Datasets:
File size: 2,959 Bytes
d846fa6 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 | {
"meta_info": {
"case_id": "BCIC2020-3_01.edf",
"bench_subset": "NeuroBench-Core",
"difficult": 1.5,
"original_dataset": "BCIC2020-3"
},
"agent_input": {
"data_path": "data/core/BCIC2020-3_01.edf",
"instruction": "Please use the first 30 seconds of EEG channels, apply a 0.5-4Hz FIR bandpass filter, resample to 200Hz, compute channel-wise variance, and list the top five channel names ranked from highest to lowest variance."
},
"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 final ranked top-5 EEG channel names by variance.\n\n### STRICT CONSTRAINTS (MANDATORY)\n1. Output ONLY a valid JSON object.\n2. DO NOT include Markdown code blocks.\n3. DO NOT include any conversational text or explanations.\n4. The key in JSON must be EXACTLY \"top5_channels\".\n5. Extract the final ranked list, not a channel inventory or an intermediate list. The value must be an array of exactly 5 strings.\n6. Normalize EACH reported channel independently: trim whitespace, remove one leading acquisition prefix EEG followed by a space, hyphen, or underscore, then uppercase the label.\n7. Remove one terminal reference suffix if present: -A1, -A2, -LE, -REF, -M1, -M2, or -AVG.\n8. Convert legacy aliases after removing the prefix and suffix: T3->T7, T4->T8, T5->P7, T6->P8.\n9. Each output label must contain only uppercase A-Z letters and digits. Examples: EEG Fp1-LE -> FP1; C3-A2 -> C3; EEG T4 -> T8; Fz -> FZ.\n10. Keep the five channels in the descending variance order reported. Do not sort the array alphabetically and do not substitute different channels.\n11. If the final list does not contain exactly 5 identifiable channel labels, return {\"top5_channels\": null}; do not guess.\n\n### OUTPUT TEMPLATE\n{\"top5_channels\": [\"CH1\", \"CH2\", \"CH3\", \"CH4\", \"CH5\"]}",
"metrics": [
{
"metric_id": "variance_top5_channel_sequence",
"type": "sequence_match_check",
"target_key": "top5_channels",
"weight": 100,
"params": {
"gt_value": [
"AF7",
"FT9",
"AF8",
"FP2",
"FP1"
],
"match_mode": "weighted_partial_order",
"top_k": 5,
"position_weights": [
1.0,
0.8,
0.6,
0.4,
0.2
],
"min_overlap": 3,
"allow_order_slip": 1
}
}
]
}
}
|