#offline rl

Seven Papers Advance RL Policy Optimization and Inverse Learning

New methods address bootstrapping error, inverse reward inference, and offline learning challenges with distributional and theoretical approaches.