Datasets:
| { | |
| "meta_info": { | |
| "case_id": "case39_01", | |
| "bench_subset": "NeuroBench-Multi", | |
| "difficult": 1, | |
| "original_dataset": "Multi_QA" | |
| }, | |
| "agent_input": { | |
| "data_path": [], | |
| "instruction": "Answer this standalone multimodal neurophysiology multiple-choice question: Compared with fMRI, which property of EEG directly supports resolving rapid neural events? Options: A: EEG inherently provides millimeter-scale cortical source localization. B: EEG provides millisecond-scale temporal resolution. C: EEG is more sensitive to deep nuclei than to cortical activity. D: EEG is unaffected by head models and volume conduction. Select exactly one option, A, B, C, or D. Return only one JSON object with the single field answer and no explanation or extra fields." | |
| }, | |
| "eval_config": { | |
| "parser_prompt": "### ROLE\nYou are a precise JSON Extraction Engine for neurophysiological data. Your sole task is to convert natural language reports into structured JSON data.\n\n### TASK\nExtract the answer choice from the agent's report.\n- answer: exactly one of A, B, C, or D.\n\n### STRICT CONSTRAINTS (MANDATORY)\n1. Output ONLY a valid JSON object.\n2. DO NOT include Markdown code blocks.\n3. DO NOT include conversational text, explanations, or extra fields.\n4. The JSON key must be EXACTLY \"answer\".\n5. The answer value must be exactly one of \"A\", \"B\", \"C\", or \"D\".\n6. If the answer is missing or unclear, return {\"answer\": null}.\n\n### OUTPUT TEMPLATE\n{\"answer\": \"<A | B | C | D or null>\"}", | |
| "metrics": [ | |
| { | |
| "metric_id": "knowledge_answer_accuracy", | |
| "type": "categorical_check", | |
| "target_key": "answer", | |
| "weight": 100, | |
| "params": { | |
| "gt_value": "B" | |
| } | |
| } | |
| ] | |
| } | |
| } | |