Reinforcement Fine-Tuning Naturally Mitigates Forgetting in Continual Post-Training

@misc{lai2025reinforcement,
  title = {Reinforcement Fine-Tuning Naturally Mitigates Forgetting in Continual Post-Training},
  author = {Song Lai and Hao-Han Zhao and Rong Feng and Changyi Ma and Wenzhuo Liu and Hongbo Zhao and Xi Lin and Dong Yi and Min Xie and Qingfu Zhang and Hongbin Liu and Gaofeng Meng and Fei Zhu},
  year = {2025},
  eprint = {2507.05386},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2507.05386},
}