Spread

Paper introduces adversarial training for robust RL policies

This paper, titled "Adversarial Latent-State Training for Robust Policies in Partially Observable Domains," introduces a new framework for reinforcement learning in partially observable environments. The authors propose…