@article{demystifyingdesignchoicesofreinforcement,
  title = {Demystifying Design Choices of Reinforcement Fine-tuning: A Batched Contextual Bandit Learning Perspective},
  author = {Hong Xie and Xiao Hu and Tao Tan and Haoran Gu and Xin Li and Jianyu Han and Defu Lian and Enhong Chen},
  year = {2026},
  eprint = {2601.22532},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2601.22532},
}