{"record": {"aliases": ["Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment"], "artifacts": [{"id": "arxiv:2609.11185", "kind": "paper", "url": "https://arxiv.org/abs/2609.11185"}], "authors": ["Jiayu Huang", "Zichen Tang", "Qianhui Ling", "Zemin Kuang", "Haihong E"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.11185", "source_url": "https://arxiv.org/abs/2609.11185"}, "categories": ["Biology & Drug Discovery", "Safety & Trustworthiness", "cs.AI"], "description": {"en": "A benchmark assessing large language models' logical consistency in medical risk-of-bias assessment under a Hierarchical Logical Consistency framework with four dimensions."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.11185", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.11185", "title": "LogiMed-RoB"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.11185", "modality": null, "name": "LogiMed-RoB", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "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_can-llms-follow-medical-expert-logic-a-ben_a2cc2223", "origin_source": "arxiv", "recognition_confidence": "0.45", "record_sha256": "d146360919d0c496ff60e46416ce0eba870750db95d351994c0349134e0b2871", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-11T09:02:33.555618Z", "source_url": "https://arxiv.org/abs/2609.11185"}, "publisher": null, "released": "2026-09-10", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.11185", "source_url": "https://arxiv.org/abs/2609.11185"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-11185", "source": "claire_radar", "source_benchmark_id": "2609.11185", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment"], "annotation": "Unknown", "applicationDomains": ["Biology & Drug Discovery"], "area": "Safety & Trustworthiness", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.11185", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Topical relevance to medical AI safety and reported performance collapse may draw interest from researchers evaluating LLM reasoning robustness.", "score": 45}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-11T09:02:33.555618Z", "inputHash": "d2868198a69130e2bda519abd3dfeed645baa62197e782044c734ae04088071f", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We introduce LogiMed-RoB, a benchmark grounded in Cochrane Risk of Bias (RoB) 2.0 expert logic", "canonicalNameSource": "abstract", "decisionReason": "Public reuse path is missing; no evidence of released task inputs, evaluator code, or scored submissions, making it a paper-only benchmark.", "evaluationEvidence": {"reuse": "Accepted to EMNLP 2026 Main Conference; report, paper, and PDF available via arXiv links.", "scoring": "Experiments on 10 state-of-the-art LLMs reveal a catastrophic Error Compounding Effect: despite the top model reaching 98.88% Atomic Consistency, its end-to-end consistency collapses to 45.13%—provides performance metrics.", "task": "comprising 860 randomized controlled trials (RCTs) and 14,820 queries. It evaluates models under the Hierarchical Logical Consistency (HLC) framework across four dimensions: Atomic Consistency, Domain Consistency, Aggregation Consistency, and Evidential Faithfulness."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-11T09:02:33.555618Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A benchmark assessing large language models' logical consistency in medical risk-of-bias assessment under a Hierarchical Logical Consistency framework with four dimensions.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence"], "snippet": "We introduce LogiMed-RoB, a benchmark grounded in Cochrane Risk of Bias (RoB) 2.0 expert logic, comprising 860 randomized controlled trials (RCTs) and 14,820 queries."}, "familyId": "bmf_1180d4e7652c", "firstSeenAt": "2026-09-11", "heat": null, "id": "bm_can-llms-follow-medical-expert-logic-a-ben_a2cc2223", "indexedAt": "2026-09-11T06:20:12.012226Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.11185", "pdf": "https://arxiv.org/pdf/2609.11185", "project": null, "report": "https://arxiv.org/abs/2609.11185"}, "metrics": [], "motivation": "Evidence-based medicine demands strict logical consistency, yet current evaluations of large language models (LLMs) prioritize superficial label matching over genuine reasoning.", "name": "LogiMed-RoB", "oneLine": "A benchmark assessing large language models' logical consistency in medical risk-of-bias assessment under a Hierarchical Logical Consistency framework with four dimensions.", "paperTitle": "Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": "Accepted to EMNLP 2026 Main Conference", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2609.11185", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026 Main Conference", "verifiedAt": "2026-09-11T06:20:12.012226Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "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": 30, "historyDate": "2026-09-05", "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "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": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 0.45, "relation": "introduces", "releasedAt": "2026-09-10", "source": {"authors": ["Jiayu Huang", "Zichen Tang", "Qianhui Ling", "Zemin Kuang", "Haihong E"], "categories": ["cs.AI"], "id": "2609.11185", "title": "Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment", "type": "arxiv", "url": "https://arxiv.org/abs/2609.11185"}, "sourceUpdatedAt": "2026-09-10T07:46:41Z", "topics": ["cs.AI"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026 Main Conference", "evidence": [{"level": "author-claim", "observedAt": "2026-09-11T06:20:12.012226Z", "rawValue": "Accepted to EMNLP 2026 Main Conference", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2609.11185"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026 Main Conference"}], "whyItMatters": "Addresses the gap between superficial accuracy and genuine reasoning in clinical AI, providing a way to detect reasoning flaws not visible in outcome-only evaluation."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
