@article{stepopsdstepawareonlinepreferencedistill,
  title = {StepOPSD: Step-Aware Online Preference Self-Distillation for Agent Reinforcement Learning},
  author = {Yanfei Zhang and Xu Lin and Chenglin Wu},
  year = {2026},
  eprint = {2605.27140},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.27140},
}