首页 > AI前沿 > PAIR-MED: Policy-reward Alternation for Iterative Refinement in Multilingual Medical Reasoning

PAIR-MED: Policy-reward Alternation for Iterative Refinement in Multilingual Medical Reasoning

arXiv自然语言 2026-11-20 13:00 28 阅读 查看原文

研究背景与问题

Large language models demonstrate strong medical question-answering capabilities, yet reasoning performance remains uneven across languages. Process reward models (PRMs) provide step-level feedback for improving reasoning, but a fixed PRM may become increasingly misaligned with the trajectory distribution of an evolving policy during reinforcement learning, thereby reducing the reliability of its supervision.

策略-奖励交替训练框架

We propose a policy--reward alternating training framework that jointly adapts a medical PRM and optimizes a multilingual reasoning policy. Following supervised initialization, the current policy generates multilingual trajectories to update the PRM, which then provides process-level rewards for the next policy update. Medical step supervision maintains the PRM's medical foundation, while process, outcome, and language-related rewards guide policy learning.

实验与结果

We develop