{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.05405", "kind": "paper", "url": "https://arxiv.org/abs/2609.05405"}], "authors": ["Ji Soo Lee", "Xilun Chen", "Pierce Chuang", "Ashish Shenoy", "Jason Wei", "Dohwan Ko", "Hyunwoo J. Kim", "Benoit Corda"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.05405", "source_url": "https://arxiv.org/abs/2609.05405"}, "categories": ["General AI", "Language & Knowledge", "Reasoning"], "description": {"en": "WearableQA evaluates AI reasoning over longitudinal wearable sensor data. It comprises 4,084 multiple-choice questions derived from 200 real users' wearable time series, blood biomarkers, and demographics. Questions are organized into 16 types along data versus health reasoning and single- versus cross-signal axes. Scoring uses multiple-choice accuracy with a known chance baseline."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.05405", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.05405", "title": "WearableQA"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Korea Advanced Institute of Science and Technology", "Korea University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.05405", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.05405", "source_url": "https://openalex.org/W7211918531"}, "key": "claire-radar:2609.05405", "modality": null, "name": "WearableQA", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "false", "origin_record_id": "bm_wearableqa_b2b26908", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "8a0dc28830ae14713070843f45fd69d1836bb8797c60ba6a1e2f583a36fa3dd4", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-08T00:56:29.478814Z", "source_url": "https://arxiv.org/abs/2609.05405"}, "publisher": null, "released": "2026-09-04", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.05405", "source_url": "https://arxiv.org/abs/2609.05405"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-05405", "source": "claire_radar", "source_benchmark_id": "2609.05405", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/facebookresearch/WearableQA", "githubScope": "benchmark_repo", "githubStars": 23, "hfDailySubmittedAt": "2026-09-10T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 44, "hfPaperUrl": "https://huggingface.co/papers/2609.05405", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Topical focus on wearable health reasoning is moderately broad, and the benchmark's scale and defined evaluation protocol are clearly detailed in the paper.", "score": 65}, "capabilities": ["Reasoning"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-08T08:59:06.803818Z", "inputHash": "85088f1ba83b9dcef06b7a4e9acea7690da1dc4d934b759f287eaa98da1fa64b", "model": "deepseek-v4-flash"}, "curation": {"decisionReason": "No artifact evidence or public release path is supplied; without a repository, dataset, or code, the benchmark cannot be verified as reusable.", "model": "deepseek-v4-pro", "policy": "source-evidence-v2", "reviewedAt": "2026-09-08T00:56:29.478814Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "WearableQA evaluates AI reasoning over longitudinal wearable sensor data. It comprises 4,084 multiple-choice questions derived from 200 real users' wearable time series, blood biomarkers, and demographics. Questions are organized into 16 types along data versus health reasoning and single- versus cross-signal axes. Scoring uses multiple-choice accuracy with a known chance baseline.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce WearableQA, a benchmark comprising 4,084 10-option multiple-choice questions constructed from the wearable time series, blood biomarkers, and demographics of 200 real users, each with up to 500 days of daily measurements."}, "familyId": "bmf_b6193a39882b", "firstSeenAt": "2026-09-07", "heat": null, "id": "bm_wearableqa_b2b26908", "indexedAt": "2026-09-07T06:22:02.605340Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.05405", "pdf": "https://arxiv.org/pdf/2609.05405", "project": null, "report": "https://arxiv.org/abs/2609.05405"}, "metrics": [], "motivation": "Recent advances in wearable sensing enable continuous monitoring of physiological and behavioral signals, yet existing benchmarks rarely evaluate whether AI systems can reason over a real user's longitudinal wearable record.", "name": "WearableQA", "oneLine": "WearableQA evaluates AI reasoning over longitudinal wearable sensor data. It comprises 4,084 multiple-choice questions derived from 200 real users' wearable time series, blood biomarkers, and demographics. Questions are organized into 16 types along data versus health reasoning and single- versus cross-signal axes. Scoring uses multiple-choice accuracy with a known chance baseline.", "paperTitle": "WearableQA: A Benchmark for Health Reasoning over Real-World Wearable Data", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.05405", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-07T06:22:02.605340Z"}, "publications": [], "ranking": {"90d": {"confidence": "Medium", "coverage": 0.7, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.4242130572709284, "value": 23}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.6726167420705262, "value": 44}}, "confidence": "Medium", "coverage": 0.7, "rank": 76, "score": 48}, "method": "attention-ranking-v13", "rank": 76, "score": 48}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-04", "source": {"authors": ["Ji Soo Lee", "Xilun Chen", "Pierce Chuang", "Ashish Shenoy", "Jason Wei", "Dohwan Ko", "Hyunwoo J. Kim", "Benoit Corda"], "categories": ["cs.CL"], "id": "2609.05405", "title": "WearableQA: A Benchmark for Health Reasoning over Real-World Wearable Data", "type": "arxiv", "url": "https://arxiv.org/abs/2609.05405"}, "sourceUpdatedAt": "2026-09-04T17:52:37Z", "topics": ["Reasoning"], "venueAttempts": [], "whyItMatters": "Most existing benchmarks do not test reasoning over authentic continuous wearable data, limiting insight into model performance on real-world physiological monitoring. WearableQA provides a diagnostic tool to differentiate model capabilities across distinct reasoning types and signals a largely unsolved challenge, as most models score below 60 percent."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
