency:** {median_latency:,} ms")
+
+        md_content.append("\n---\n")
+
+        # Check if NEET results are present (or any dataset with overall_score and section_breakdown)
+        if "overall_score" in summary and "section_breakdown" in summary: # Generic check for score-based summary
+            total_processed = summary.get("total_questions_processed", 0)
+            
+            overall_score = summary.get('overall_score', 'N/A')
+            total_possible_score = summary.get('total_possible_score_for_processed_questions', 'N/A')
+            correct_full_count = summary.get('overall_correct_full', 'N/A')
+            partial_correct_count = summary.get('overall_partial_correct', 'N/A')
+            incorrect_choice_count = summary.get('overall_incorrect_choice', 'N/A')
+            skipped_count = summary.get('overall_skipped', 'N/A')
+            failures_count = summary.get('overall_api_parse_failures', 'N/A')
+            unmapped_count = summary.get('unmapped_section_questions', 'N/A')
+
+            md_content.append("## Exam Scoring Results")
+            md_content.append(f"**Overall Score:** **{overall_score}** / **{total_possible_score}**")
+            md_content.append(f"- **Fully Correct Answers:** {correct_full_count}")
+            if partial_correct_count != 'N/A' and partial_correct_count > 0 :
+                md_content.append(f"- **Partially Correct Answers:** {partial_correct_count}")
+            md_content.append(f"- **Incorrectly Answered (Choice Made):** {incorrect_choice_count}")
+            md_content.append(f"- **Skipped Questions:** {skipped_count}")
+            md_content.append(f"- **API/Parse Failures:** {failures_count}")
+            md_content.append(f"- **Total Questions Processed:** {total_processed}")
+            if unmapped_count > 0:
+                md_content.append(f"- **Unmapped Section Questions:** {unmapped_count} *(Not included in section breakdown)*")
+
+            md_content.append("\n### Detailed Score Calculation by Question Type")
+            question_type_breakdown = summary.get("question_type_breakdown", {})
+            if question_type_breakdown:
+                sorted_q_types = sorted(question_type_breakdown.keys())
+                for q_type in sorted_q_types:
+                    stats = question_type_breakdown[q_type]
+                    q_type_display = q_type.replace('_', ' ').title()
+
+                    correct_count_q = stats.get('correct_full', 0)
+                    partial_count_q = stats.get('partial_correct', 0)
+                    incorrect_count_q = stats.get('incorrect_choice', 0)
+                    skipped_count_q = stats.get('skipped', 0)
+                    api_fail_count_q = stats.get('api_parse_failures', 0)
+                    score_q = stats.get('score', 0)
+
+                    breakdown_parts = []
+                    if correct_count_q > 0:
+                        breakdown_parts.append(f"{correct_count_q} Correct")
+                    if partial_count_q > 0:
+                        breakdown_parts.append(f"{partial_count_q} Partial")
+                    if incorrect_count_q > 0:
+                        breakdown_parts.append(f"{incorrect_count_q} Incorrect")
+                    if skipped_count_q > 0:
+                        breakdown_parts.append(f"{skipped_count_q} Skipped")
+                    if api_fail_count_q > 0:
+                        breakdown_parts.append(f"{api_fail_count_q} API/Parse Fail")
+
+                    breakdown_str = ", ".join(breakdown_parts) if breakdown_parts else "No questions processed"
+
+                    md_content.append(f"**{q_type_display} ({stats.get('count', 0)} questions):** {score_q} marks")
+                    md_content.append(f"  *Breakdown:* {breakdown_str}")
+            else:
+                md_content.append("No question type breakdown available.")
+
+            md_content.append("\n### Section Breakdown")
+            md_content.append("| Section       | Score | Fully Correct | Partially Correct | Incorrect Choice | Skipped | API/Parse Failures |")
+            md_content.append("|---------------|-------|---------------|-------------------|------------------|---------|--------------------|")
+            section_breakdown = summary.get("section_breakdown", {})
+            
+            sorted_section_names = sorted(section_breakdown.keys())
+            if not sorted_section_names and section_breakdown:
+                logging.warning("Could not sort section names for Markdown summary; using unsorted.")
+                sorted_section_names = list(section_breakdown.keys())
+
+            for section_name in sorted_section_names:
+                stats = section_breakdown.get(section_name, {})
+                score = stats.get('score', 'N/A')
+                s_correct = stats.get('correct', 'N/A')
+                s_partial = stats.get('partial_correct', 'N/A')
+                s_incorrect = stats.get('incorrect', 'N/A')
+                s_skipped = stats.get('skipped', 'N/A')
+                s_failures = stats.get('api_parse_failures', 'N/A')
+                display_section_name = section_name.replace('_', ' ')
+                md_content.append(f"| {display_section_name:<13} | {score:<5} | {s_correct:<13} | {s_partial:<17} | {s_incorrect:<16} | {s_skipped:<7} | {s_failures:<18} |")
+            if not sorted_section_names:
+                md_content.append("| No section data available | N/A   | N/A           | N/A               | N/A              | N/A     | N/A                |")
+        
+        # Fallback for simple accuracy (if exam scoring wasn't applicable or failed)
+        elif "accuracy_on_parsed" in summary:
+            md_content.append("## Simple Accuracy Results (Fallback)")
+            md_content.append(f"- **Accuracy (on successfully parsed non-skipped):** {summary.get('accuracy_on_parsed', 'N/A'):.4f}")
+            md_content.append(f"- **Total Processed Attempts:** {summary.get('total_processed_attempts', 'N/A')}")
+            # Add other relevant simple stats if available
+        else:
+            md_content.append("## Summary")
+            md_content.append("*(No specific Exam Scoring or Accuracy metrics found in summary)*")
+
+
+        with open(filepath, 'w') as f:
+            f.write("\n".join(md_content))
+        logging.info(f"Markdown summary saved to {filepath}")
+
+    except IOError as e:
+        logging.error(f"Failed to save markdown summary to {filepath}: {e}")
+    except Exception as e:
+        logging.error(f"Unexpected error generating or saving markdown summary to {filepath}: {e}")
+
+
+def process_question(
+    model_id: str,
+    api_key: str,
+    config: dict,
+    example: dict,
+    image: PILImage.Image,
+    predictions_path: str,
+    summary_details_path: str,
+    attempt: int = 1,
+) -> Dict[str, Any]:
+    """Handles a single question: API call, re-prompt on parse failure, scoring, writing to disk.
+
+    Returns the result_data dict. On API exception, raises so the caller can handle retry queueing.
+    """
+    question_id = example["question_id"]
+    subject = example["subject"]
+    exam_name_from_data = example.get("exam_name", "UNKNOWN_EXAM")
+    question_type_from_data = example.get("question_type", "MCQ_SINGLE_CORRECT")
+    truth = json.loads(example["correct_answer"])
+
+    result_data = {
+        "question_id": question_id,
+        "subject": subject,
+        "exam_name": exam_name_from_data,
+        "question_type": question_type_from_data,
+        "ground_truth": truth,
+        "predicted_answer": None,
+        "raw_response": None,
+        "parse_successful": False,
+        "api_call_successful": False,
+        "error": None,
+        "attempt": attempt,
+        "previous_raw_response_on_reprompt": None,
+        "response_metadata": None,
+    }
+
+    # --- API Call ---
+    logging.info(f"Attempting API call for question: {question_id} with model: {model_id}")
+    parsed_answer, raw_response, response_metadata = get_openrouter_prediction(
+        model_identifier=model_id,
+        api_key=api_key,
+        image=image,
+        exam_name=exam_name_from_data,
+        exam_year=str(example.get("exam_year", "UNKNOWN_YEAR")),
+        question_type=question_type_from_data,
+        max_tokens=config.get("max_tokens", 100),
+        request_timeout=config.get("request_timeout", 60),
+        temperature=config.get("temperature", 0),
+    )
+
+    api_success = raw_response is not None  # False if API returned empty content
+    parse_success = parsed_answer is not None
+
+    # --- Re-prompt Logic ---
+    if api_success and not parse_success and raw_response is not None:
+        logging.warning(f"Question {question_id}: Initial parse failed. Attempting re-prompt.")
+        result_data["previous_raw_response_on_reprompt"] = raw_response
+        try:
+            parsed_answer_rp, raw_response_rp, rp_metadata = get_openrouter_prediction(
+                model_identifier=model_id,
+                api_key=api_key,
+                previous_raw_response=raw_response,
+                question_type=question_type_from_data,
+                max_tokens=config.get("max_tokens", 100),
+                request_timeout=config.get("request_timeout", 60),
+                temperature=config.get("temperature", 0),
+            )
+            if isinstance(parsed_answer_rp, list):
+                parsed_answer_rp = [str(item) for item in parsed_answer_rp]
+
+            result_data.update({
+                "predicted_answer": parsed_answer_rp,
+                "raw_response": raw_response_rp,
+                "parse_successful": parsed_answer_rp is not None,
+                "api_call_successful": True,
+                "attempt": attempt + 1,
+                "response_metadata": rp_metadata,
+            })
+            logging.info(f"Question {question_id}: Re-prompt {'succeeded' if result_data['parse_successful'] else 'failed to parse'}.")
+        except Exception as e_rp:
+            logging.error(f"Re-prompt API call failed for question {question_id}: {e_rp}")
+            result_data.update({
+                "predicted_answer": None,
+                "raw_response": raw_response,
+                "parse_successful": False,
+                "api_call_successful": True,
+                "error": f"Initial parse failed. Re-prompt API call failed: {str(e_rp)}",
+                "attempt": attempt,
+                "response_metadata": response_metadata,
+            })
+    else:
+        current_error = result_data.get("error")
+        if not api_success:
+            current_error = "API call returned empty content."
+
+        if isinstance(parsed_answer, list):
+            parsed_answer = [str(item) for item in parsed_answer]
+
+        result_data.update({
+            "predicted_answer": parsed_answer,
+            "raw_response": raw_response,
+            "parse_successful": parse_success,
+            "api_call_successful": api_success,
+            "error": current_error,
+            "attempt": attempt,
+            "response_metadata": response_metadata,
+        })
+
+    # --- Score and write to disk ---
+    score_details = calculate_single_question_score_details(result_data)
+    result_data["marks_awarded"] = score_details.get("marks_awarded")
+    result_data["evaluation_status"] = score_details.get("evaluation_status")
+
+    metadata = result_data.get("response_metadata") or {}
+    append_summary_detail(
+        {
+            "question_id": question_id,
+            "marks_awarded": result_data["marks_awarded"],
+            "evaluation_status": result_data["evaluation_status"],
+            "predicted_answer": result_data["predicted_answer"],
+            "ground_truth": result_data["ground_truth"],
+            "attempt": result_data["attempt"],
+            "prompt_tokens": metadata.get("prompt_tokens"),
+            "completion_tokens": metadata.get("completion_tokens"),
+            "cost": metadata.get("cost"),
+            "response_latency_ms": metadata.get("response_latency_ms"),
+        },
+        summary_details_path,
+    )
+    append_prediction(result_data, predictions_path)
+
+    return result_data
+
+
+def log_question_result(result_data: Dict[str, Any], prefix: str = "") -> str:
+    """Logs a color-coded result for a single question.
+
+    Returns a category string: 'correct', 'incorrect', 'skipped', 'parse_fail', or 'api_fail'.
+    """
+    question_id = result_data["question_id"]
+    log_message_prefix = f"{prefix}Question {question_id}:"
+    log_message_suffix = f"(Attempt {result_data['attempt']})"
+
+    if not result_data["api_call_successful"]:
+        print(f"{MAGENTA}{log_message_prefix} API Call Failed {log_message_suffix}{RESET}")
+        return "api_fail"
+
+    if not result_data["parse_successful"]:
+        print(f"{CYAN}{log_message_prefix} Failed to parse answer {log_message_suffix}{RESET}")
+        return "parse_fail"
+
+    if result_data["predicted_answer"] == "SKIP":
+        print(f"{YELLOW}{log_message_prefix} Skipped {log_message_suffix}{RESET}")
+        return "skipped"
+
+    marks_awarded = result_data.get("marks_awarded", 0)
+    evaluation_status_value = result_data.get("evaluation_status")
+
+    is_considered_correct = False
+    log_display_status = "N/A"
+    status_check_string = ""
+
+    if evaluation_status_value is True:
+        is_considered_correct = True
+        log_display_status = "True (Boolean)"
+        status_check_string = "CORRECT_TRUE_BOOLEAN"
+    elif isinstance(evaluation_status_value, str):
+        log_display_status = evaluation_status_value
+        status_check_string = evaluation_status_value.strip().upper()
+        if "CORRECT" in status_check_string:
+            is_considered_correct = True
+    elif evaluation_status_value is None:
+        log_display_status = "None"
+        status_check_string = "NONE_STATUS"
+    else:
+        log_display_status = str(evaluation_status_value)
+        status_check_string = str(evaluation_status_value).strip().upper()
+
+    known_eval_skip_statuses = ["SKIPPED_BY_EVAL", "SKIPPED"]
+
+    if is_considered_correct:
+        print(f"{GREEN}{log_message_prefix} Correct - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
+        return "correct"
+    elif status_check_string in known_eval_skip_statuses:
+        print(f"{YELLOW}{log_message_prefix} Skipped by Eval - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
+        return "skipped"
+    else:
+        print(f"{RED}{log_message_prefix} Incorrect - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
+        return "incorrect"
+
+
+def load_completed_question_ids(summary_details_path: str) -> set:
+    """Reads summary.jsonl and returns a set of question_ids that have already been processed."""
