@article{reliablecriticsmonotonicimprovementand, title = {Reliable Critics: Monotonic Improvement and Convergence Guarantees for Reinforcement Learning}, author = {Eshwar S. R. and Gugan Thoppe and Aditya Gopalan and Gal Dalal}, year = {2025}, eprint = {2506.07134}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2506.07134v1}, }