让医疗推理更可靠：用多智能体系统结合过程监督，提升基因病关联判断准确率与临床合规性。
Process-Supervised Multi-Agent Reinforcement Learning for Reliable Clinical Reasoning
- 设计分层多智能体系统，用过程监督确保推理符合临床路径。
- 在ClinGen数据集上，结果准确率从19.5%提升至75.0%，过程对齐度达52.0% F1。
- 适合医学知识图谱构建、临床决策支持系统研发者参考。
临床决策需基于异构证据进行细致推理并提供可追溯的解释。现有大模型多智能体系统虽提升结果准确性，但忽视与临床标准一致的过程推理。以基因-疾病因果关系注释为例，专家需整合多样生物医学证据判断基因是否致病。本文提出一种代理作为工具的强化学习框架，双重目标为：(i) 过程层面监督，确保推理遵循有效临床路径；(ii) 通过分层多智能体系统实现高效协作。在ClinGen数据集上的评估显示，仅使用结果奖励时，采用GRPO训练的Qwen3-4B监督代理使最终结果准确率从基础模型的0.195提升至0.732，但过程对齐度仅为0.392 F1；而同时引入过程与结果奖励后，准确率达0.750，过程一致性显著提升至0.520 F1。代码已开源。
原文摘要 · Abstract (English)
Clinical decision-making requires nuanced reasoning over heterogeneous evidence and traceable justifications. While recent LLM multi-agent systems (MAS) show promise, they largely optimise for outcome accuracy while overlooking process-grounded reasoning aligned with clinical standards. One critical real-world case of this is gene-disease validity curation, where experts must determine whether a gene is causally implicated in a disease by synthesising diverse biomedical evidence. We introduce an agent-as-tool reinforcement learning framework for this task with two objectives: (i) process-level supervision to ensure reasoning follows valid clinical pathways, and (ii) efficient coordination via a hierarchical multi-agent system. Our evaluation on the ClinGen dataset shows that with outcome-only rewards, MAS with a GRPO-trained Qwen3-4B supervisor agent substantially improves final outcome accuracy from 0.195 with a base model supervisor to 0.732, but results in poor process alignment (0.392 F1). Conversely, with process + outcome rewards, MAS with GRPO-trained supervisor achieves higher outcome accuracy (0.750) while significantly improving process fidelity to 0.520 F1. Our code is available at https://github.com/chaeeunlee-io/GeneDiseaseCurationAgents.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。