+    completed_ids = set()
+    if not os.path.exists(summary_details_path):
+        return completed_ids
+    try:
+        with open(summary_details_path, 'r') as f:
+            for line in f:
+                try:
+                    data = json.loads(line)
+                    qid = data.get("question_id")
+                    if qid:
+                        completed_ids.add(qid)
+                except json.JSONDecodeError:
+                    continue
+    except IOError as e:
+        logging.warning(f"Could not read {summary_details_path} for resume: {e}")
+    return completed_ids
+
+
+def run_benchmark(
+    config: dict,
+    api_key: str,
+    model_to_run: str,
+    output_dir_override: str | None = None,
+    exam_name_choice: str | None = None,
+    exam_year_choice: str | None = None,
+    question_ids_str: str | None = None,
+    resume_dir: str | None = None
+):
+    """Runs the benchmark evaluation loop with incremental saving and retries."""
+
+    # Determine models to run - now it's a single model
+    models_to_run = [model_to_run] # Benchmark will run for the single specified model
+    logging.info(f"Target model for this run: {model_to_run}")
+
+     # Determine base output directory
+    base_output_dir = output_dir_override if output_dir_override else config.get("results_base_dir", "results")
+    os.makedirs(base_output_dir, exist_ok=True)
+
+     # Load dataset directly from metadata.jsonl and images/
+    metadata_path = config.get("metadata_path", "images/metadata.jsonl")
+    images_base_dir = config.get("images_base_dir", "images")
+    try:
+        records = []
+        with open(metadata_path, 'r', encoding='utf-8') as f:
+            for line in f:
+                row = json.loads(line)
+                row["image"] = os.path.join(images_base_dir, row.pop("file_name"))
+                row.setdefault("paper_id", None)
+                records.append(row)
+
+        features = Features({
+            "image": HFImage(decode=True),
+            "question_id": Value("string"),
+            "exam_name": Value("string"),
+            "exam_year": Value("int32"),
+            "subject": Value("string"),
+            "question_type": Value("string"),
+            "correct_answer": Value("string"),
+            "paper_id": Value("int64"),
+        })
+        dataset = Dataset.from_list(records, features=features)
+        logging.info(f"Dataset loaded successfully from {metadata_path}. Original number of questions: {len(dataset)}")
+    except Exception as e:
+        logging.error(f"Failed to load dataset from '{metadata_path}': {e}")
+        logging.error("Ensure 'metadata.jsonl' exists and image paths are valid.")
+        return
+
+    # Filter dataset based on choices
+    original_dataset_size = len(dataset)
+    
+    # Filter by exam_name
+    if exam_name_choice and exam_name_choice.lower() != "all":
+        logging.info(f"Filtering dataset for exam_name: '{exam_name_choice}'")
+        dataset = dataset.filter(lambda example: example.get('exam_name') == exam_name_choice)
+        logging.info(f"Dataset size after exam_name filter: {len(dataset)} questions.")
+
+    # Filter by exam_year
+    if exam_year_choice and exam_year_choice.lower() != "all":
+        try:
+            filter_year_int = int(exam_year_choice)
+            logging.info(f"Filtering dataset for exam_year: {filter_year_int}")
+            dataset = dataset.filter(lambda example: example.get('exam_year') == filter_year_int)
+            logging.info(f"Dataset size after exam_year filter: {len(dataset)} questions.")
+        except ValueError:
+            logging.error(f"Invalid exam_year provided: '{exam_year_choice}'. Must be an integer or 'all'. Year filtering skipped.")
+
+    # Filter by specific question IDs if provided
+    if question_ids_str:
+        try:
+            target_question_ids = {q_id.strip() for q_id in question_ids_str.split(',') if q_id.strip()}
+            if target_question_ids:
+                logging.info(f"Filtering dataset for specific question IDs: {target_question_ids}")
+                dataset = dataset.filter(lambda example: example.get('question_id') in target_question_ids)
+                logging.info(f"Dataset size after question_id filter: {len(dataset)} questions.")
+            else:
+                logging.warning("Empty or invalid question_ids string provided. No question ID filtering applied.")
+        except Exception as e:
+            logging.error(f"Error processing question_ids_str '{question_ids_str}': {e}. No question ID filtering applied.")
+            
+    if len(dataset) < original_dataset_size:
+        logging.info(f"Final dataset size after all filters: {len(dataset)} (originally {original_dataset_size}).")
+    
+    if len(dataset) == 0:
+        logging.warning("No questions to process after filtering. Skipping model benchmark.")
+        return
+
+     # --- Main Loop: Iterate through models ---
+    for model_id in models_to_run:
+        logging.info(f"--- Starting benchmark for model: {model_id} ---")
+
+        # Set up output directory (resume existing or create new)
+        if resume_dir:
+            model_output_dir = resume_dir
+            timestamp = os.path.basename(resume_dir).rsplit('_', 2)[-2] + '_' + os.path.basename(resume_dir).rsplit('_', 1)[-1] if '_' in os.path.basename(resume_dir) else datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
+        else:
+            timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
+            safe_model_name = model_id.replace('/', '_')
+            dir_name_parts = [safe_model_name]
+
+            current_exam_name_for_dir = exam_name_choice if exam_name_choice and exam_name_choice.lower() != "all" else "AllExams"
+            current_exam_year_for_dir = exam_year_choice if exam_year_choice and exam_year_choice.lower() != "all" else "AllYears"
+
+            if current_exam_name_for_dir != "AllExams":
+                dir_name_parts.append(current_exam_name_for_dir.replace('/', '_'))
+            if current_exam_year_for_dir != "AllYears":
+                dir_name_parts.append(str(current_exam_year_for_dir))
+
+            dir_name_parts.append(timestamp)
+
+            model_output_dir_name = "_".join(filter(None, dir_name_parts))
+            model_output_dir = os.path.join(base_output_dir, model_output_dir_name)
+
+        os.makedirs(model_output_dir, exist_ok=True)
+        predictions_path = os.path.join(model_output_dir, "predictions.jsonl")
+        summary_details_path = os.path.join(model_output_dir, "summary.jsonl")
+        markdown_summary_path = os.path.join(model_output_dir, "summary.md")
+        logging.info(f"Results for {model_id} will be saved to: {model_output_dir}")
+
+        # Resume: skip already-completed questions
+        if resume_dir:
+            completed_ids = load_completed_question_ids(summary_details_path)
+            if completed_ids:
+                logging.info(f"Resuming: found {len(completed_ids)} already-completed questions. Skipping them.")
+                dataset = dataset.filter(lambda example: example.get('question_id') not in completed_ids)
+                logging.info(f"Remaining questions to process: {len(dataset)}")
+                if len(dataset) == 0:
+                    logging.info("All questions already completed. Nothing to resume.")
+                    return
+
+        current_total_questions = len(dataset)
+        logging.info(f"Processing {current_total_questions} questions for model: {model_id}")
+
+        model_results = [] # Stores results in memory for final calculation
+        failed_questions_data = [] # Stores data needed to retry failed questions
+
+        # Counters for tqdm postfix
+        initial_correct_count = 0
+        initial_incorrect_count = 0
+        initial_skipped_count = 0
+        initial_parse_fail_count = 0
+        initial_api_fail_count = 0
+
+        # --- Initial Pass: Iterate through questions ---
+        pbar_initial = tqdm(dataset, desc=f"Processing {model_id} (Initial Pass)", total=current_total_questions)
+        for example in pbar_initial:
+            try:
+                result_data = process_question(
+                    model_id=model_id, api_key=api_key, config=config,
+                    example=example, image=example["image"],
+                    predictions_path=predictions_path, summary_details_path=summary_details_path,
+                    attempt=1,
+                )
+                model_results.append(result_data)
+                category = log_question_result(result_data)
+            except Exception as e:
+                logging.error(f"Initial API call failed for question {example['question_id']} (Attempt 1): {e}")
+                failed_questions_data.append(example)
+                category = "api_fail"
+
+            if category == "correct":
+                initial_correct_count += 1
+            elif category == "incorrect":
+                initial_incorrect_count += 1
+            elif category == "skipped":
+                initial_skipped_count += 1
+            elif category == "parse_fail":
+                initial_parse_fail_count += 1
+            elif category == "api_fail":
+                initial_api_fail_count += 1
+            pbar_initial.set_postfix_str(f"✓:{initial_correct_count} ✗:{initial_incorrect_count} S:{initial_skipped_count} P!:{initial_parse_fail_count} A!:{initial_api_fail_count}")
+
+        pbar_initial.close()
+
+        # --- Retry Pass for questions with initial API failures ---
+        if failed_questions_data:
+            logging.info(f"--- Retrying {len(failed_questions_data)} questions with initial API failures for model: {model_id} ---")
+
+            retry_correct_count = 0
+            retry_incorrect_count = 0
+            retry_skipped_count = 0
+            retry_parse_fail_count = 0
+            retry_api_fail_count = 0
+
+            pbar_retry = tqdm(failed_questions_data, desc=f"Processing {model_id} (API Retry Pass)", total=len(failed_questions_data))
+            for example_retry in pbar_retry:
+                try:
+                    result_data_retry = process_question(
+                        model_id=model_id, api_key=api_key, config=config,
+                        example=example_retry, image=example_retry["image"],
+                        predictions_path=predictions_path, summary_details_path=summary_details_path,
+                        attempt=2,
+                    )
+                    model_results.append(result_data_retry)
+                    category = log_question_result(result_data_retry, prefix="(Retry) ")
+                except Exception as e_retry_api:
+                    logging.error(f"API call failed permanently for question {example_retry['question_id']} (Attempt 2 API Retry): {e_retry_api}")
+                    # Score and write the failure so it appears in output files
+                    fail_data = {
+                        "question_id": example_retry["question_id"],
+                        "subject": example_retry["subject"],
+                        "exam_name": example_retry.get("exam_name", "UNKNOWN_EXAM"),
+                        "question_type": example_retry.get("question_type", "MCQ_SINGLE_CORRECT"),
+                        "ground_truth": json.loads(example_retry["correct_answer"]),
+                        "predicted_answer": None, "raw_response": None,
+                        "parse_successful": False, "api_call_successful": False,
+                        "error": f"Initial API fail. Retry API call also failed: {str(e_retry_api)}",
+                        "attempt": 2, "previous_raw_response_on_reprompt": None,
+                    }
+                    score_details = calculate_single_question_score_details(fail_data)
+                    fail_data["marks_awarded"] = score_details.get("marks_awarded")
+                    fail_data["evaluation_status"] = score_details.get("evaluation_status")
+                    append_summary_detail(
+                        {"question_id": fail_data["question_id"], "marks_awarded": fail_data["marks_awarded"],
+                         "evaluation_status": fail_data["evaluation_status"], "predicted_answer": None,
+                         "ground_truth": fail_data["ground_truth"], "attempt": 2},
+                        summary_details_path,
+                    )
+                    append_prediction(fail_data, predictions_path)
+                    model_results.append(fail_data)
+                    category = "api_fail"
+
+                if category == "correct":
+                    retry_correct_count += 1
+                elif category == "incorrect":
+                    retry_incorrect_count += 1
+                elif category == "skipped":
+                    retry_skipped_count += 1
+                elif category == "parse_fail":
+                    retry_parse_fail_count += 1
+                elif category == "api_fail":
+                    retry_api_fail_count += 1
+                pbar_retry.set_postfix_str(f"✓:{retry_correct_count} ✗:{retry_incorrect_count} S:{retry_skipped_count} P!:{retry_parse_fail_count} A!:{retry_api_fail_count}")
+            pbar_retry.close()
+
+        # --- Final Evaluation for the current model ---
+        logging.info(f"--- Calculating final results for model: {model_id} ---")
+        
+        # Always use calculate_exam_scores now
+        evaluation_summary = calculate_exam_scores(model_results) # model_results modified in-place
+        
+        # Use the actual choices for the summary, defaulting to "All" if not specified or "all"
+        summary_exam_name_display = exam_name_choice if exam_name_choice and exam_name_choice.lower() != "all" else "All_Exams"
+        summary_exam_year_display = exam_year_choice if exam_year_choice and exam_year_choice.lower() != "all" else "All_Years"
+
+        # Aggregate response metadata across all questions
+        total_prompt_tokens = 0
+        total_completion_tokens = 0
+        total_cost = 0.0
+        latencies = []
+        for r in model_results:
+            meta = r.get("response_metadata") or {}
+            if meta.get("prompt_tokens"):
+                total_prompt_tokens += meta["prompt_tokens"]
+            if meta.get("completion_tokens"):
+                total_completion_tokens += meta["completion_tokens"]
+            if meta.get("cost"):
+                total_cost += meta["cost"]
+            if meta.get("response_latency_ms"):
+                latencies.append(meta["response_latency_ms"])
+
+        summary = {
+            "model_name": model_id, # This is model_to_run
+            "exam_name": summary_exam_name_display,
+            "exam_year": summary_exam_year_display,
+            "question_ids_filter": question_ids_str if question_ids_str else "None", # Add question ID filter info
+            "timestamp": timestamp,
+            "git_sha": get_git_sha(),
+            "temperature": config.get("temperature", 0),
+            "max_tokens": config.get("max_tokens", 100),
+            "total_questions_in_dataset": original_dataset_size, # Total before any filtering
+            "total_questions_processed_in_run": len(dataset), # Total after filtering for this run
+            "total_prompt_tokens": total_prompt_tokens,
+            "total_completion_tokens": total_completion_tokens,
+            "total_tokens": total_prompt_tokens + total_completion_tokens,
+            "total_cost_usd": round(total_cost, 6),
+            "avg_response_latency_ms": round(sum(latencies) / len(latencies)) if latencies else None,
+            "median_response_latency_ms": round(sorted(latencies)[len(latencies) // 2]) if latencies else None,
+            **evaluation_summary
+        }
+        logging.info(f"Overall Score: {summary.get('overall_score')}")
+        logging.info(f"Full Correct: {summary.get('overall_correct_full')}, Partial Correct: {summary.get('overall_partial_correct')}, Incorrect Choice: {summary.get('overall_incorrect_choice')}, Skipped: {summary.get('overall_skipped')}, API/Parse Failures: {summary.get('overall_api_parse_failures')}")
+        
+        logging.info(f"--- Results Summary for model: {model_id} ---")
+        logging.info(json.dumps(summary, indent=2, sort_keys=True))
+        logging.info("-------------------------------------")
+
+        # The model_results list was modified in-place by calculate_exam_scores
+        # to include evaluation_status and marks_awarded.
