Robust Multi-Agent Reinforcement Learning by Mutual Information Regularization

no code implementations • 15 Oct 2023 • Simin Li, Ruixiao Xu, Jingqiao Xiu, Yuwei Zheng, Pu Feng, Yaodong Yang, Xianglong Liu

Existing robust MARL methods either approximate or enumerate all possible threat scenarios against worst-case adversaries, leading to computational intensity and reduced robustness.

Multi-agent Reinforcement Learning Off-policy evaluation +3