{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.01961", "kind": "paper", "url": "https://arxiv.org/abs/2606.01961"}, {"id": "gh:automedbench/automedbench", "kind": "repo", "url": "https://github.com/AutoMedBench/AutoMedBench"}], "authors": ["Junqi Liu", "Selena Song", "Yuhan Wang", "Jiawei Mao", "Hardy Chen", "Xiaoke Huang", "Tianhao Qi", "Pengfei Guo", "Yucheng Tang", "Yufan He", "Can Zhao", "Andriy Myronenko", "Dong Yang", "Daguang Xu", "Yuyin Zhou"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.01961", "source_url": "https://arxiv.org/abs/2606.01961"}, "categories": ["Agents", "Agents & Tool Use", "Biology & Drug Discovery", "Multimodal"], "description": {"en": "AutoMedBench evaluates autonomous AI agents on end-to-end medical-AI research tasks spanning segmentation, image enhancement, VQA, report generation, and lesion detection. Tasks follow a five-stage workflow (Plan, Setup, Validate, Inference, Submit), with scoring based on both final task performance and stage-level rubric scores."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.01961", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.01961", "title": "AutoMedBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["University of California, Santa Cruz"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.01961", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.01961", "source_url": "https://openalex.org/W7163140972"}, "key": "claire-radar:2606.01961", "modality": null, "name": "AutoMedBench", "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_automedbench_61e5be52", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "59e723feb8e3f7e9ce9c8b48c6582f837aea676d22938f126dddd98a1adbc0f1", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.01961"}, "publisher": null, "released": "2026-06-01", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.01961", "source_url": "https://arxiv.org/abs/2606.01961"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-01961", "source": "claire_radar", "source_benchmark_id": "2606.01961", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Biology & Drug Discovery"], "area": "Agents & Tool Use", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/AutoMedBench/AutoMedBench", "githubScope": "benchmark_repo", "githubStars": 62, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 26, "hfPaperUrl": "https://huggingface.co/papers/2606.01961", "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": "2026-09-30", "state": "stale"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "8b76663e82520ad458808d2fdc47e2e21c699407edea64cc4bc7e6c2b7db4452", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "AutoMedBench evaluates autonomous AI agents on end-to-end medical-AI research tasks spanning segmentation, image enhancement, VQA, report generation, and lesion detection. Tasks follow a five-stage workflow (Plan, Setup, Validate, Inference, Submit), with scoring based on both final task performance and stage-level rubric scores.", "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"], "snippet": "To address this gap, we present AutoMedBench, a workflow-aware benchmark for autonomous medical-AI research across diverse medical imaging and multimodal inference tasks, organizing agent execution into a unified five-stage workflow (S1-S5): Plan, Setup, Validate, Inference, and Submit."}, "familyId": "bmf_bbd322eeddb0", "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/AutoMedBench/AutoMedBench", "url": "https://github.com/AutoMedBench/AutoMedBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/AutoMedBench/AutoMedBench", "url": "https://huggingface.co/papers/2606.01961"}, {"basis": "huggingface-paper-association", "linkedUrl": "https://github.com/AutoMedBench/AutoMedBench", "url": "https://huggingface.co/api/papers/2606.01961"}], "scope": "benchmark_repo", "url": "https://github.com/AutoMedBench/AutoMedBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_automedbench_61e5be52", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": "https://github.com/AutoMedBench/AutoMedBench", "data": null, "hfPaper": "https://huggingface.co/papers/2606.01961", "paper": "https://arxiv.org/abs/2606.01961", "pdf": "https://arxiv.org/pdf/2606.01961", "project": null, "report": "https://arxiv.org/abs/2606.01961"}, "metrics": [], "motivation": "Autonomous agents are increasingly expected to support end-to-end medical-AI research workflows, moving beyond isolated prediction tasks or short-form clinical question answering.", "name": "AutoMedBench", "oneLine": "AutoMedBench evaluates autonomous AI agents on end-to-end medical-AI research tasks spanning segmentation, image enhancement, VQA, report generation, and lesion detection. Tasks follow a five-stage workflow (Plan, Setup, Validate, Inference, Submit), with scoring based on both final task performance and stage-level rubric scores.", "paperTitle": "AutoMedBench: Towards Medical AutoResearch with Agentic AI Models", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.01961", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-01", "source": {"authors": ["Junqi Liu", "Selena Song", "Yuhan Wang", "Jiawei Mao", "Hardy Chen", "Xiaoke Huang", "Tianhao Qi", "Pengfei Guo", "Yucheng Tang", "Yufan He", "Can Zhao", "Andriy Myronenko", "Dong Yang", "Daguang Xu", "Yuyin Zhou"], "categories": ["cs.AI"], "id": "2606.01961", "title": "AutoMedBench: Towards Medical AutoResearch with Agentic AI Models", "type": "arxiv", "url": "https://arxiv.org/abs/2606.01961"}, "sourceUpdatedAt": "2026-06-03T05:43:27Z", "topics": ["Agents", "Multimodal"], "venueAttempts": [], "whyItMatters": "Existing medical agent benchmarks focus on final outputs, obscuring failure points. AutoMedBench provides granular stage-level scoring and error analysis, enabling targeted assessment of agent capabilities and highlighting bottlenecks like verification and submission, which can guide improvement priorities in automated medical research systems."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