+        # predictions.jsonl is now written incrementally without evaluation details.
+        # No need to overwrite predictions.jsonl here.
+
+
+        # Save final summary (Markdown) for the current model
+        # The summary.json file is no longer generated as per user request.
+        generate_markdown_summary(summary, markdown_summary_path) # Call the new function
+
+    logging.info("Benchmark run completed.")
+
+
+if __name__ == "__main__":
+    # Get available choices for arguments
+    # Assuming benchmark_config.yaml is in a 'configs' directory relative to script or a fixed path
+    default_config_path = "configs/benchmark_config.yaml"
+    default_metadata_path = "images/metadata.jsonl"
+
+    available_models = get_available_models(default_config_path)
+    available_exam_names, available_exam_years = get_available_exam_details(default_metadata_path)
+
+    # Add "all" option for exams and years
+    exam_name_choices = ["all"] + available_exam_names
+    exam_year_choices = ["all"] + available_exam_years
+
+    parser = argparse.ArgumentParser(description="Run JEE/NEET LLM Benchmark.")
+    parser.add_argument(
+        "--config",
+        type=str,
+        default=default_config_path,
+        help=f"Path to the benchmark configuration YAML file (default: {default_config_path})."
+    )
+    parser.add_argument(
+        "--model", # Changed from --models
+        type=str,
+        required=True if available_models else False, # Required if models are available
+        choices=available_models if available_models else None,
+        help="Select the model to run." + (f" Available: {', '.join(available_models)}." if available_models else " (No models found in config)")
+    )
+    parser.add_argument(
+        "--output_dir",
+        type=str,
+        help="Override the base output directory specified in the config file."
+    )
+    parser.add_argument(
+        "--exam_name",
+        type=str,
+        default="all",
+        choices=exam_name_choices if exam_name_choices else ["all"],
+        help="Select the exam name to run, or 'all' for all exams." + (f" Available: {', '.join(available_exam_names)}." if available_exam_names else "")
+    )
+    parser.add_argument(
+        "--exam_year",
+        type=str, 
+        default="all",
+        choices=exam_year_choices if exam_year_choices else ["all"],
+        help="Select the exam year to run, or 'all' for all years." + (f" Available: {', '.join(available_exam_years)}." if available_exam_years else "")
+    )
+    parser.add_argument(
+        "--question_ids",
+        type=str,
+        default=None,
+        help="Optional: Comma-separated list of specific question IDs to run (e.g., ID1,ID2,ID3)."
+    )
+    parser.add_argument(
+        "--resume",
+        type=str,
+        default=None,
+        help="Optional: Path to an existing results directory to resume an interrupted run."
+    )
+    parser.add_argument(
+        "--temperature",
+        type=float,
+        default=None,
+        help="Override sampling temperature (default: from config, typically 0 for deterministic output)."
+    )
+    parser.add_argument(
+        "--num_runs",
+        type=int,
+        default=1,
+        help="Number of independent runs per model (default: 1). Use 3+ for publication-grade variance analysis."
+    )
+    args = parser.parse_args()
+
+    # Dynamically update config path if user provides a different one
+    if args.config != default_config_path:
+        logging.info(f"User provided config path: {args.config}. Re-fetching models if necessary.")
+        # If models were not found with default, or if user specified a different config, try to load models from it.
+        if not available_models or args.model not in available_models: 
+            user_config_models = get_available_models(args.config)
+            if args.model not in user_config_models:
+                logging.error(f"Selected model '{args.model}' not found in the specified config '{args.config}'. Exiting.")
+                exit(1) # Or handle more gracefully
+            # Potentially update choices if parser allowed any string due to no initial models
+            # This is complex with argparse after parsing. For now, we rely on the initial check or error out.
+
+    try:
+        # Load API key first - fail fast if not set
+        api_key = load_api_key()
+        # Load configuration using the (potentially user-overridden) config path
+        config = load_config(args.config)
+
+        # Apply CLI temperature override
+        if args.temperature is not None:
+            config["temperature"] = args.temperature
+
+        # Ensure the selected model is in the config
+        if args.model not in config.get("openrouter_models", []):
+            logging.error(f"The model '{args.model}' is not listed in '{args.config}'. Please check the model name or the config file.")
+            exit(1)
+
+        # Run the benchmark (potentially multiple runs)
+        for run_num in range(1, args.num_runs + 1):
+            if args.num_runs > 1:
+                logging.info(f"=== Starting run {run_num}/{args.num_runs} ===")
+            run_benchmark(
+                config=config,
+                api_key=api_key,
+                model_to_run=args.model,
+                output_dir_override=args.output_dir,
+                exam_name_choice=args.exam_name,
+                exam_year_choice=args.exam_year,
+                question_ids_str=args.question_ids,
+                resume_dir=args.resume if run_num == 1 else None,  # Only resume on first run
+            )
+    except (ValueError, FileNotFoundError, yaml.YAMLError) as e:
+        logging.error(f"Setup failed: {e}")
+    except Exception as e:
+        logging.error(f"An unexpected error occurred during benchmark execution: {e}", exc_info=True)

src/evaluation.py

@@ -0,0 +1,415 @@
+import logging
+import re
+from typing import List, Optional, Union, Dict, Any
+
+# Configure logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+
+
+def get_subject_as_section(subject: str, question_num_for_log: int) -> Optional[str]:
+    """
+    Returns the subject name directly as the section identifier.
+    question_num_for_log is only used for logging context if subject is invalid.
+    """
+    if subject and isinstance(subject, str) and subject.strip():
+        return subject.strip()
+    else:
+        logging.warning(f"Invalid or missing subject ('{subject}') for question_num '{question_num_for_log}'. Cannot determine section.")
+        return None
+
+def is_within_range(predicted_value_str: str, lower_bound_str: str, upper_bound_str: str) -> bool:
+    """
+    Checks if a predicted numerical value (as a string) falls within a specified range.
+    The comparison is inclusive.
+    """
+    try:
+        predicted_value = float(predicted_value_str)
+        lower_bound = float(lower_bound_str)
+        upper_bound = float(upper_bound_str)
+    except ValueError:
+        logging.debug(f"Could not convert predicted value '{predicted_value_str}' or bounds ('{lower_bound_str}', '{upper_bound_str}') to numbers.")
+        return False
+
+    return lower_bound <= predicted_value <= upper_bound
+
+
+def calculate_single_question_score_details(result_item: Dict[str, Any]) -> Dict[str, Any]:
+    """
+    Calculates marks_awarded and evaluation_status for a single question result.
+
+    Args:
+        result_item (Dict[str, Any]): A dictionary for a single question, must contain:
+            'question_id' (str)
+            'exam_name' (str)
+            'question_type' (str)
+            'ground_truth' (List[str] | str)  # Changed to string
+            'predicted_answer' (List[str] | str | None) # Changed to string
+            'api_call_successful' (bool)
+
+    Returns:
+        Dict[str, Any]: A dictionary with 'marks_awarded' (int) and 'evaluation_status' (str).
+    """
+    question_id = result_item.get("question_id", "UNKNOWN_QID") # Provide default for logging
+    exam_name = result_item.get("exam_name", "").upper()
+    question_type = result_item.get("question_type", "").upper()
+    pred = result_item.get("predicted_answer")
+    truth = result_item.get("ground_truth")
+    api_success = result_item.get("api_call_successful", False)
+
+    current_score_change = 0
+    evaluation_status = "unknown"
+
+    # Ensure truth is a set of uppercase strings for consistent processing.
+    # Ground truth from metadata.jsonl is expected to be a list of strings.
+    # e.g., ["1"], ["A"], ["12.75"], ["A", "C"]
+    # For integer ranges, it will be like [["0.7", "0.8"]]
+    truth_processed: List[Union[str, List[str]]] = []
+    if isinstance(truth, list):
+        for t_item in truth:
+            if isinstance(t_item, str):
+                truth_processed.append(t_item.upper())
+            elif isinstance(t_item, list) and len(t_item) == 2 and all(isinstance(x, str) for x in t_item):
+                truth_processed.append([x.upper() for x in t_item]) # Store range as list of uppercase strings
+            else:
+                logging.error(f"Invalid item in ground_truth list for {question_id}: {t_item}. Skipping.")
+    elif isinstance(truth, str):
+        truth_processed.append(truth.upper())
+    else:
+        logging.error(f"Invalid ground_truth format for {question_id}: {truth} (type: {type(truth)}). Assigning 0 marks.")
+        return {"marks_awarded": 0, "evaluation_status": "error_bad_ground_truth"}
+
+
+    if not api_success or pred is None: # pred is None means our internal parsing failed
+        evaluation_status = "failure_api_or_parse"
+        current_score_change = 0  # No penalty: parse/API failure is not a deliberate wrong choice
+    elif isinstance(pred, str) and pred.upper() == "SKIP": # Standardize SKIP comparison
+        current_score_change = 0
+        evaluation_status = "skipped"
+    elif isinstance(pred, list) and all(isinstance(p, str) for p in pred):
+        pred_set = {p.upper() for p in pred} # Convert to uppercase strings
+        
+        # Handle MCQ_SINGLE_CORRECT first, as it has special logic for multiple truths.
+        # The parser (`parse_llm_answer`) returns `pred` as `list[str]` with one element for valid single answers.
+        if question_type == "MCQ_SINGLE_CORRECT":
+            # A prediction is correct if its single element is present in the truth_set.
+            # This accommodates metadata entries where `correct_answer` for an MCQ_SINGLE_CORRECT
+            # might list multiple acceptable options (e.g., if a question had two official correct answers).
