@article{opsdderivingwithpolicyoptimizationtraini,
  title = {β-OPSD: Deriving with Policy Optimization, Training with Self-Distillation},
  author = {Jiawei Xu and Minghui Liu and Juzheng Zhang and Tom Goldstein and Furong Huang},
  year = {2026},
  eprint = {2607.28582},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2607.28582},
}