# Disentangling Reasoning and Knowledge in Medical Large Language Models

arXiv: https://arxiv.org/abs/2505.11462 · HTML: https://mlsanity.com/paper/2505.11462/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning.
- Builds on 26 of its 50 references that are in the map.
- First posted on arXiv 16 May 2025.

## Placed in

- LLM reasoning / [Clinical and biomedical applications](https://mlsanity.com/lr/clinical-and-biomedical-applications/)

## Builds on (most cited first, 26 of 26)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [PubMedQA: A Dataset for Biomedical Research Question Answering](https://mlsanity.com/paper/1909.06146/) (2019)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [Kimi k1.5: Scaling Reinforcement Learning with LLMs](https://mlsanity.com/paper/2501.12599/) (2025)
- [SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training](https://mlsanity.com/paper/2501.17161/) (2025)
- [Humanity's Last Exam](https://mlsanity.com/paper/2501.14249/) (2025)
- [Training Language Models to Self-Correct via Reinforcement Learning](https://mlsanity.com/paper/2409.12917/) (2024)
- [R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization](https://mlsanity.com/paper/2503.10615/) (2025)
- [Generating Sequences by Learning to Self-Correct](https://mlsanity.com/paper/2211.00053/) (2022)
- [HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs](https://mlsanity.com/paper/2412.18925/) (2024)
- [MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding](https://mlsanity.com/paper/2501.18362/) (2025)
- [SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Models](https://mlsanity.com/paper/2504.11468/) (2025)
- [Evaluation of OpenAI o1: Opportunities and Challenges of AGI](https://mlsanity.com/paper/2409.18486/) (2024)
- [OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles](https://mlsanity.com/paper/2503.17352/) (2025)
- [MedReason: Eliciting Factual Medical Reasoning Steps in LLMs via Knowledge Graphs](https://mlsanity.com/paper/2504.00993/) (2025)
- [Benchmarking Large Language Models on Answering and Explaining Challenging Medical Questions](https://mlsanity.com/paper/2402.18060/) (2024)
- [HEAD-QA: A Healthcare Dataset for Complex Reasoning](https://mlsanity.com/paper/1906.04701/) (2019)
- [Navigating the Grey Area: How Expressions of Uncertainty and Overconfidence Affect Language Models](https://mlsanity.com/paper/2302.13439/) (2023)
- [m1: Unleash the Potential of Test-Time Scaling for Medical Reasoning with Large Language Models](https://mlsanity.com/paper/2504.00869/) (2025)
- [Disentangling Memory and Reasoning Ability in Large Language Models](https://mlsanity.com/paper/2411.13504/) (2024)
- [Backtracking Improves Generation Safety](https://mlsanity.com/paper/2409.14586/) (2024)
- [Lessons From Red Teaming 100 Generative AI Products](https://mlsanity.com/paper/2501.07238/) (2025)

## Built on by (most cited first, 18 of 18)

- [MedAgentGym: A Scalable Agentic Training Environment for Code-Centric Reasoning in Biomedical Data Science](https://mlsanity.com/paper/2506.04405/) (2025)
- [Medical Reasoning in the Era of LLMs: A Systematic Review of Enhancement Techniques and Applications](https://mlsanity.com/paper/2508.00669/) (2025)
- [Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning](https://mlsanity.com/paper/2508.16129/) (2025)
- [Evo-MedAgent: Beyond One-Shot Diagnosis with Agents That Remember, Reflect, and Improve](https://mlsanity.com/paper/2604.14475/) (2026)
- [RCP-Merging: Merging Long Chain-of-Thought Models with Domain-Specific Models by Considering Reasoning Capability as Prior](https://mlsanity.com/paper/2508.03140/) (2025)
- [Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoning](https://mlsanity.com/paper/2508.19202/) (2025)
- [Training and Evaluation of Guideline-Based Medical Reasoning in LLMs](https://mlsanity.com/paper/2512.03838/) (2025)
- [Medical Reasoning with Large Language Models: A Survey and MR-Bench](https://mlsanity.com/paper/2604.08559/) (2026)
- [Rethinking Test-Time Scaling for Medical AI: Model and Task-Aware Strategies for LLMs and VLMs](https://mlsanity.com/paper/2506.13102/) (2025)
- [How Far Have Medical Vision-Language Models Come? A Comprehensive Benchmarking Study](https://mlsanity.com/paper/2507.11200/) (2025)
- [Towards Efficient Medical Reasoning with Minimal Fine-Tuning Data](https://mlsanity.com/paper/2508.01450/) (2025)
- [Multilingual Medical Reasoning for Question Answering with Large Language Models](https://mlsanity.com/paper/2512.05658/) (2025)
- [Quark Medical Alignment: A Holistic Multi-Dimensional Alignment and Collaborative Optimization Paradigm](https://mlsanity.com/paper/2602.11661/) (2026)
- [ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging](https://mlsanity.com/paper/2601.05560/) (2026)
- [Process Reward Agents for Steering Knowledge-Intensive Reasoning](https://mlsanity.com/paper/2604.09482/) (2026)
- [Eliciting Medical Reasoning with Knowledge-enhanced Data Synthesis: A Semi-Supervised Reinforcement Learning Approach](https://mlsanity.com/paper/2604.11547/) (2026)
- [ActiveMedAgent: Cost-Aware Trajectory Learning for Multimodal Medical Diagnosis](https://mlsanity.com/paper/2610.11140/) (2026)
- [MedEvo: Medical Reasoning Evolution through Knowledge Navigation](https://mlsanity.com/paper/2610.17690/) (2026)

## Abstract (arXiv)

Medical reasoning in large language models (LLMs) aims to emulate clinicians' diagnostic thinking, but current benchmarks such as MedQA-USMLE, MedMCQA, and PubMedQA often mix reasoning with factual recall. We address this by separating 11 biomedical QA benchmarks into reasoning- and knowledge-focused subsets using a PubMedBERT classifier that reaches 81 percent accuracy, comparable to human performance. Our analysis shows that only 32.8 percent of questions require complex reasoning. We evaluate biomedical models (HuatuoGPT-o1, MedReason, m1) and general-domain models (DeepSeek-R1, o4-mini, Qwen3), finding consistent gaps between knowledge and reasoning performance. For example, HuatuoGPT-o1 scores 56.9 on knowledge but only 44.8 on reasoning. In adversarial tests where models are misled with incorrect initial reasoning, biomedical models degrade sharply, while larger or RL-trained general models show more robustness. To address this, we train BioMed-R1 using fine-tuning and reinforcement learning on reasoning-heavy examples. It achieves the strongest performance among similarly sized models. Further gains may come from incorporating clinical case reports and training with adversarial and backtracking scenarios.