+            is_correct = False
+            if len(pred_set) == 1: # Ensure prediction is indeed a single option
+                single_pred_answer = list(pred_set)[0] # Get the single predicted option
+                # Check against all processed truths (which are single strings for MCQ)
+                if single_pred_answer in truth_processed: 
+                    is_correct = True
+            
+            if is_correct:
+                evaluation_status = "correct"
+                if exam_name == "NEET": current_score_change = 4
+                elif exam_name == "JEE_MAIN": current_score_change = 4
+                elif exam_name == "JEE_ADVANCED": current_score_change = 3
+                else: current_score_change = 1 # Default positive score for unknown exam
+            else:
+                evaluation_status = "incorrect"
+                if exam_name == "NEET": current_score_change = -1
+                elif exam_name == "JEE_MAIN": current_score_change = -1
+                elif exam_name == "JEE_ADVANCED": current_score_change = -1
+                else: current_score_change = 0 # Default no penalty
+
+        elif exam_name == "JEE_MAIN" and question_type == "INTEGER": # Integer answers are now strings in a list e.g. ["14"]
+            # For JEE_MAIN INTEGER, we expect truth_processed to contain single strings
+            is_correct = False
+            if len(pred_set) == 1:
+                predicted_answer_str = list(pred_set)[0]
+                if predicted_answer_str in truth_processed: # Check against single string truths
+                    is_correct = True
+            
+            if is_correct:
+                current_score_change = 4; evaluation_status = "correct"
+            else:
+                current_score_change = 0; evaluation_status = "incorrect"
+        
+        elif exam_name == "JEE_ADVANCED":
+            # Note: MCQ_SINGLE_CORRECT for JEE_ADVANCED is handled by the common block above
+            if question_type == "INTEGER":
+                is_correct = False
+                if len(pred_set) == 1:
+                    predicted_answer_str = list(pred_set)[0] # Get the single predicted string
+
+                    # Iterate through each ground truth entry in the 'truth_processed' list
+                    for gt_entry in truth_processed:
+                        if isinstance(gt_entry, list) and len(gt_entry) == 2: # This is a range [lower, upper]
+                            lower_bound_str, upper_bound_str = gt_entry[0], gt_entry[1]
+                            if is_within_range(predicted_answer_str, lower_bound_str, upper_bound_str):
+                                is_correct = True
+                                break # Found a matching range, no need to check others
+                        elif isinstance(gt_entry, str): # This is an exact integer match
+                            if predicted_answer_str == gt_entry: # gt_entry is already uppercase
+                                is_correct = True
+                                break # Found an exact match, no need to check others
+                
+                if is_correct:
+                    current_score_change = 4; evaluation_status = "correct"
+                else:
+                    current_score_change = 0; evaluation_status = "incorrect"
+            elif question_type == "MCQ_MULTIPLE_CORRECT":
+                # For MCQ_MULTIPLE_CORRECT, truth_processed contains single strings
+                truth_set_mcq = set(truth_processed) # Convert to set for intersection operations
+
+                num_correct_options_in_truth = len(truth_set_mcq)
+                num_chosen_options = len(pred_set)
+                correct_chosen_options = pred_set.intersection(truth_set_mcq)
+                incorrect_chosen_options = pred_set.difference(truth_set_mcq)
+                num_correct_chosen = len(correct_chosen_options)
+                num_incorrect_chosen = len(incorrect_chosen_options)
+
+                if num_incorrect_chosen > 0:
+                    current_score_change = -2; evaluation_status = "incorrect_negative"
+                elif num_correct_chosen == num_correct_options_in_truth and num_chosen_options == num_correct_options_in_truth:
+                    current_score_change = 4; evaluation_status = "correct_full"
+                elif num_correct_options_in_truth == 4 and num_correct_chosen == 3 and num_chosen_options == 3:
+                    current_score_change = 3; evaluation_status = "partial_3_of_4"
+                elif num_correct_options_in_truth >= 3 and num_correct_chosen == 2 and num_chosen_options == 2:
+                    current_score_change = 2; evaluation_status = "partial_2_of_3_plus"
+                elif num_correct_options_in_truth >= 2 and num_correct_chosen == 1 and num_chosen_options == 1:
+                    current_score_change = 1; evaluation_status = "partial_1_of_2_plus"
+                else:
+                    current_score_change = 0; evaluation_status = "no_marks_no_penalty"
+        else:
+            logging.warning(f"Unknown exam_name/question_type combination for scoring: {exam_name}/{question_type} for QID {question_id}. Assigning 0 marks.")
+            current_score_change = 0
+            evaluation_status = "unknown_exam_type"
+    else:
+        logging.error(f"Unexpected prediction type for {question_id}: {pred}. Treating as API/Parse Failure.")
+        current_score_change = -1 # Default penalty
+        evaluation_status = "failure_unexpected_type"
+
+    return {"marks_awarded": current_score_change, "evaluation_status": evaluation_status}
+
+
+def calculate_max_score_for_question(exam_name: str, question_type: str) -> int:
+    """
+    Returns the maximum possible score for a given exam and question type.
+    """
+    exam_name = exam_name.upper()
+    question_type = question_type.upper()
+
+    if exam_name == "NEET" and question_type == "MCQ_SINGLE_CORRECT":
+        return 4
+    elif exam_name == "JEE_MAIN":
+        if question_type == "MCQ_SINGLE_CORRECT":
+            return 4
+        elif question_type == "INTEGER":
+            return 4
+    elif exam_name == "JEE_ADVANCED":
+        if question_type == "MCQ_SINGLE_CORRECT":
+            return 3
+        elif question_type == "INTEGER":
+            return 4
+        elif question_type == "MCQ_MULTIPLE_CORRECT":
+            return 4 # Max score for multiple correct is 4
+    return 0 # Default for unknown types
+
+
+def calculate_exam_scores(results: List[Dict[str, Any]]) -> Dict[str, Any]:
+    """
+    Calculates exam scores based on exam_name and question_type, providing section-wise breakdown
+    and detailed question type statistics.
+
+    Args:
+        results (List[Dict[str, Any]]): A list of result dictionaries. Each dict must contain:
+            'question_id' (str)
+            'subject' (str)
+            'exam_name' (str) e.g., "NEET", "JEE_MAIN", "JEE_ADVANCED"
+            'question_type' (str) e.g., "MCQ_SINGLE_CORRECT", "MCQ_MULTIPLE_CORRECT", "INTEGER"
+            'ground_truth' (List[str] | str): Correct answer(s). For INTEGER, it's a single str.
+            'predicted_answer' (List[str] | str | None): Model's prediction.
+            'api_call_successful' (bool): Whether the API call succeeded.
+            This list will be modified in-place to add 'evaluation_status' and 'marks_awarded' by calling
+            calculate_single_question_score_details for each item.
+    Returns:
+        Dict[str, Any]: A dictionary containing overall and section-wise scores and counts,
+                        plus question type breakdowns and total possible score.
+    """
+    if not results:
+        return {"error": "No results provided."}
+
+    overall_stats = {
+        "score": 0,
+        "correct": 0,
+        "incorrect": 0,
+        "skipped": 0,
+        "api_parse_failures": 0,
+        "partial_correct": 0,
+        "total_possible_score": 0 # New field
+    }
+    
+    # Initialize question type breakdown
+    question_type_breakdown: Dict[str, Dict[str, Any]] = {}
+
+    valid_subjects_from_data = [r.get("subject") for r in results if r.get("subject") and isinstance(r.get("subject"), str) and r.get("subject").strip()]
+    if not valid_subjects_from_data and results:
+        logging.warning("No valid subjects found in results data to initialize section_stats.")
+    
+    unique_subjects = sorted(list(set(s.strip() for s in valid_subjects_from_data)))
+    section_stats = {
+        subj: {"score": 0, "correct": 0, "incorrect": 0, "skipped": 0, "api_parse_failures": 0, "partial_correct": 0}
+        for subj in unique_subjects
+    }
+    if not unique_subjects and results:
+        logging.warning("section_stats is empty because no unique, valid subjects were found.")
+    
+    unmapped_section_questions = 0
+
+    for result in results:
+        question_id = result.get("question_id") # For logging within loop
+        subject = result.get("subject") # For section mapping
+        exam_name = result.get("exam_name", "").upper()
+        question_type = result.get("question_type", "").upper()
+
+        # Calculate score details for the single question
+        score_details = calculate_single_question_score_details(result)
+        current_score_change = score_details.get("marks_awarded", 0)
+        evaluation_status = score_details.get("evaluation_status", "unknown_error_in_scoring")
+
+        # Update the result dictionary in-place (as original function did)
+        result['evaluation_status'] = evaluation_status
+        result['marks_awarded'] = current_score_change
+        
+        # Accumulate total possible score
+        overall_stats["total_possible_score"] += calculate_max_score_for_question(exam_name, question_type)
+
+        # Determine boolean flags based on evaluation_status for aggregation
+        is_correct_full = evaluation_status in ["correct", "correct_full"]
+        is_partial_correct = evaluation_status.startswith("partial_")
+        is_incorrect_choice = evaluation_status in ["incorrect", "incorrect_negative"]
+        is_skipped = evaluation_status == "skipped"
+        is_api_parse_failure = evaluation_status in ["failure_api_or_parse", "failure_unexpected_type", "error_bad_ground_truth"]
+
+
+        overall_stats["score"] += current_score_change
+        if is_correct_full: overall_stats["correct"] += 1
+        if is_incorrect_choice: overall_stats["incorrect"] += 1
+        if is_skipped: overall_stats["skipped"] += 1
+        if is_api_parse_failure: overall_stats["api_parse_failures"] += 1
+        if is_partial_correct: overall_stats["partial_correct"] +=1
+        
+        # Aggregate by question type
+        if question_type not in question_type_breakdown:
+            question_type_breakdown[question_type] = {
+                "count": 0,
+                "score": 0,
+                "correct_full": 0,
+                "partial_correct": 0,
+                "incorrect_choice": 0,
+                "skipped": 0,
+                "api_parse_failures": 0,
+                "max_score_per_question": calculate_max_score_for_question(exam_name, question_type)
+            }
+        
+        q_type_stats = question_type_breakdown[question_type]
+        q_type_stats["count"] += 1
+        q_type_stats["score"] += current_score_change
+        if is_correct_full: q_type_stats["correct_full"] += 1
+        if is_incorrect_choice: q_type_stats["incorrect_choice"] += 1
+        if is_skipped: q_type_stats["skipped"] += 1
+        if is_api_parse_failure: q_type_stats["api_parse_failures"] += 1
+        if is_partial_correct: q_type_stats["partial_correct"] += 1
+
+        section = None
+        if subject:
+            question_num_for_log = -1 # Placeholder, as QID might not have number
+            if question_id:
+                match_num = re.search(r'_(\d+)$', question_id)
+                if match_num:
+                    try: question_num_for_log = int(match_num.group(1))
+                    except ValueError: pass
+            section = get_subject_as_section(subject, question_num_for_log)
+
+        if section and section in section_stats:
+            section_stats[section]["score"] += current_score_change
+            if is_correct_full: section_stats[section]["correct"] += 1
+            if is_incorrect_choice: section_stats[section]["incorrect"] += 1
+            if is_skipped: section_stats[section]["skipped"] += 1
+            if is_api_parse_failure: section_stats[section]["api_parse_failures"] += 1
+            if is_partial_correct: section_stats[section]["partial_correct"] +=1
+        elif section is None and not is_api_parse_failure : # only count as unmapped if not already an API/parse failure (which might lack subject)
+            unmapped_section_questions += 1
+            # logging.warning(f"Could not map question to section: ID={question_id}, Subject={subject}") # Already logged by get_subject_as_section if subject is bad
+
+    logging.info(f"Exam Score Calculation Complete. Overall Score: {overall_stats['score']}")
+    if unmapped_section_questions > 0:
+        logging.warning(f"{unmapped_section_questions} questions could not be mapped to a section.")
+
+    return {
+        "overall_score": overall_stats["score"],
+        "overall_correct_full": overall_stats["correct"],
+        "overall_partial_correct": overall_stats["partial_correct"],
+        "overall_incorrect_choice": overall_stats["incorrect"],
+        "overall_skipped": overall_stats["skipped"],
+        "overall_api_parse_failures": overall_stats["api_parse_failures"],
+        "total_questions_processed": len(results),
+        "total_possible_score_for_processed_questions": overall_stats["total_possible_score"], # New field
+        "unmapped_section_questions": unmapped_section_questions,
+        "section_breakdown": section_stats,
+        "question_type_breakdown": question_type_breakdown # New field
+    }
+
+
+# Example Usage (for testing)
+if __name__ == '__main__':
+    print("Running evaluation tests...")
