Geometric Value Iteration: Dynamic Error-Aware KL Regularization for Reinforcement Learning

Publication
Asian Conference on Machine Learning (ACML)