MedOmni-45°: A Safety-Performance Benchmark for Reasoning-Oriented LLMs in Medicine

• URL: http://arxiv.org/abs/2508.16213v1
• Date: Fri, 22 Aug 2025 08:38:16 GMT
• Title: MedOmni-45°: A Safety-Performance Benchmark for Reasoning-Oriented LLMs in Medicine
• Authors: Kaiyuan Ji, Yijin Guo, Zicheng Zhang, Xiangyang Zhu, Yuan Tian, Ning Liu, Guangtao Zhai,
• Abstract summary: 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.
• Score: 69.08855631283829
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: With the increasing use of large language models (LLMs) in medical decision-support, it is essential to evaluate not only their final answers but also the reliability of their reasoning. Two key risks are Chain-of-Thought (CoT) faithfulness -- whether reasoning aligns with responses and medical facts -- and sycophancy, where models follow misleading cues over correctness. Existing benchmarks often collapse such vulnerabilities into single accuracy scores. To address this, we introduce MedOmni-45 Degrees, a benchmark and workflow designed to quantify safety-performance trade-offs under manipulative hint conditions. It contains 1,804 reasoning-focused medical questions across six specialties and three task types, including 500 from MedMCQA. Each question is paired with seven manipulative hint types and a no-hint baseline, producing about 27K inputs. We evaluate seven LLMs spanning open- vs. closed-source, general-purpose vs. medical, and base vs. reasoning-enhanced models, totaling over 189K inferences. Three metrics -- Accuracy, CoT-Faithfulness, and Anti-Sycophancy -- are combined into a composite score visualized with a 45 Degrees plot. Results show a consistent safety-performance trade-off, with no model surpassing the diagonal. The open-source QwQ-32B performs closest (43.81 Degrees), balancing safety and accuracy but not leading in both. MedOmni-45 Degrees thus provides a focused benchmark for exposing reasoning vulnerabilities in medical LLMs and guiding safer model development.

