Revisiting Regularized Policy Optimization for Stable and Efficient Reinforcement Learning in Two-Player Games

ArXi:2602.10894v2 Announce Type: replace-cross Two-player games such as board games have long been used as traditional benchmarks for reinforcement learning. This work revisits a policy optimization method with reverse Kullback-Leibler regularization and entropy regularization and analyzes this combination in two-player zero-sum settings from theoretical and empirical perspectives. From a theoretical perspective, we investigate the stability of the policy update rule in two theoretical settings: game-theoretic normal-form games and finite-length games.