AMPO: Active Multi-Preference Optimization for Self-play Preference Selection

@misc{gupta2025ampoactivemultipreference,
      title={AMPO: Active Multi-Preference Optimization}, 
      author={Taneesh Gupta and Saravan Rajmohan and Xuchao Zhang and Chetan Bansal and Rahul Madhavan},
      year={2025},
      eprint={2502.18293},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2502.18293}, 
}