+
+    # --- Test calculate_exam_scores (now with strings) ---
+    print("\n--- Testing calculate_exam_scores ---")
+    test_results_exam = [
+        # NEET - Answers as strings "1", "2", "A", "B" etc.
+        {"question_id": "N001", "subject": "Physics", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["1"], "predicted_answer": ["1"], "api_call_successful": True}, # Correct +4
+        {"question_id": "N002", "subject": "Physics", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["D"], "predicted_answer": ["B"], "api_call_successful": True}, # Incorrect -1
+        {"question_id": "N003", "subject": "Chemistry", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["4"], "predicted_answer": "SKIP", "api_call_successful": True}, # Skipped 0
+        {"question_id": "N004", "subject": "Chemistry", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["C"], "predicted_answer": None, "api_call_successful": False}, # API Fail -1
+        {"question_id": "N005", "subject": "Botany", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["A"], "predicted_answer": None, "api_call_successful": True}, # Parse Fail -1
+
+        # JEE Main - MCQ - Answers as strings "1", "2", "A", "B" etc.
+        {"question_id": "JM001", "subject": "Maths", "exam_name": "JEE_MAIN", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["2"], "predicted_answer": ["2"], "api_call_successful": True}, # Correct +4
+        {"question_id": "JM002", "subject": "Maths", "exam_name": "JEE_MAIN", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["C"], "predicted_answer": ["a"], "api_call_successful": True}, # Incorrect -1 (C vs a)
+        # JEE Main - Integer - Answers as strings e.g., "5", "10"
+        {"question_id": "JM003", "subject": "Physics", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["5"], "predicted_answer": ["5"], "api_call_successful": True}, # Correct +4
+        {"question_id": "JM004", "subject": "Physics", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["10"], "predicted_answer": ["8"], "api_call_successful": True}, # Incorrect 0
+        {"question_id": "JM005", "subject": "Chemistry", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["7"], "predicted_answer": None, "api_call_successful": True}, # Parse Fail 0
+
+        # JEE Advanced - MCQ Single Correct - Answers as strings "A", "B" etc.
+        {"question_id": "JA001", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["A"], "predicted_answer": ["a"], "api_call_successful": True}, # Correct +3 (case-insensitive)
+        {"question_id": "JA002", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["B"], "predicted_answer": ["C"], "api_call_successful": True}, # Incorrect -1
+        # JEE Advanced - Integer - Answers as strings e.g., "12", "0"
+        {"question_id": "JA003", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "ground_truth": ["12"], "predicted_answer": ["12"], "api_call_successful": True}, # Correct +4
+        {"question_id": "JA004", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "ground_truth": ["0"], "predicted_answer": ["1"], "api_call_successful": True}, # Incorrect 0
+        # JEE Advanced - MCQ Multiple Correct - Answers as strings "A", "C" etc.
+        {"question_id": "JA005", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["a", "c"], "api_call_successful": True}, # All Correct +4
+        {"question_id": "JA006", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B", "C"], "predicted_answer": ["a", "b"], "api_call_successful": True}, # Partial +2 (3 correct, 2 chosen)
+        {"question_id": "JA007", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B", "C", "D"], "predicted_answer": ["a", "b", "c"], "api_call_successful": True}, # Partial +3 (4 correct, 3 chosen)
+        {"question_id": "JA008", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B"], "predicted_answer": ["a"], "api_call_successful": True}, # Partial +1 (2 correct, 1 chosen)
+        {"question_id": "JA009", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["a", "b"], "api_call_successful": True}, # Incorrect option chosen -2
+        {"question_id": "JA010", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["b", "d"], "api_call_successful": True}, # All incorrect options chosen -2
+        {"question_id": "JA011", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": "SKIP", "api_call_successful": True}, # Skipped 0
+        {"question_id": "JA012", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A"], "predicted_answer": ["a"], "api_call_successful": True}, # Single correct in multi-choice, full marks +4
+        {"question_id": "JA013", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A","B","C"], "predicted_answer": ["a","d"], "api_call_successful": True}, # One correct, one incorrect -> -2
+    ]
+
+    exam_summary = calculate_exam_scores(test_results_exam)
+    print("\nExam Score Summary:")
+    import json
+    print(json.dumps(exam_summary, indent=2, sort_keys=True))
+
+    # Basic assertions - can be expanded
+    assert exam_summary["overall_score"] == (4-1+0+0+0) + (4-1) + (4+0+0) + (3-1) + (4+0) + (4+2+3+1-2-2+0+4-2)
+    assert exam_summary["overall_correct_full"] == 7
+    assert exam_summary["overall_partial_correct"] == 3
+    assert exam_summary["overall_incorrect_choice"] == 8
+    assert exam_summary["overall_skipped"] == 2
+    assert exam_summary["overall_api_parse_failures"] == 3 # N004, N005, JM005
+
+    assert exam_summary["section_breakdown"]["Physics"]["score"] == (4-1) + (4+0) + (4+0) - 2 # N001,N002 + JM003,JM004 + JA003,JA004 + JA013
+    assert exam_summary["section_breakdown"]["Chemistry"]["score"] == (0+0) + (0) + (4+2+3+1-2-2+0) # N003,N004 + JM005 + JA005-JA011
+    assert exam_summary["section_breakdown"]["Botany"]["score"] == 0 # N005 (parse fail = 0 penalty)
+    assert exam_summary["section_breakdown"]["Maths"]["score"] == (4-1) + (3-1) + 4 # JM001,JM002 + JA001,JA002 + JA012
+
+    print("\nEvaluation tests completed.")

src/llm_interface.py

@@ -0,0 +1,283 @@
+import base64
+import requests
+import time
+import logging
+from io import BytesIO
+from PIL import Image
+from tenacity import retry, stop_after_attempt, wait_exponential, retry_if_exception_type
+
+from utils import parse_llm_answer
+from prompts import (
+    INITIAL_PROMPT_TEMPLATE,
+    REPROMPT_PROMPT_TEMPLATE,
+    get_answer_format_instruction,
+    get_example_instruction,
+    get_specific_instructions_reprompt,
+    get_reprompt_example_instruction
+)
+
+# Configure logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+
+# OpenRouter API endpoint
+OPENROUTER_API_ENDPOINT = "https://openrouter.ai/api/v1/chat/completions"
+
+# Define exceptions for retry logic
+RETRYABLE_EXCEPTIONS = (
+    requests.exceptions.Timeout,
+    requests.exceptions.ConnectionError,
+    requests.exceptions.RequestException # Catch broader request errors for retries
+)
+
+# Define status codes that warrant a retry
+RETRYABLE_STATUS_CODES = {429, 500, 502, 503, 504}
+
+# Retry decorator configuration
+retry_config = dict(
+    stop=stop_after_attempt(3),  # Retry up to 3 times
+    wait=wait_exponential(multiplier=1, min=2, max=10),  # Exponential backoff: 2s, 4s, 8s...
+    retry=(retry_if_exception_type(RETRYABLE_EXCEPTIONS)) # Retry on specific exceptions
+    # We will handle status code retries manually within the function for more control
+)
+
+def encode_image_to_base64(image: Image.Image) -> str:
+    """Encodes a PIL Image object to a base64 string."""
+    buffered = BytesIO()
+    # Ensure image is in RGB format for broad compatibility
+    if image.mode != 'RGB':
+        image = image.convert('RGB')
+    image.save(buffered, format="JPEG") # Save as JPEG for potentially smaller size
+    img_str = base64.b64encode(buffered.getvalue()).decode('utf-8')
+    return img_str
+
+def construct_reprompt_prompt(previous_raw_response: str, question_type: str) -> list:
+    """Constructs the message list for a re-prompt API call based on question_type."""
+    specific_instructions = get_specific_instructions_reprompt(question_type)
+    reprompt_example_instruction = get_reprompt_example_instruction(question_type)
+
+    prompt_text = REPROMPT_PROMPT_TEMPLATE.format(
+        previous_raw_response=previous_raw_response,
+        question_type=question_type,
+        specific_instructions=specific_instructions,
+        reprompt_example_instruction=reprompt_example_instruction
+    )
+    messages = [{"role": "user", "content": prompt_text}]
+    return messages
+
+
+def construct_initial_prompt(base64_image: str, exam_name: str, exam_year: str, question_type: str) -> list:
+    """Constructs the initial message list with image for the OpenRouter API call, tailored by question_type."""
+    
+    answer_format_instruction = get_answer_format_instruction(question_type)
+    example_instruction = get_example_instruction(question_type)
+
+    prompt_text = INITIAL_PROMPT_TEMPLATE.format(
+        exam_name=exam_name,
+        exam_year=exam_year,
+        question_type=question_type,
+        answer_format_instruction=answer_format_instruction,
+        example_instruction=example_instruction
+    )
+
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {"type": "text", "text": prompt_text},
+                {
+                    "type": "image_url",
+                    "image_url": {
+                        "url": f"data:image/jpeg;base64,{base64_image}"
+                    }
+                }
+            ]
+        }
+    ]
+    return messages
+
+@retry(**retry_config)
+def get_openrouter_prediction(
+    model_identifier: str,
+    api_key: str,
+    image: Image.Image | None = None, # Image is now optional
+    previous_raw_response: str | None = None, # Added for re-prompting
+    exam_name: str | None = None,
+    exam_year: str | None = None,
+    question_type: str = "MCQ_SINGLE_CORRECT", # New parameter with default
+    max_tokens: int = 100,
+    request_timeout: int = 60,
+    temperature: float = 0.0,
+) -> tuple[list[str] | str | None, str | None, dict | None]:
+    """
+    Gets a prediction from an OpenRouter model. Handles initial image prompts and text-only re-prompts.
+
+    Args:
+        model_identifier (str): The OpenRouter model identifier (e.g., "openai/gpt-4o").
+        api_key (str): The OpenRouter API key.
+        image (Image.Image | None): The question image (for initial prompt). Default None.
+        previous_raw_response (str | None): The raw response from a previous failed parse attempt (for re-prompt). Default None.
+        exam_name (str | None): The name of the exam (e.g., "NEET", "JEE"). Required if 'image' is provided for initial prompt.
+        exam_year (str | None): The year of the exam. Required if 'image' is provided for initial prompt.
+        question_type (str): Type of question, e.g., "MCQ_SINGLE_CORRECT", "MCQ_MULTIPLE_CORRECT", "INTEGER".
+        max_tokens (int): Max tokens for the response.
+        request_timeout (int): Timeout for the API request in seconds.
+        temperature (float): Sampling temperature. 0.0 for deterministic output.
+
+    Returns:
+        tuple[list[str] | str | None, str | None, dict | None]: A tuple containing:
+            - The parsed answer as a list of strings, the string "SKIP", or None if failed.
+            - The raw response text from the LLM (or None if API call failed).
+            - Response metadata dict (tokens, latency, model version, cost) or None.
+
+    Raises:
+        ValueError: If arguments are inconsistent (e.g., image provided without exam details for initial prompt).
+        requests.exceptions.RequestException: If the API call fails after retries.
+    """
+    logging.info(f"Requesting prediction from model: {model_identifier} for question_type: {question_type}")
+
+    if image is not None and previous_raw_response is None:
+        # Initial prompt with image
+        if not exam_name or not exam_year: # exam_name and exam_year are crucial for initial prompt context
+            raise ValueError("'exam_name' and 'exam_year' must be provided when 'image' is specified for an initial prompt.")
+        logging.debug(f"Constructing initial prompt with image for {exam_name} {exam_year}, type: {question_type}.")
+        base64_image = encode_image_to_base64(image)
+        messages = construct_initial_prompt(base64_image, exam_name, exam_year, question_type)
+    elif image is None and previous_raw_response is not None:
+        # Re-prompt based on previous response
+        logging.debug(f"Constructing re-prompt based on previous response for type: {question_type}.")
+        messages = construct_reprompt_prompt(previous_raw_response, question_type)
+    else:
+        # This condition means either both image and previous_raw_response are None, or both are provided.
+        # The latter (both provided) is ambiguous for which prompt to use.
+        # The former (both None) means no input to act on.
+        raise ValueError("Provide 'image' (with 'exam_name' and 'exam_year') for an initial call, OR 'previous_raw_response' for a re-prompt. Not neither or both.")
