Mitigating Forgetting Between Supervised and Reinforcement Learning Yields Stronger Reasoners

@misc{yuan2025mitigating,
  title = {Mitigating Forgetting Between Supervised and Reinforcement Learning Yields Stronger Reasoners},
  author = {Xiangchi Yuan and Xiang Chen and Tong Yu and Dachuan Shi and Can Jin and Wenke Lee and Saayan Mitra},
  year = {2025},
  eprint = {2510.04454},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2510.04454},
}