#learning rate selection

Six Papers Advance RL Optimization Theory With New Learning-Rate Bounds

Recent arXiv papers establish finite-iteration convergence guarantees and learning-rate constraints for reinforcement learning algorithms.