+
+
+    try:
+        headers = {
+            "Authorization": f"Bearer {api_key}",
+            "Content-Type": "application/json"
+        }
+
+        data = {
+            "model": model_identifier,
+            "messages": messages,
+            "max_tokens": max_tokens,
+            "temperature": temperature,
+            "usage": {"include": True},
+        }
+
+        request_start_time = time.time()
+        response = requests.post(
+            OPENROUTER_API_ENDPOINT,
+            headers=headers,
+            json=data,
+            timeout=request_timeout
+        )
+        response_latency_ms = round((time.time() - request_start_time) * 1000)
+
+        if response.status_code in RETRYABLE_STATUS_CODES:
+            logging.warning(f"Received retryable status code {response.status_code} from {model_identifier} for {question_type}. Retrying might occur.")
+            response.raise_for_status()
+
+        if not response.ok:
+            logging.error(f"API Error for model {model_identifier} ({question_type}): Status {response.status_code} - {response.text}")
+            return None, None, None
+
+        response_json = response.json()
+        raw_response_text = response_json.get("choices", [{}])[0].get("message", {}).get("content")
+
+        # Extract response metadata
+        usage = response_json.get("usage", {})
+        response_metadata = {
+            "generation_id": response_json.get("id"),
+            "prompt_tokens": usage.get("prompt_tokens"),
+            "completion_tokens": usage.get("completion_tokens"),
+            "total_tokens": usage.get("total_tokens"),
+            "cost": usage.get("cost"),
+            "response_latency_ms": response_latency_ms,
+            "model_version": response_json.get("model"),
+            "temperature": temperature,
+        }
+
+        if not raw_response_text:
+            logging.warning(f"Empty response content received from model: {model_identifier} for {question_type}")
+            return None, None, response_metadata
+
+        logging.info(f"Raw response received from {model_identifier} ({question_type}): '{raw_response_text[:100]}...'")
+        # Pass question_type to parse_llm_answer
+        parsed_answer = parse_llm_answer(raw_response_text, question_type=question_type)
+
+        if parsed_answer is None:
+             logging.warning(f"Failed to parse answer from model {model_identifier} for {question_type}.")
+
+        return parsed_answer, raw_response_text, response_metadata
+
+    except requests.exceptions.Timeout as e:
+        logging.error(f"Request timed out for model {model_identifier} ({question_type}): {e}")
+        raise
+    except requests.exceptions.RequestException as e:
+        logging.error(f"Request failed for model {model_identifier} ({question_type}): {e}")
+        raise
+    except Exception as e:
+        logging.error(f"An unexpected error occurred for model {model_identifier} ({question_type}): {e}")
+        return None, None, None
+
+# Example Usage (requires a valid API key in .env and Pillow/requests/tenacity installed)
+if __name__ == '__main__':
+    from src.utils import load_api_key
+    try:
+        dummy_image = Image.new('RGB', (60, 30), color = 'black')
+        api_key = load_api_key()
+        test_model = "anthropic/claude-3-haiku" 
+
+        print(f"\n--- Testing with model: {test_model} ---")
+
+        # Test Case 1: Initial call - MCQ_SINGLE_CORRECT
+        print("\nTest Case 1: Initial - MCQ_SINGLE_CORRECT")
+        parsed_ans_1, raw_resp_1, meta_1 = get_openrouter_prediction(
+            model_identifier=test_model, api_key=api_key, image=dummy_image,
+            exam_name="DUMMY_EXAM", exam_year="2024", question_type="MCQ_SINGLE_CORRECT"
+        )
+        print(f"Parsed: {parsed_ans_1}, Raw: {raw_resp_1[:60] if raw_resp_1 else None}...")
+        print(f"Metadata: {meta_1}")
+
+        # Test Case 2: Initial call - MCQ_MULTIPLE_CORRECT
+        print("\nTest Case 2: Initial - MCQ_MULTIPLE_CORRECT")
+        parsed_ans_2, raw_resp_2, meta_2 = get_openrouter_prediction(
+            model_identifier=test_model, api_key=api_key, image=dummy_image,
+            exam_name="DUMMY_EXAM", exam_year="2024", question_type="MCQ_MULTIPLE_CORRECT"
+        )
+        print(f"Parsed: {parsed_ans_2}, Raw: {raw_resp_2[:60] if raw_resp_2 else None}...")
+
+        # Test Case 3: Initial call - INTEGER
+        print("\nTest Case 3: Initial - INTEGER")
+        parsed_ans_3, raw_resp_3, meta_3 = get_openrouter_prediction(
+            model_identifier=test_model, api_key=api_key, image=dummy_image,
+            exam_name="DUMMY_EXAM", exam_year="2024", question_type="INTEGER"
+        )
+        print(f"Parsed: {parsed_ans_3}, Raw: {raw_resp_3[:60] if raw_resp_3 else None}...")
+
+
+        # Test Case 4: Re-prompt - MCQ_SINGLE_CORRECT (simulating bad initial response)
+        print("\nTest Case 4: Re-prompt - MCQ_SINGLE_CORRECT")
+        bad_initial_resp_mcq_single = "<answer>1 2</answer> This is some extra text."
+        reprompt_ans_4, reprompt_raw_4, meta_4 = get_openrouter_prediction(
+            model_identifier=test_model, api_key=api_key,
+            previous_raw_response=bad_initial_resp_mcq_single, question_type="MCQ_SINGLE_CORRECT"
+        )
+        print(f"Parsed: {reprompt_ans_4}, Raw: {reprompt_raw_4[:60] if reprompt_raw_4 else None}...")
+
+        # Test Case 5: Re-prompt - MCQ_MULTIPLE_CORRECT (simulating bad initial response)
+        print("\nTest Case 5: Re-prompt - MCQ_MULTIPLE_CORRECT")
+        bad_initial_resp_mcq_multi = "The answer is <answer>option 1 and 4</answer> because reasons."
+        reprompt_ans_5, reprompt_raw_5, meta_5 = get_openrouter_prediction(
+            model_identifier=test_model, api_key=api_key,
+            previous_raw_response=bad_initial_resp_mcq_multi, question_type="MCQ_MULTIPLE_CORRECT"
+        )
+        print(f"Parsed: {reprompt_ans_5}, Raw: {reprompt_raw_5[:60] if reprompt_raw_5 else None}...")
+
+    except ValueError as e:
+        print(f"Setup or Argument Error: {e}")
+    except Exception as e:
+        print(f"Runtime Error during example execution: {e}")

src/prompts.py

@@ -0,0 +1,69 @@
+# System prompts for LLM interaction
+
+# --- Initial Prompt Components ---
+
+ANSWER_FORMAT_INSTRUCTIONS = {
+    "MCQ_SINGLE_CORRECT": "Pick the single correct option (e.g., A, B, 1, 2).",
+    "INTEGER": "Find the numerical answer (integer or decimal).",
+    "MCQ_MULTIPLE_CORRECT": "Pick all correct options (e.g., A,C or 1,3), comma-separated.",
+    "DEFAULT": "Answer based on the question's format."
+}
+
+EXAMPLE_INSTRUCTIONS = {
+    "MCQ_SINGLE_CORRECT": "- <answer>A</answer>\n- <answer>2</answer>",
+    "INTEGER": "- <answer>5</answer>\n- <answer>12.75</answer>",
+    "MCQ_MULTIPLE_CORRECT": "- <answer>A,C</answer>\n- <answer>1,3</answer>\n- <answer>B</answer>",
+    "DEFAULT": "- <answer>Your Answer</answer>"
+}
+
+INITIAL_PROMPT_TEMPLATE = """Look at this {exam_name} {exam_year} exam question ({question_type}).
+
+{answer_format_instruction}
+
+Respond with ONLY an <answer> tag. No other text.
+
+{example_instruction}
+- Unsure: <answer>SKIP</answer>"""
+
+
+# --- Reprompt Components ---
+
+SPECIFIC_INSTRUCTIONS_REPROMPT = {
+    "MCQ_SINGLE_CORRECT": "provide the correct option (e.g., A or 2)",
+    "INTEGER": "provide the numerical answer",
+    "MCQ_MULTIPLE_CORRECT": "provide all correct options comma-separated (e.g., A,C)",
+    "DEFAULT": "provide the answer"
+}
+
+REPROMPT_EXAMPLE_INSTRUCTIONS = {
+    "MCQ_SINGLE_CORRECT": "<answer>A</answer>",
+    "MCQ_MULTIPLE_CORRECT": "<answer>A,C</answer>",
+    "INTEGER": "<answer>42</answer>",
+    "DEFAULT": "<answer>Your Answer</answer>",
+}
+
+REPROMPT_PROMPT_TEMPLATE = """Your previous response was not in the correct format.
+
+Previous response:
+{previous_raw_response}
+
+This is a '{question_type}' question. Please {specific_instructions} in an <answer> tag.
+
+Example: {reprompt_example_instruction}
+Unsure: <answer>SKIP</answer>
+
+Respond with ONLY the <answer> tag. No other text."""
+
+# --- Helper functions to get instructions ---
+
+def get_answer_format_instruction(question_type: str) -> str:
+    return ANSWER_FORMAT_INSTRUCTIONS.get(question_type, ANSWER_FORMAT_INSTRUCTIONS["DEFAULT"])
+
+def get_example_instruction(question_type: str) -> str:
+    return EXAMPLE_INSTRUCTIONS.get(question_type, EXAMPLE_INSTRUCTIONS["DEFAULT"])
+
+def get_specific_instructions_reprompt(question_type: str) -> str:
+    return SPECIFIC_INSTRUCTIONS_REPROMPT.get(question_type, SPECIFIC_INSTRUCTIONS_REPROMPT["DEFAULT"])
+
+def get_reprompt_example_instruction(question_type: str) -> str:
+    return REPROMPT_EXAMPLE_INSTRUCTIONS.get(question_type, REPROMPT_EXAMPLE_INSTRUCTIONS["DEFAULT"])

src/utils.py

@@ -0,0 +1,194 @@
+import os
+import re
+from dotenv import load_dotenv
+import logging
+
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+
+def load_api_key(key_name="OPENROUTER_API_KEY"):
+    """
+    Loads the specified API key from environment variables.
+    Loads .env file first if it exists.
+
+    Args:
+        key_name (str): The name of the environment variable holding the API key.
+
+    Returns:
+        str: The API key.
+
+    Raises:
+        ValueError: If the API key environment variable is not set.
+    """
+    load_dotenv() # Load environment variables from .env file if it exists
+    api_key = os.getenv(key_name)
+    if not api_key:
+        raise ValueError(
+            f"'{key_name}' environment variable not set. Please create a .env file or set the variable."
+        )
+    return api_key
+
+
+def parse_llm_answer(response_text: str, question_type: str = "MCQ_SINGLE_CORRECT") -> list[str] | str | None:
+    """
+    Parses the LLM response text to extract answers within <answer> tags.
+    The parsing logic adapts based on the question_type.
+
+    Handles:
+    - MCQ_SINGLE_CORRECT: Single option identifier (integer like "1", "2" or letter like "A", "B").
+    - INTEGER: Single numerical value, which can be an integer or a decimal (e.g., "5", "12.75", "0.5").
+    - MCQ_MULTIPLE_CORRECT: Multiple option identifiers (integers or letters), comma-separated.
+    - The specific string "SKIP" for skipped questions (case-insensitive content within tags).
+    - Potential newlines and varied spacing within the tags.
+
+    Args:
+        response_text (str): The raw text response from the LLM.
+        question_type (str): The type of question, e.g., "MCQ_SINGLE_CORRECT",
+                             "MCQ_MULTIPLE_CORRECT", "INTEGER".
+                             Defaults to "MCQ_SINGLE_CORRECT".
+
+    Returns:
+        list[str] | str | None:
+            - A list containing string answer(s) if found and valid.
+              (e.g., ["1"], ["A"], ["12.75"], ["A", "C"])
+            - The string "SKIP" if the response indicates a skip.
+            - None if parsing fails (no tag, invalid content, type mismatch, etc.).
+    """
+    if not response_text:
+        return None
+
+    # Check for exact SKIP response first (case-insensitive for the tag and content)
+    # Using regex to be more flexible with whitespace around SKIP
+    skip_match = re.search(r"<answer>\s*SKIP\s*</answer>", response_text, re.IGNORECASE)
+    if skip_match:
+        logging.info(f"Parsed answer as SKIP for question_type: {question_type}.")
+        return "SKIP"
+
+    match = re.search(r"<answer>(.*?)</answer>", response_text, re.DOTALL | re.IGNORECASE)
+
+    if not match:
+        logging.warning(f"Could not find <answer> tag in response for question_type: {question_type} in response: '{response_text[:200]}...'")
+        return None
+
+    extracted_content = match.group(1).strip()
+    if not extracted_content:
+        logging.warning(f"Found <answer> tag but content is empty for question_type: {question_type}.")
