Episodic Memory Temporal Consistency for Cooperative Multi-Agent Reinforcement Learning

Zicheng Zhao, Yu Lan, Chengzhengxu Li +2

Cooperative Multi-Agent Reinforcement Learning (MARL) frequently suffers from severe reward sparsity and exploration bottlenecks. While episodic memory mechanisms mitigate these is…