@article{utilizingandcalibratinghindsightprocessr, title = {Utilizing and Calibrating Hindsight Process Rewards via Reinforcement with Mutual Information Self-Evaluation}, author = {Jiashu Yao and Heyan Huang and Zeming Liu and Yuhang Guo}, year = {2026}, eprint = {2604.11611}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2604.11611}, }