+        return None
+
+    potential_answers_str_list = [item.strip() for item in extracted_content.split(',')]
+    parsed_answers_final = []
+
+    for ans_str_raw in potential_answers_str_list:
+        ans_str = ans_str_raw.strip()
+        if not ans_str: # Skip empty strings that might result from "1, ,2" or trailing commas
+            continue
+
+        if question_type == "INTEGER":
+            try:
+                # Try to parse as float to validate it's a number (integer or decimal).
+                # The value is kept as a string to preserve original formatting (e.g., "0.50", "5.0")
+                # for exact string comparison with ground truth if needed, and for consistent type handling.
+                float(ans_str) 
+                parsed_answers_final.append(ans_str)
+            except ValueError:
+                logging.warning(f"Could not parse '{ans_str}' as a valid number (integer or decimal) for INTEGER type. Full content: '{extracted_content}'")
+                return None # If any part is not a valid number for INTEGER type, fail parsing.
+        
+        elif question_type in ["MCQ_SINGLE_CORRECT", "MCQ_MULTIPLE_CORRECT"]:
+            # For MCQs, accept single-digit numbers (1-9) or letters A-D.
+            if re.fullmatch(r"[1-9]", ans_str): # Single digit options
+                parsed_answers_final.append(ans_str)
+            elif re.fullmatch(r"[a-dA-D]", ans_str): # Letter options A-D
+                parsed_answers_final.append(ans_str.upper())
+            else:
+                logging.warning(f"Could not parse '{ans_str}' as a valid MCQ option (expected 1-9 or A-D). Full content: '{extracted_content}'")
+                return None
+        else: # Should not happen if question_type is validated before calling
+            logging.error(f"Unknown question_type '{question_type}' encountered in parse_llm_answer logic.")
+            return None
+
+
+    if not parsed_answers_final: # If list is empty after processing (e.g. content was just commas)
+        logging.warning(f"No valid answer items found after parsing content: '{extracted_content}' for question_type: {question_type}.")
+        return None
+
+    # Apply rules based on question_type for number of answers
+    if question_type in ["MCQ_SINGLE_CORRECT", "INTEGER"]:
+        if len(parsed_answers_final) == 1:
+            return parsed_answers_final # Returns list[str] with one element
+        else:
+            logging.warning(f"Expected single answer for {question_type}, but found {len(parsed_answers_final)} items: {parsed_answers_final}. Content: '{extracted_content}'")
+            return None
+    elif question_type == "MCQ_MULTIPLE_CORRECT":
+        # For multiple correct, any number of valid items is acceptable.
+        # Return them sorted and unique.
+        return sorted(list(set(parsed_answers_final)))
+    else:
+        # This case should ideally be caught by earlier checks or input validation.
+        logging.error(f"Unknown question_type '{question_type}' provided to parse_llm_answer at final stage.")
+        return None
+
+# Example Usage (for testing)
+if __name__ == '__main__':
+    test_cases = [
+        # MCQ_SINGLE_CORRECT (can be number or letter)
+        {"resp": "<answer>2</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": ["2"]},
+        {"resp": "<answer>B</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": ["B"]},
+        {"resp": "<answer> c </answer>", "type": "MCQ_SINGLE_CORRECT", "expected": ["C"]},
+        {"resp": "<answer>1,3</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": None}, # Fail: multiple for single
+        {"resp": "<answer>A,C</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": None}, # Fail: multiple for single
+        {"resp": "<answer>X</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": None}, # Fail: not A-D
+        {"resp": "<answer>5</answer>", "type": "MCQ_SINGLE_CORRECT", "expected": ["5"]}, # Accept single digits 1-9
+
+        # INTEGER (can be int or decimal string)
+        {"resp": "<answer>42</answer>", "type": "INTEGER", "expected": ["42"]},
+        {"resp": "<answer>0</answer>", "type": "INTEGER", "expected": ["0"]},
+        {"resp": "<answer>12.75</answer>", "type": "INTEGER", "expected": ["12.75"]},
+        {"resp": "<answer>0.5</answer>", "type": "INTEGER", "expected": ["0.5"]},
+        {"resp": "<answer>-5</answer>", "type": "INTEGER", "expected": ["-5"]}, # Assuming negative integers are valid if problem allows
+        {"resp": "<answer>5.00</answer>", "type": "INTEGER", "expected": ["5.00"]},
+        {"resp": "<answer>abc</answer>", "type": "INTEGER", "expected": None}, # Fail: not a number
+        {"resp": "<answer>1,2</answer>", "type": "INTEGER", "expected": None}, # Fail: multiple for single int
+
+        # MCQ_MULTIPLE_CORRECT (can be numbers or letters, mixed is not typical but parser allows if items are valid individually)
+        {"resp": "<answer>1,3</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["1", "3"]},
+        {"resp": "<answer>A,C</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["A", "C"]},
+        {"resp": "<answer> b , d </answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["B", "D"]},
+        {"resp": "<answer>2</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["2"]}, # Single is valid
+        {"resp": "<answer>D</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["D"]}, # Single is valid
+        {"resp": "<answer>1,C,3</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["1", "3", "C"]}, # Mixed, sorted
+        {"resp": "<answer>C,1,A,1</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["1", "A", "C"]}, # Unique and sorted
+        {"resp": "<answer>1,X,3</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": None}, # Fail: X not in A-D
+
+        # SKIP and general failures
+        {"resp": "<ANSWER>SKIP</ANSWER>", "type": "MCQ_SINGLE_CORRECT", "expected": "SKIP"},
+        {"resp": "  <answer>  SKIP  </answer>  ", "type": "INTEGER", "expected": "SKIP"},
+        {"resp": "No answer tag here.", "type": "MCQ_SINGLE_CORRECT", "expected": None},
+        {"resp": "<answer></answer>", "type": "MCQ_SINGLE_CORRECT", "expected": None},
+        {"resp": "<answer>  </answer>", "type": "MCQ_SINGLE_CORRECT", "expected": None},
+        {"resp": None, "type": "MCQ_SINGLE_CORRECT", "expected": None},
+        {"resp": "", "type": "MCQ_SINGLE_CORRECT", "expected": None},
+        {"resp": "<answer>5</answer>", "type": "UNKNOWN_TYPE", "expected": None}, # Unknown type
+        {"resp": "<answer>1,,2</answer>", "type": "MCQ_MULTIPLE_CORRECT", "expected": ["1", "2"]}, # Handles empty item from double comma
+    ]
+
+    print("\n--- Testing parse_llm_answer (revised) ---")
+    all_passed = True
+    for i, case in enumerate(test_cases):
+        parsed = parse_llm_answer(case["resp"], case["type"])
+        if parsed == case["expected"]:
+            print(f"Test {i+1} PASSED: Response: '{str(case['resp'])[:50]}...', Type: {case['type']} -> Parsed: {parsed}")
+        else:
+            print(f"Test {i+1} FAILED: Response: '{str(case['resp'])[:50]}...', Type: {case['type']} -> Parsed: {parsed} (Expected: {case['expected']})")
+            all_passed = False
+    
+    if all_passed:
+        print("\nAll revised parse_llm_answer tests passed!")
+    else:
+        print("\nSome revised parse_llm_answer tests FAILED.")
+
+    # Test API key loading (will raise error if .env or env var not set)
+    # try:
+    #     key = load_api_key()
+    #     print(f"\nAPI Key loaded successfully (partially hidden): {key[:4]}...{key[-4:]}")
+    # except ValueError as e:
+    #     print(f"\nError loading API key: {e}")

tests/test_evaluation.py

@@ -0,0 +1,291 @@
+"""Tests for src/evaluation.py - scoring and evaluation functions."""
+
+import sys
+import os
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+
+import pytest
+from evaluation import (
+    calculate_single_question_score_details,
+    calculate_exam_scores,
+    calculate_max_score_for_question,
+    is_within_range,
+)
+
+
+# --- calculate_single_question_score_details ---
+
+class TestSingleQuestionScoring:
+    """Tests for individual question scoring across exam types."""
+
+    # NEET MCQ
+    def test_neet_correct(self):
+        result = calculate_single_question_score_details({
+            "question_id": "N001", "exam_name": "NEET",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["1"], "predicted_answer": ["1"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 4
+        assert result["evaluation_status"] == "correct"
+
+    def test_neet_incorrect(self):
+        result = calculate_single_question_score_details({
+            "question_id": "N002", "exam_name": "NEET",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["D"], "predicted_answer": ["B"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == -1
+        assert result["evaluation_status"] == "incorrect"
+
+    def test_neet_skipped(self):
+        result = calculate_single_question_score_details({
+            "question_id": "N003", "exam_name": "NEET",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["4"], "predicted_answer": "SKIP",
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 0
+        assert result["evaluation_status"] == "skipped"
+
+    def test_api_failure_no_penalty(self):
+        """API/parse failures should score 0, not -1."""
+        result = calculate_single_question_score_details({
+            "question_id": "N004", "exam_name": "NEET",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["C"], "predicted_answer": None,
+            "api_call_successful": False,
+        })
+        assert result["marks_awarded"] == 0
+        assert result["evaluation_status"] == "failure_api_or_parse"
+
+    def test_parse_failure_no_penalty(self):
+        """Parse failure (api_success=True but pred=None) should score 0."""
+        result = calculate_single_question_score_details({
+            "question_id": "N005", "exam_name": "NEET",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["A"], "predicted_answer": None,
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 0
+        assert result["evaluation_status"] == "failure_api_or_parse"
+
+    # JEE Main
+    def test_jee_main_mcq_correct(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JM001", "exam_name": "JEE_MAIN",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["2"], "predicted_answer": ["2"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 4
+
+    def test_jee_main_integer_correct(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JM003", "exam_name": "JEE_MAIN",
+            "question_type": "INTEGER",
+            "ground_truth": ["5"], "predicted_answer": ["5"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 4
+
+    def test_jee_main_integer_incorrect_no_penalty(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JM004", "exam_name": "JEE_MAIN",
+            "question_type": "INTEGER",
+            "ground_truth": ["10"], "predicted_answer": ["8"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 0
+
+    # JEE Advanced
+    def test_jee_advanced_mcq_single_correct(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA001", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_SINGLE_CORRECT",
+            "ground_truth": ["A"], "predicted_answer": ["a"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 3  # JEE Advanced MCQ single = +3
+
+    def test_jee_advanced_mcq_multiple_all_correct(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA005", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_MULTIPLE_CORRECT",
+            "ground_truth": ["A", "C"], "predicted_answer": ["a", "c"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 4
+        assert result["evaluation_status"] == "correct_full"
+
+    def test_jee_advanced_mcq_multiple_partial_2_of_3(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA006", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_MULTIPLE_CORRECT",
+            "ground_truth": ["A", "B", "C"], "predicted_answer": ["a", "b"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 2
+        assert result["evaluation_status"] == "partial_2_of_3_plus"
+
+    def test_jee_advanced_mcq_multiple_partial_3_of_4(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA007", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_MULTIPLE_CORRECT",
+            "ground_truth": ["A", "B", "C", "D"], "predicted_answer": ["a", "b", "c"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 3
+        assert result["evaluation_status"] == "partial_3_of_4"
+
+    def test_jee_advanced_mcq_multiple_partial_1_of_2(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA008", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_MULTIPLE_CORRECT",
+            "ground_truth": ["A", "B"], "predicted_answer": ["a"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 1
+        assert result["evaluation_status"] == "partial_1_of_2_plus"
+
+    def test_jee_advanced_mcq_multiple_incorrect_chosen(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA009", "exam_name": "JEE_ADVANCED",
+            "question_type": "MCQ_MULTIPLE_CORRECT",
+            "ground_truth": ["A", "C"], "predicted_answer": ["a", "b"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == -2
+        assert result["evaluation_status"] == "incorrect_negative"
+
+    def test_jee_advanced_integer_with_range(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA_RANGE", "exam_name": "JEE_ADVANCED",
+            "question_type": "INTEGER",
+            "ground_truth": [["0.7", "0.8"]],
+            "predicted_answer": ["0.75"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 4
+        assert result["evaluation_status"] == "correct"
+
+    def test_jee_advanced_integer_outside_range(self):
+        result = calculate_single_question_score_details({
+            "question_id": "JA_RANGE2", "exam_name": "JEE_ADVANCED",
+            "question_type": "INTEGER",
+            "ground_truth": [["0.7", "0.8"]],
+            "predicted_answer": ["0.9"],
+            "api_call_successful": True,
+        })
+        assert result["marks_awarded"] == 0
+        assert result["evaluation_status"] == "incorrect"
+
+
+# --- is_within_range ---
+
+class TestIsWithinRange:
+    def test_within_range(self):
+        assert is_within_range("0.75", "0.7", "0.8") is True
+
+    def test_at_lower_bound(self):
+        assert is_within_range("0.7", "0.7", "0.8") is True
+
+    def test_at_upper_bound(self):
+        assert is_within_range("0.8", "0.7", "0.8") is True
+
+    def test_below_range(self):
+        assert is_within_range("0.5", "0.7", "0.8") is False
+
+    def test_above_range(self):
+        assert is_within_range("0.9", "0.7", "0.8") is False
+
+    def test_invalid_value(self):
+        assert is_within_range("abc", "0.7", "0.8") is False
+
+
+# --- calculate_max_score_for_question ---
+
+class TestMaxScore:
+    def test_neet_mcq(self):
+        assert calculate_max_score_for_question("NEET", "MCQ_SINGLE_CORRECT") == 4
+
+    def test_jee_main_mcq(self):
+        assert calculate_max_score_for_question("JEE_MAIN", "MCQ_SINGLE_CORRECT") == 4
+
+    def test_jee_main_integer(self):
+        assert calculate_max_score_for_question("JEE_MAIN", "INTEGER") == 4
+
+    def test_jee_advanced_mcq_single(self):
+        assert calculate_max_score_for_question("JEE_ADVANCED", "MCQ_SINGLE_CORRECT") == 3
+
+    def test_jee_advanced_mcq_multiple(self):
+        assert calculate_max_score_for_question("JEE_ADVANCED", "MCQ_MULTIPLE_CORRECT") == 4
+
+    def test_jee_advanced_integer(self):
+        assert calculate_max_score_for_question("JEE_ADVANCED", "INTEGER") == 4
+
+    def test_unknown(self):
+        assert calculate_max_score_for_question("UNKNOWN", "UNKNOWN") == 0
+
+
+# --- calculate_exam_scores (integration) ---
+
+class TestExamScoresIntegration:
+    """Integration test using the full test dataset from evaluation.py __main__."""
