FaithMed: Training LLMs For Faithful Evidence-Based Medical Reasoning

2026-07-01

AI Summary

Current large language models in healthcare often lack active evidence acquisition and effective supervision during reasoning, resulting in insufficient explainability and reliability for clinical decision-making. This work proposes a novel step-level process supervision framework that formalizes evidence-based medicine principles into procedural standards. By integrating clinician-designed automated scoring rules, a retrieval-augmented agent architecture, step-level reinforcement learning, and an advantage grouping strategy, the approach jointly optimizes reasoning faithfulness and task performance. Evaluated across seven medical benchmarks, the model outperforms agent-based search baselines by an average of 9%, surpasses outcome-only reinforcement learning by 5.8%, and achieves a 15.5% improvement in evidence-based medicine scoring.

Abstract

Faithful reasoning is essential in medicine, where clinical decisions require transparent justification grounded in reliable evidence. Current medical LLMs either lack active access to evidence or use retrieved evidence without supervising how it should be appraised and applied during reasoning. To address this, we formalize evidence-based medicine principles as process-level criteria and introduce FaithMed, a framework that combines clinician-designed, automatically refined rubrics with reinforcement learning using step-level process reward assignment and advantage grouping. Across seven medical benchmarks, FaithMed improves over agentic-search baselines (+9% on average) and outcome-only RL (+5.8%), while raising average evidence-based medicine rubric scores over agentic-search Qwen3 baselines (+15.5%). This work demonstrates that explicit step-level supervision can improve both task success and the faithfulness of the reasoning process. Code is available at https://github.com/cxcscmu/FaithMed.

Problem

• faithful reasoning
• evidence-based medicine
• medical LLMs
• reasoning process
• clinical decision-making

Innovation

• evidence-based medicine
• faithful reasoning
• process-level supervision
• reinforcement learning
• step-level reward