Accelerating Safe Reinforcement Learning with Constraint-mismatched Baseline Policies.

Tsung-Yen Yang, Justinian Rosca, Karthik Narasimhan, Peter J. Ramadge

Browse the full ICML paper archive.