Maximum Entropy Heterogeneous-Agent Reinforcement Learning

arXiv:2306.10715v4 Announce Type: replace-cross
Abstract: Multi-agent reinforcement learning (MARL) has been shown effective for cooperative games in recent years. However, existing state-of-the-art methods face challenges related to sample complexity, training instability, and the risk of converging to a suboptimal Nash Equilibrium. In this paper, we propose a unified framework for learning \emph{stochastic} policies to resolve these issues. We embed cooperative MARL problems into probabilistic graphical models, from which we derive the maximum entropy (MaxEnt) objective for MARL. Based on the …

abstract agent art arxiv challenges complexity cs.lg cs.ma entropy equilibrium face framework games however multi-agent nash equilibrium paper reinforcement reinforcement learning risk sample state stochastic training type