Aravind Venugopal, Stephanie Milani, Fei Fang and Balaraman Ravindran

MABL: Bi-Level Latent-Variable World Model for Sample-Efficient Multi-Agent Reinforcement Learning