{"success":true,"database":"eegdash","data":{"_id":"6953f4249276ef1ee07a3398","dataset_id":"ds004854","associated_paper_doi":null,"authors":["Tony Johnson","Stephen Gordon","Jon Touryan","Kevin King"],"bids_version":"1.8.0","contact_info":["Kevin King"],"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.18112/openneuro.ds004854.v1.0.0","datatypes":["eeg"],"demographics":{"subjects_count":1,"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/ds004854","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"0682ea8c38c10dce9376665c56f9a6c031cabeb6ebd738e958512d0283cc65e4","license":"CC0","n_contributing_labs":null,"name":"TX18","readme":"TX18 dataset\nThis is a placeholder dataset.","recording_modality":["eeg"],"senior_author":"Kevin King","sessions":[],"size_bytes":83056632,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["nback"],"timestamps":{"digested_at":"2026-04-22T12:27:03.532758+00:00","dataset_created_at":"2023-11-14T15:35:17.746Z","dataset_modified_at":"2023-11-14T15:36:07.000Z"},"total_files":1,"storage":{"backend":"s3","base":"s3://openneuro.org/ds004854","raw_key":"dataset_description.json","dep_keys":["CHANGES","README","participants.tsv","task-nback_events.json"]},"tagger_meta":{"config_hash":"3557b68bca409f28","metadata_hash":"e34c034ce7723d87","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":0.6},"reasoning":{"few_shot_analysis":"Closest few-shot by cognitive paradigm is the digit span dataset (Healthy / Auditory / Memory), which labels a classic working-memory span task as Type=Memory. An n-back task is also a canonical working-memory paradigm, so by convention it maps to Type=Memory. None of the few-shot examples provide a convention for inferring stimulus Modality from only a task name with no description, so Modality cannot be confidently inferred.","metadata_analysis":"Very limited metadata; dataset appears non-descriptive/placeholder. Key snippets: (1) readme: \"This is a placeholder dataset.\" (2) tasks: [\"nback\"]. (3) participants_overview: \"Subjects: 1\". There is no explicit statement of participant diagnosis/clinical recruitment, and no description of stimulus type (visual/auditory/etc.) or trial structure.","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology: Metadata says nothing beyond \"Subjects: 1\" and does not mention any diagnosis; few-shot conventions require explicit clinical recruitment to assign a disorder label, otherwise use Healthy only when clearly stated as healthy controls—here it is not stated. => Aligns with choosing Unknown.\nModality: Metadata provides only \"nback\" with no stimulus description; few-shot examples infer Modality from explicit stimulus descriptions (e.g., \"presented auditorily\", \"visual discrimination task\"). Here there is no such fact. => Aligns with choosing Unknown.\nType: Metadata indicates task \"nback\"; few-shot convention (digit span example) maps working-memory span paradigms to Type=Memory. N-back is a working-memory task, so the pattern suggests Memory and does not conflict with any metadata facts.","decision_summary":"Top-2 candidates and decision:\n- Pathology: (1) Unknown—supported by lack of any diagnosis/health statement (\"This is a placeholder dataset.\", \"Subjects: 1\"); (2) Healthy—possible but unsupported because no explicit 'healthy' recruitment is stated. Winner: Unknown. Confidence based on absence of evidence.\n- Modality: (1) Unknown—no stimulus info provided (\"This is a placeholder dataset.\"); (2) Visual—common for n-back but would be an unjustified assumption from the task name alone. Winner: Unknown. Confidence low due to no direct modality facts.\n- Type: (1) Memory—supported by task label \"nback\" (canonical working-memory paradigm) and few-shot convention that working-memory tasks (digit span) are labeled Memory; (2) Attention—also plausible because n-back involves sustained attention, but primary construct is typically working memory. Winner: Memory. Confidence moderate because inference is based mainly on task name without detailed task description."}},"nemar_citation_count":0,"computed_title":"TX18","nchans_counts":[{"val":64,"count":1}],"sfreq_counts":[{"val":128.0,"count":1}],"stats_computed_at":"2026-04-22T23:16:00.308593+00:00","total_duration_s":1926.1796875,"canonical_name":null,"name_confidence":0.95,"name_meta":{"suggested_at":"2026-04-14T10:18:35.343Z","model":"openai/gpt-5.2 + openai/gpt-5.4-mini + deterministic_fallback"},"name_source":"canonical","author_year":"Johnson2023_TX18"}}