Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment

• URL: http://arxiv.org/abs/2609.11185v1
• Date: Thu, 10 Sep 2026 07:46:41 GMT
• Title: Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment
• Abstract summary: We introduce LogiMed-RoB, a benchmark grounded in Cochrane Risk of Bias (RoB) 2.0 expert logic.<n>It evaluates models under the Hierarchical Logical Consistency framework across four dimensions.<n> Experiments on 10 state-of-the-art LLMs reveal a catastrophic Error Compounding Effect.
• Score: 6.781586922950944
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Evidence-based medicine demands strict logical consistency, yet current evaluations of large language models (LLMs) prioritize superficial label matching over genuine reasoning. 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. It evaluates models under the Hierarchical Logical Consistency (HLC) framework across four dimensions: Atomic Consistency, Domain Consistency, Aggregation Consistency, and Evidential Faithfulness. 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%, with several open-weight architectures plummeting to nearly 0%. We further uncover a systematic evidence-reasoning gap: even when models retrieve high-quality evidence, they fail to deduce correct outcomes in 18.63-40.05% of cases, while Blind Guess Rates reach 48.28%. LogiMed-RoB demonstrates that high outcome accuracy can conceal critical reasoning flaws, underscoring the necessity of white-box logical verification for clinical deployment.

Related papers

• NovGauge: A Fine-Grained Benchmark for Diagnosing LLMs' Capability in Paper Novelty Assessment [54.4265627247104]
  Large language models (LLMs) are increasingly used in peer review at major AI conferences.<n>Existing benchmarks assess novelty as a single holistic score.<n>We present NovGauge, a human-anchored benchmark for fine-grained novelty assessment diagnosis.
  arXiv  Detail & Related papers  (2026-09-10T08:34:56Z)
• Enhancing Brain MRI Anomaly Detection and Reasoning with ROI Rethink and Synthetic Data [9.98520042320808]
  We present BrReMark, a framework that introduces explicit region marking into brain MRI diagnosis.<n>The model first generates hypotheses about potential abnormalities and grounds them through explicit bounding box marking.<n>We integrate a domain randomization-based pathology synthesis augmentation strategy to improve the model's generalizability to out-of-distribution (OOD) data.
  arXiv  Detail & Related papers  (2026-06-24T14:41:27Z)
• ChaosBench-Logic v2: Evaluating LLM Logical Reasoning over Dynamical Systems at Scale [0.0]
  We present ChaosBench-Logic v2, a 40,886-question benchmark over 165 dynamical systems.<n>We find that regime-transition reasoning remains near random even for frontier models.<n>Per-family decomposition shows that the proprietary-model advantage concentrates on cross-indicator and consistency tasks.
  arXiv  Detail & Related papers  (2026-05-23T00:34:17Z)
• Eligibility-Aware Evidence Synthesis: An Agentic Framework for Clinical Trial Meta-Analysis [49.83108591873481]
  EligMeta is an agentic framework that integrates automated trial discovery with eligibility-aware meta-analysis.<n>The framework structures eligibility criteria and computes similarity-based study weights reflecting population alignment between target trials.
  arXiv  Detail & Related papers  (2026-04-03T03:18:50Z)
• PanCanBench: A Comprehensive Benchmark for Evaluating Large Language Models in Pancreatic Oncology [48.732366302949515]
  Large language models (LLMs) have achieved expert-level performance on standardized examinations, yet multiple-choice accuracy poorly reflects real-world clinical utility and safety.<n>We developed a human-in-the-loop pipeline to create expert rubrics for de-identified patient questions.<n>We evaluated 22 proprietary and open-source LLMs using an LLM-as-a-judge framework, measuring clinical completeness, factual accuracy, and web-search integration.
  arXiv  Detail & Related papers  (2026-03-02T00:50:39Z)
• CausalT5K: Diagnosing and Informing Refusal for Trustworthy Causal Reasoning of Skepticism, Sycophancy, Detection-Correction, and Rung Collapse [1.4608214000864057]
  CausalT5K is a diagnostic benchmark of over 5,000 cases across 10 domains.<n>Unlike synthetic benchmarks, CausalT5K embeds causal traps in realistic narratives.<n>Preliminary experiments reveal a Four-Quadrant Control Landscape where static audit policies universally fail.
  arXiv  Detail & Related papers  (2026-02-09T17:36:56Z)
• MedOmni-45°: A Safety-Performance Benchmark for Reasoning-Oriented LLMs in Medicine [69.08855631283829]
  We introduce Med Omni-45 Degrees, a benchmark designed to quantify safety-performance trade-offs under manipulative hint conditions.<n>It contains 1,804 reasoning-focused medical questions across six specialties and three task types, including 500 from MedMCQA.<n>Results show a consistent safety-performance trade-off, with no model surpassing the diagonal.
  arXiv  Detail & Related papers  (2025-08-22T08:38:16Z)
• Mitigating Hallucinations in Large Language Models via Causal Reasoning [16.830640147232295]
  Large language models (LLMs) exhibit logically inconsistent hallucinations that appear coherent yet violate reasoning principles.<n>We introduce causal-DAG construction and reasoning (CDCR-SFT)<n>CDCR-SFT trains LLMs to explicitly construct variable-level directed acyclic graph (DAG) and then perform reasoning over it.
  arXiv  Detail & Related papers  (2025-08-17T20:51:06Z)
• ReasonMed: A 370K Multi-Agent Generated Dataset for Advancing Medical Reasoning [54.30630356786752]
  ReasonMed is the largest medical reasoning dataset to date, with 370k high-quality examples.<n>It is built through a multi-agent generation, verification, and refinement process.<n>Using ReasonMed, we find that integrating detailed CoT reasoning with concise answer summaries yields the most robust fine-tuning results.
  arXiv  Detail & Related papers  (2025-06-11T08:36:55Z)
• Benchmarking Chinese Medical LLMs: A Medbench-based Analysis of Performance Gaps and Hierarchical Optimization Strategies [11.0505830548286]
  This study introduces a granular error taxonomy through systematic analysis of top 10 models on MedBench.<n> Evaluation of 10 leading models reveals vulnerabilities, despite achieving 0.86 accuracy in medical knowledge recall.<n>Our analysis uncovers systemic weaknesses in knowledge boundary enforcement and multi-step reasoning.
  arXiv  Detail & Related papers  (2025-03-10T13:28:25Z)
• Quantifying the Reasoning Abilities of LLMs on Real-world Clinical Cases [48.87360916431396]
  We introduce MedR-Bench, a benchmarking dataset of 1,453 structured patient cases, annotated with reasoning references.<n>We propose a framework encompassing three critical examination recommendation, diagnostic decision-making, and treatment planning, simulating the entire patient care journey.<n>Using this benchmark, we evaluate five state-of-the-art reasoning LLMs, including DeepSeek-R1, OpenAI-o3-mini, and Gemini-2.0-Flash Thinking, etc.
  arXiv  Detail & Related papers  (2025-03-06T18:35:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.