Breaking the Curse of Multiagency: Provably Efficient Decentralized Multi-Agent RL with Function Approximation

2023·arXiv