@article{onetokenrolloutguidingsupervisedfinetuni,
  title = {One-Token Rollout: Guiding Supervised Fine-Tuning of LLMs with Policy Gradient},
  author = {Rui Ming and Haoyuan Wu and Shoubo Hu and Zhuolun He and Bei Yu},
  year = {2025},
  eprint = {2509.26313},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.26313},
}