Related papers

• 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)
• Health-ORSC-Bench: A Benchmark for Measuring Over-Refusal and Safety Completion in Health Context [82.32380418146656]
  Health-ORSC-Bench is the first large-scale benchmark designed to measure textbfOver-Refusal and textbfSafe Completion quality in healthcare.<n>Our framework uses an automated pipeline with human validation to test models at varying levels of intent ambiguity.<n>Health-ORSC-Bench provides a rigorous standard for calibrating the next generation of medical AI assistants.
  arXiv  Detail & Related papers  (2026-01-25T01:28:52Z)
• What Matters For Safety Alignment? [38.86339753409445]
  This paper presents a comprehensive empirical study on the safety alignment capabilities of AI systems.<n>We systematically investigate and compare the influence of six critical intrinsic model characteristics and three external attack techniques.<n>We identify the LRMs GPT-OSS-20B, Qwen3-Next-80B-A3B-Thinking, and GPT-OSS-120B as the top-three safest models.
  arXiv  Detail & Related papers  (2026-01-07T12:31:52Z)
• SafeMed-R1: Adversarial Reinforcement Learning for Generalizable and Robust Medical Reasoning in Vision-Language Models [0.0]
  We introduce SafeMed-R1, a hybrid defense framework that ensures robust performance while preserving high-quality, interpretable medical reasoning.<n>We demonstrate that models trained with explicit chain-of-thought reasoning exhibit superior adversarial robustness compared to instruction-only variants.
  arXiv  Detail & Related papers  (2025-12-22T12:07:33Z)
• MedBench v4: A Robust and Scalable Benchmark for Evaluating Chinese Medical Language Models, Multimodal Models, and Intelligent Agents [10.963960571170643]
  We present MedBench v4, a nationwide, cloud-based benchmarking infrastructure comprising over 700,000 expert-curated tasks.<n>Items undergo multi-stage refinement and multi-round review by clinicians from more than 500 institutions, and open-ended responses are scored by an LLM-as-a-judge to human ratings.<n>LLMs reach a mean overall score of 54.1/100 (best: Claude Sonnet 4.5, 62.5/100), but safety and ethics remain low.<n>Multimodal models perform worse overall (mean 47.5/100; best: GPT-5, 54.9/100), with solid perception yet weaker cross-
  arXiv  Detail & Related papers  (2025-11-18T12:37:32Z)
• EchoBench: Benchmarking Sycophancy in Medical Large Vision-Language Models [82.43729208063468]
  Recent benchmarks for medical Large Vision-Language Models (LVLMs) emphasize leaderboard accuracy, overlooking reliability and safety.<n>We study sycophancy -- models' tendency to uncritically echo user-provided information.<n>We introduce EchoBench, a benchmark to systematically evaluate sycophancy in medical LVLMs.
  arXiv  Detail & Related papers  (2025-09-24T14:09:55Z)
• A Novel Evaluation Benchmark for Medical LLMs: Illuminating Safety and Effectiveness in Clinical Domains [15.73821689524201]
  Large language models (LLMs) hold promise in clinical decision support but face major challenges in safety evaluation and effectiveness validation.<n>We developed the Clinical Safety-Effectiveness Dual-Track Benchmark (CSEDB), a multidimensional framework built on clinical expert consensus.<n>Thirty-two specialist physicians developed and reviewed 2,069 open-ended Q&A items aligned with these criteria, spanning 26 clinical departments to simulate real-world scenarios.
  arXiv  Detail & Related papers  (2025-07-31T12:10:00Z)
• Beyond Benchmarks: Dynamic, Automatic And Systematic Red-Teaming Agents For Trustworthy Medical Language Models [87.66870367661342]
  Large language models (LLMs) are used in AI applications in healthcare.<n>Red-teaming framework that continuously stress-test LLMs can reveal significant weaknesses in four safety-critical domains.<n>A suite of adversarial agents is applied to autonomously mutate test cases, identify/evolve unsafe-triggering strategies, and evaluate responses.<n>Our framework delivers an evolvable, scalable, and reliable safeguard for the next generation of medical AI.
  arXiv  Detail & Related papers  (2025-07-30T08:44:22Z)
• 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)
• VADER: A Human-Evaluated Benchmark for Vulnerability Assessment, Detection, Explanation, and Remediation [0.8087612190556891]
  VADER comprises 174 real-world software vulnerabilities, each carefully curated from GitHub and annotated by security experts.<n>For each vulnerability case, models are tasked with identifying the flaw, classifying it using Common Weaknession (CWE), explaining its underlying cause, proposing a patch, and formulating a test plan.<n>Using a one-shot prompting strategy, we benchmark six state-of-the-art LLMs (Claude 3.7 Sonnet, Gemini 2.5 Pro, GPT-4.1, GPT-4.5, Grok 3 Beta, and o3) on VADER.<n>Our results show that current state-of-the-
  arXiv  Detail & Related papers  (2025-05-26T01:20:44Z)
• Disentangling Reasoning and Knowledge in Medical Large Language Models [23.401484250342158]
  Medical reasoning in large language models aims to emulate clinicians' diagnostic thinking.<n>Current benchmarks such as MedQA-USMLE, MedMCQA, and PubMedQA often mix reasoning with factual recall.<n>We evaluate biomedical models (HuatuoGPT-o1, MedReason, m1) and general-domain models (DeepSeek-R1, o4-mini, Qwen3)<n>We train BioMed-R1 using fine-tuning and reinforcement learning on reasoning-heavy examples.
  arXiv  Detail & Related papers  (2025-05-16T17:16:27Z)
• CARES: A Comprehensive Benchmark of Trustworthiness in Medical Vision Language Models [92.04812189642418]
  We introduce CARES and aim to evaluate the Trustworthiness of Med-LVLMs across the medical domain. We assess the trustworthiness of Med-LVLMs across five dimensions, including trustfulness, fairness, safety, privacy, and robustness.
  arXiv  Detail & Related papers  (2024-06-10T04:07:09Z)
• Flames: Benchmarking Value Alignment of LLMs in Chinese [86.73527292670308]
  This paper proposes a value alignment benchmark named Flames. It encompasses both common harmlessness principles and a unique morality dimension that integrates specific Chinese values. Our findings indicate that all the evaluated LLMs demonstrate relatively poor performance on Flames.
  arXiv  Detail & Related papers  (2023-11-12T17:18:21Z)
• Can large language models reason about medical questions? [7.95779617839642]
  We investigate whether close- and open-source models can be applied to answer and reason about difficult real-world-based questions. We focus on three popular medical benchmarks (MedQA-USMLE, MedMCQA, and PubMedQA) and multiple prompting scenarios. Based on an expert annotation of the generated CoTs, we found that InstructGPT can often read, reason and recall expert knowledge.
  arXiv  Detail & Related papers  (2022-07-17T11:24:44Z)

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.