Filtered Direct Preference Optimization

Tetsuro Morimura, Mitsuki Sakamoto, Yuu Jinnai, Kenshi Abe, Kaito Ariu

Keywords: Reinforcement Learning from Human Feedback (RLHF), Direct Preference Optimization (DPO), Filtered Direct Preference Optimization (fDPO)

TL;DR: Introducing filtered Direct Preference Optimization (fDPO) that enhances language model alignment with human preferences by discarding lower-quality samples compared to those generated by the learning model

Abstract: Reinforcement learning from human feedback (RLHF) plays a crucial role in aligning language models with human preferences. While the significance of dataset quality is generally recognized, explicit investigations into its impact within the RLHF framework, to our knowledge, have been limited. This paper addresses the issue of text quality within the preference dataset by focusing on direct preference optimization (DPO), an increasingly adopted reward-model-free RLHF method. We confirm that text quality significantly influences the performance of models optimized with DPO more than those optimized with reward-model-based RLHF. Building on this new insight, we propose an extension of DPO, termed filtered direct preference optimization (fDPO). fDPO uses a trained reward model to monitor the quality of texts within the preference dataset during DPO training. Samples of lower quality are discarded based on comparisons with texts generated by the model being optimized, resulting in a more accurate dataset. Experimental results demonstrate that fDPO enhances the final model performance. Our code is available at https://github.com/CyberAgentAILab/filtered-dpo.

Submission Number: 42