MAPoRL: Post-Training of Collaborative Large Language Models Based on Reinforcement Learning for Multi-Agent Cooperation

MAPoRL models 合作性能 多智能体 大语言模型 领域泛化