{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.31179", "kind": "paper", "url": "https://arxiv.org/abs/2606.31179"}, {"id": "gh:microsoft/healthagentbench", "kind": "repo", "url": "https://github.com/microsoft/HealthAgentBench"}], "categories": ["Agents & Tool Use", "Biology & Drug Discovery", "cs.AI"], "description": {"en": "Evaluates AI agents on 54 realistic healthcare tasks across 7 categories, including medical imaging, EHR analysis, and clinical trial matching. Agents operate in terminal environments with task-specific verifiers and a final task success rate."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.31179", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.31179", "title": "HealthAgentBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.31179", "modality": null, "name": "HealthAgentBench", "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": "unknown", "origin_record_id": "bm_healthagentbench_400912dd", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "89beab789b8bac7f818b42074b94efbf2326177fd5de31538307129abe1cd42b", "relation": "aggregates", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.31179"}, "publisher": null, "released": "2026-06-30", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.31179", "source_url": "https://arxiv.org/abs/2606.31179"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-31179", "source": "claire_radar", "source_benchmark_id": "2606.31179", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Mixed", "applicationDomains": ["Biology & Drug Discovery"], "area": "Agents & Tool Use", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/microsoft/HealthAgentBench", "githubScope": "benchmark_repo", "githubStars": 55, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 5, "hfPaperUrl": "https://huggingface.co/papers/2606.31179", "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": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-10-04", "state": "stale"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Aggregate Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "65afc00a2c6c32b161975e9b50e9fdf9dcd8514bb9fac5e2b1ae58b65a1b4129", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates AI agents on 54 realistic healthcare tasks across 7 categories, including medical imaging, EHR analysis, and clinical trial matching. Agents operate in terminal environments with task-specific verifiers and a final task success rate.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce HealthAgentBench, a suite of 54 agentic healthcare tasks across 7 categories each with its unique environment."}, "familyId": "bmf_38ba71a1d0f7", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/microsoft/HealthAgentBench", "url": "https://github.com/microsoft/HealthAgentBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/microsoft/HealthAgentBench", "url": "https://huggingface.co/papers/2606.31179"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/microsoft/HealthAgentBench", "url": "https://arxiv.org/abs/2606.31179"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/microsoft/HealthAgentBench.", "url": "https://arxiv.org/abs/2606.31179"}], "scope": "benchmark_repo", "url": "https://github.com/microsoft/HealthAgentBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_healthagentbench_400912dd", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": "https://github.com/microsoft/HealthAgentBench", "data": null, "hfPaper": "https://huggingface.co/papers/2606.31179", "paper": "https://arxiv.org/abs/2606.31179", "pdf": "https://arxiv.org/pdf/2606.31179", "project": null, "report": "https://arxiv.org/abs/2606.31179"}, "metrics": [], "motivation": "As AI agents become increasingly capable of complex, long-horizon reasoning, rigorous and holistic evaluation is essential for measuring progress toward real-world healthcare applications.", "name": "HealthAgentBench", "oneLine": "Evaluates AI agents on 54 realistic healthcare tasks across 7 categories, including medical imaging, EHR analysis, and clinical trial matching. Agents operate in terminal environments with task-specific verifiers and a final task success rate.", "paperTitle": "HealthAgentBench: A Unified Benchmark Suite of Realistic Agentic Healthcare Environments for Challenging Frontier AI Agents", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.31179", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "Microsoft", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/microsoft/HealthAgentBench"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "aggregates", "releasedAt": "2026-06-30", "source": {"authors": ["Qianchu Liu", "Sheng Zhang", "Guanghui Qin", "Jeya Maria Jose Valanarasu", "Maximilian Rokuss", "Mingyu Lu", "Timothy Ossowski", "Juan Manuel Zambrano Chaves", "Cliff Wong", "Peniel Argaw", "Yashna Hasija", "Mu Wei", "Wen-wai Yim", "Qin Liu", "Zilin Jing", "Jason Entenmann", "Naoto Usuyama", "Tristan Naumann", "Hoifung Poon"], "categories": ["cs.AI", "cs.CL", "cs.CV"], "id": "2606.31179", "title": "HealthAgentBench: A Unified Benchmark Suite of Realistic Agentic Healthcare Environments for Challenging Frontier AI Agents", "type": "arxiv", "url": "https://arxiv.org/abs/2606.31179"}, "sourceUpdatedAt": "2026-06-30T06:09:01Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Provides a standardized, realistic evaluation for agentic healthcare AI, revealing performance gaps across task types and informing model selection for clinical applications."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
