作者位次 #8/10ACL 2026Findings· Findings of ACL
Graph- GRPO : Stabilizing Multi-Agent Topology Learning via Group Relative Policy Optimization
Yueyang Cang · Xiaoteng Zhang · Erlu Zhao · Zehua Ji · Yuhang Liu · Yuchen He · Zhiyuan Ning · Chen Yijun · Wenge Que · Li Shi
摘要
Optimizing communication topology is fundamental to the efficiency and effectiveness of Large Language Model (LLM)-based Multi-Agent Systems (MAS). While recent approaches utilize reinforcement learning to dynamically construct task-specific graphs, they typically rely on single-sample policy gra...