@article{zhong2026diagnosing,
  title={Diagnosing Training Inference Mismatch in LLM Reinforcement Learning},
  author={Zhong, Tianle and Ling, Neiwen and Pi, Yifan and Wei, Zijun and Yu, Tianshu and Fox, Geoffrey and Wu, Peng and Yu, Xiao},
  journal={arXiv preprint arXiv:2605.14220},
  year={2026}
}