+
+    @pytest.fixture
+    def test_results(self):
+        return [
+            {"question_id": "N001", "subject": "Physics", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["1"], "predicted_answer": ["1"], "api_call_successful": True},
+            {"question_id": "N002", "subject": "Physics", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["D"], "predicted_answer": ["B"], "api_call_successful": True},
+            {"question_id": "N003", "subject": "Chemistry", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["4"], "predicted_answer": "SKIP", "api_call_successful": True},
+            {"question_id": "N004", "subject": "Chemistry", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["C"], "predicted_answer": None, "api_call_successful": False},
+            {"question_id": "N005", "subject": "Botany", "exam_name": "NEET", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["A"], "predicted_answer": None, "api_call_successful": True},
+            {"question_id": "JM001", "subject": "Maths", "exam_name": "JEE_MAIN", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["2"], "predicted_answer": ["2"], "api_call_successful": True},
+            {"question_id": "JM002", "subject": "Maths", "exam_name": "JEE_MAIN", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["C"], "predicted_answer": ["a"], "api_call_successful": True},
+            {"question_id": "JM003", "subject": "Physics", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["5"], "predicted_answer": ["5"], "api_call_successful": True},
+            {"question_id": "JM004", "subject": "Physics", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["10"], "predicted_answer": ["8"], "api_call_successful": True},
+            {"question_id": "JM005", "subject": "Chemistry", "exam_name": "JEE_MAIN", "question_type": "INTEGER", "ground_truth": ["7"], "predicted_answer": None, "api_call_successful": True},
+            {"question_id": "JA001", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["A"], "predicted_answer": ["a"], "api_call_successful": True},
+            {"question_id": "JA002", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_SINGLE_CORRECT", "ground_truth": ["B"], "predicted_answer": ["C"], "api_call_successful": True},
+            {"question_id": "JA003", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "ground_truth": ["12"], "predicted_answer": ["12"], "api_call_successful": True},
+            {"question_id": "JA004", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "INTEGER", "ground_truth": ["0"], "predicted_answer": ["1"], "api_call_successful": True},
+            {"question_id": "JA005", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["a", "c"], "api_call_successful": True},
+            {"question_id": "JA006", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B", "C"], "predicted_answer": ["a", "b"], "api_call_successful": True},
+            {"question_id": "JA007", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B", "C", "D"], "predicted_answer": ["a", "b", "c"], "api_call_successful": True},
+            {"question_id": "JA008", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B"], "predicted_answer": ["a"], "api_call_successful": True},
+            {"question_id": "JA009", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["a", "b"], "api_call_successful": True},
+            {"question_id": "JA010", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": ["b", "d"], "api_call_successful": True},
+            {"question_id": "JA011", "subject": "Chemistry", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "C"], "predicted_answer": "SKIP", "api_call_successful": True},
+            {"question_id": "JA012", "subject": "Maths", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A"], "predicted_answer": ["a"], "api_call_successful": True},
+            {"question_id": "JA013", "subject": "Physics", "exam_name": "JEE_ADVANCED", "question_type": "MCQ_MULTIPLE_CORRECT", "ground_truth": ["A", "B", "C"], "predicted_answer": ["a", "d"], "api_call_successful": True},
+        ]
+
+    def test_overall_score(self, test_results):
+        summary = calculate_exam_scores(test_results)
+        # NEET: 4-1+0+0+0=3, JEE Main MCQ: 4-1=3, JEE Main INT: 4+0+0=4,
+        # JEE Adv MCQ Single: 3-1=2, JEE Adv INT: 4+0=4,
+        # JEE Adv MCQ Multi: 4+2+3+1-2-2+0+4-2=8 => Total=24
+        assert summary["overall_score"] == 24
+
+    def test_overall_counts(self, test_results):
+        summary = calculate_exam_scores(test_results)
+        assert summary["overall_correct_full"] == 7
+        assert summary["overall_partial_correct"] == 3
+        assert summary["overall_incorrect_choice"] == 8
+        assert summary["overall_skipped"] == 2
+        assert summary["overall_api_parse_failures"] == 3
+
+    def test_section_scores(self, test_results):
+        summary = calculate_exam_scores(test_results)
+        sections = summary["section_breakdown"]
+        assert sections["Physics"]["score"] == 9    # (4-1)+(4+0)+(4+0)-2
+        assert sections["Chemistry"]["score"] == 6  # (0+0)+(0)+(4+2+3+1-2-2+0)
+        assert sections["Botany"]["score"] == 0     # parse fail = 0
+        assert sections["Maths"]["score"] == 9      # (4-1)+(3-1)+4
+
+    def test_empty_results(self):
+        summary = calculate_exam_scores([])
+        assert "error" in summary

tests/test_utils.py

@@ -0,0 +1,133 @@
+"""Tests for src/utils.py - parse_llm_answer function."""
+
+import sys
+import os
+
+# Add src to path so we can import utils directly
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+
+import pytest
+from utils import parse_llm_answer
+
+
+# --- MCQ_SINGLE_CORRECT ---
+
+class TestMCQSingleCorrect:
+    def test_numeric_option(self):
+        assert parse_llm_answer("<answer>2</answer>", "MCQ_SINGLE_CORRECT") == ["2"]
+
+    def test_letter_option(self):
+        assert parse_llm_answer("<answer>B</answer>", "MCQ_SINGLE_CORRECT") == ["B"]
+
+    def test_lowercase_letter_uppercased(self):
+        assert parse_llm_answer("<answer> c </answer>", "MCQ_SINGLE_CORRECT") == ["C"]
+
+    def test_multiple_numeric_rejected(self):
+        assert parse_llm_answer("<answer>1,3</answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_multiple_letter_rejected(self):
+        assert parse_llm_answer("<answer>A,C</answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_invalid_letter_rejected(self):
+        """Letters beyond A-D should be rejected."""
+        assert parse_llm_answer("<answer>X</answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_single_digit_5_accepted(self):
+        """Single digits 1-9 are accepted as valid options."""
+        assert parse_llm_answer("<answer>5</answer>", "MCQ_SINGLE_CORRECT") == ["5"]
+
+    def test_multi_digit_rejected(self):
+        """Multi-digit numbers like 99 should be rejected."""
+        assert parse_llm_answer("<answer>99</answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_zero_rejected(self):
+        """0 is not a valid MCQ option."""
+        assert parse_llm_answer("<answer>0</answer>", "MCQ_SINGLE_CORRECT") is None
+
+
+# --- INTEGER ---
+
+class TestInteger:
+    def test_positive_integer(self):
+        assert parse_llm_answer("<answer>42</answer>", "INTEGER") == ["42"]
+
+    def test_zero(self):
+        assert parse_llm_answer("<answer>0</answer>", "INTEGER") == ["0"]
+
+    def test_decimal(self):
+        assert parse_llm_answer("<answer>12.75</answer>", "INTEGER") == ["12.75"]
+
+    def test_small_decimal(self):
+        assert parse_llm_answer("<answer>0.5</answer>", "INTEGER") == ["0.5"]
+
+    def test_negative_integer(self):
+        assert parse_llm_answer("<answer>-5</answer>", "INTEGER") == ["-5"]
+
+    def test_trailing_zeros(self):
+        assert parse_llm_answer("<answer>5.00</answer>", "INTEGER") == ["5.00"]
+
+    def test_text_rejected(self):
+        assert parse_llm_answer("<answer>abc</answer>", "INTEGER") is None
+
+    def test_multiple_values_rejected(self):
+        assert parse_llm_answer("<answer>1,2</answer>", "INTEGER") is None
+
+
+# --- MCQ_MULTIPLE_CORRECT ---
+
+class TestMCQMultipleCorrect:
+    def test_two_numeric_options(self):
+        assert parse_llm_answer("<answer>1,3</answer>", "MCQ_MULTIPLE_CORRECT") == ["1", "3"]
+
+    def test_two_letter_options(self):
+        assert parse_llm_answer("<answer>A,C</answer>", "MCQ_MULTIPLE_CORRECT") == ["A", "C"]
+
+    def test_lowercase_uppercased_with_spaces(self):
+        assert parse_llm_answer("<answer> b , d </answer>", "MCQ_MULTIPLE_CORRECT") == ["B", "D"]
+
+    def test_single_numeric_valid(self):
+        assert parse_llm_answer("<answer>2</answer>", "MCQ_MULTIPLE_CORRECT") == ["2"]
+
+    def test_single_letter_valid(self):
+        assert parse_llm_answer("<answer>D</answer>", "MCQ_MULTIPLE_CORRECT") == ["D"]
+
+    def test_mixed_sorted(self):
+        assert parse_llm_answer("<answer>1,C,3</answer>", "MCQ_MULTIPLE_CORRECT") == ["1", "3", "C"]
+
+    def test_duplicates_removed(self):
+        assert parse_llm_answer("<answer>C,1,A,1</answer>", "MCQ_MULTIPLE_CORRECT") == ["1", "A", "C"]
+
+    def test_invalid_letter_rejected(self):
+        """X is not in A-D, so entire parse fails."""
+        assert parse_llm_answer("<answer>1,X,3</answer>", "MCQ_MULTIPLE_CORRECT") is None
+
+    def test_double_comma_handled(self):
+        assert parse_llm_answer("<answer>1,,2</answer>", "MCQ_MULTIPLE_CORRECT") == ["1", "2"]
+
+
+# --- SKIP and edge cases ---
+
+class TestSkipAndEdgeCases:
+    def test_skip_uppercase_tag(self):
+        assert parse_llm_answer("<ANSWER>SKIP</ANSWER>", "MCQ_SINGLE_CORRECT") == "SKIP"
+
+    def test_skip_with_whitespace(self):
+        assert parse_llm_answer("  <answer>  SKIP  </answer>  ", "INTEGER") == "SKIP"
+
+    def test_no_tag(self):
+        assert parse_llm_answer("No answer tag here.", "MCQ_SINGLE_CORRECT") is None
+
+    def test_empty_tag(self):
+        assert parse_llm_answer("<answer></answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_whitespace_only_tag(self):
+        assert parse_llm_answer("<answer>  </answer>", "MCQ_SINGLE_CORRECT") is None
+
+    def test_none_input(self):
+        assert parse_llm_answer(None, "MCQ_SINGLE_CORRECT") is None
+
+    def test_empty_string_input(self):
+        assert parse_llm_answer("", "MCQ_SINGLE_CORRECT") is None
+
+    def test_unknown_type(self):
+        assert parse_llm_answer("<answer>5</answer>", "UNKNOWN_TYPE") is None