@article{understandingtheperformancegapin,
  title = {Understanding the Performance Gap in Preference Learning: A Dichotomy of RLHF and DPO},
  author = {Ruizhe Shi and Minhak Song and Runlong Zhou and Zihan Zhang and Maryam Fazel and Simon S. Du},
  year = {2025},
  eprint = {2505.19770},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2505.19770v1},
}