{"success":true,"database":"eegdash","data":{"_id":"6953f4239276ef1ee07a32e6","dataset_id":"ds003694","associated_paper_doi":null,"authors":["Benjamin J. Griffiths","María Carmen Martín-Buro","Bernhard Staresina","Simon Hanslmayr"],"bids_version":"1.0.2","contact_info":["Benjamin James Griffiths"],"contributing_labs":null,"data_processed":false,"dataset_doi":"10.18112/openneuro.ds003694.v1.0.0","datatypes":["meg"],"demographics":{"subjects_count":28,"ages":[],"age_min":null,"age_max":null,"age_mean":null,"species":null,"sex_distribution":null,"handedness_distribution":null},"experimental_modalities":null,"external_links":{"source_url":"https://openneuro.org/datasets/ds003694","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"98372048a7cd782018b4bb6e219f1f5f3dce7c00c9ae8f87660acf1d6a33ff27","license":"CC0","n_contributing_labs":null,"name":"MEGMEM","readme":null,"recording_modality":["meg"],"senior_author":"Simon Hanslmayr","sessions":[],"size_bytes":234581975924,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["MEM"],"timestamps":{"digested_at":"2026-04-22T12:25:39.124485+00:00","dataset_created_at":"2021-06-12T09:48:35.040Z","dataset_modified_at":"2021-06-14T06:56:33.000Z"},"total_files":132,"storage":{"backend":"s3","base":"s3://openneuro.org/ds003694","raw_key":"dataset_description.json","dep_keys":["CHANGES","participants.tsv"]},"tagger_meta":{"config_hash":"3557b68bca409f28","metadata_hash":"eb3d4813c5789a74","model":"openai/gpt-5.2","tagged_at":"2026-04-07T09:32:40.872789+00:00"},"tags":{"pathology":["Unknown"],"modality":["Unknown"],"type":["Memory"],"confidence":{"pathology":0.4,"modality":0.4,"type":0.7},"reasoning":{"few_shot_analysis":"Closest few-shot by research construct is the digit-span dataset (Healthy / Auditory / Memory), where an explicitly memory-focused task maps to Type=Memory. This guides labeling convention that when the paradigm is clearly a memory task (even if exact stimuli are not described), Type should be Memory rather than Attention/Perception. However, unlike that few-shot, this dataset’s metadata does not specify stimulus modality (auditory vs visual) or participant clinical status, so those labels cannot be copied and must remain unknown unless explicitly stated.","metadata_analysis":"Available metadata is very sparse. Key lines: (1) Title indicates memory theme: \"title\": \"MEGMEM\". (2) Task name suggests memory: \"tasks\": [\"MEM\"]. (3) Participants only gives count with no diagnosis/health screening: \"participants_overview\": \"Subjects: 28\". No stimulus description (e.g., pictures/words/sounds), no resting/sleep indicators, and no clinical recruitment terms are present.","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology: Metadata says only \"Subjects: 28\" (no patient/control/diagnosis stated). Few-shot patterns: many non-clinical cognitive task datasets are labeled Healthy when explicitly stated as healthy/controls, but here it is not stated. ALIGNMENT: insufficient evidence; cannot assert Healthy.\nModality: Metadata provides no stimulus channel description; only task label \"MEM\". Few-shot patterns: memory tasks can be auditory (digit span) or visual (word/picture encoding), so pattern does not disambiguate. ALIGNMENT: insufficient evidence; must be Unknown.\nType: Metadata says \"MEGMEM\" and task \"MEM\", both consistent with a memory experiment. Few-shot convention: explicit memory task (digit span) maps to Type=Memory. ALIGNMENT: aligns at the construct level (memory-focused task) even though modality/population details are missing.","decision_summary":"Pathology top-2: (a) Unknown—supported by lack of any diagnosis/health quote beyond \"Subjects: 28\"; (b) Healthy—plausible given typical cognitive MEG studies, but not stated. Winner: Unknown (metadata does not specify recruitment/diagnosis). Confidence justified by absence of explicit evidence.\nModality top-2: (a) Unknown—no stimulus info provided; (b) Visual—common for MEG memory paradigms, but not stated. Winner: Unknown. Confidence low due to missing stimulus description.\nType top-2: (a) Memory—supported by \"title\": \"MEGMEM\" and task \"MEM\"; (b) Learning—sometimes memory experiments involve learning/encoding, but no indication of reinforcement/training. Winner: Memory. Confidence moderate because evidence is suggestive but lacks task description details."}},"nemar_citation_count":1,"computed_title":"MEGMEM","nchans_counts":[{"val":327,"count":109},{"val":319,"count":19},{"val":336,"count":4}],"sfreq_counts":[{"val":1000.0,"count":132}],"stats_computed_at":"2026-04-22T23:16:00.222399+00:00","total_duration_s":null,"canonical_name":null,"name_confidence":0.72,"name_meta":{"suggested_at":"2026-04-14T10:18:35.342Z","model":"openai/gpt-5.2 + openai/gpt-5.4-mini + deterministic_fallback"},"name_source":"canonical","author_year":"Griffiths2021"}}