Datasets:
| { | |
| "meta_info": { | |
| "case_id": "MentalArithmetic_05.edf", | |
| "bench_subset": "NeuroBench-Core", | |
| "difficult": 1.5, | |
| "original_dataset": "MentalArithmetic" | |
| }, | |
| "agent_input": { | |
| "data_path": "data/core/MentalArithmetic_05.edf", | |
| "instruction": "Please select only the parietal (P*) and occipital (O*) EEG channels from the raw signal. Apply a 0.5-40 Hz FIR bandpass filter, then a 60 Hz FIR notch filter, and resample to 200 Hz. Compute the SEF95 value for each selected channel over the full available signal, then report the region-mean values as sef95_p_mean and sef95_o_mean." | |
| }, | |
| "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 two final values from the agent report:\n1) Mean SEF95 of parietal channels\n2) Mean SEF95 of occipital channels\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 units.\n4. The keys in JSON must be EXACTLY \"sef95_p_mean\" and \"sef95_o_mean\".\n5. Both values must be FLOAT.\n6. If one value is missing, return it as null.\n\n### OUTPUT TEMPLATE\n{\"sef95_p_mean\": <float|null>, \"sef95_o_mean\": <float|null>}", | |
| "metrics": [ | |
| { | |
| "metric_id": "sef95_p_accuracy", | |
| "type": "numeric_check", | |
| "target_key": "sef95_p_mean", | |
| "weight": 50, | |
| "params": { | |
| "gt_value": 18.25, | |
| "tolerance": 0.9125 | |
| } | |
| }, | |
| { | |
| "metric_id": "sef95_o_accuracy", | |
| "type": "numeric_check", | |
| "target_key": "sef95_o_mean", | |
| "weight": 50, | |
| "params": { | |
| "gt_value": 17.125, | |
| "tolerance": 0.85625 | |
| } | |
| } | |
| ] | |
| } | |
| } | |