AAAI · 2025 · Conference paper

Pre-DPO: Improving Data Utilization in Direct Preference Optimization Using a Guiding Reference Model

Jun-Shu Pan, Wei Shen, Shulin Huang, Qiji Zhou, Yue Zhang

Published in
AAAI Conference on Artificial Intelligence
Date
2025-04-22
Citations
9
arXiv
2504.15843
Cite
@inproceedings{pan2025predpo,
  title = {Pre-DPO: Improving Data Utilization in Direct Preference Optimization Using a Guiding Reference Model},
  author = {Jun-Shu Pan and Wei Shen and Shulin Huang and Qiji Zhou and Yue Zhang},
  year = {2025},
  booktitle = {AAAI Conference on Artificial Intelligence},
  eprint = {2504.15843},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2504.15843},
}