cs.LGJun 1, 2025

Doubly Robust Alignment for Large Language Models

Authors: Erhan Xu, Kai Ye, Hongyi Zhou, Luhan Zhu, Francesco Quinzan, Chengchun Shi

Organizations: Department of Statistics, LSE, London, UK · Department of Mathematics, Tsinghua University, Beijing, China · School of Design, LCC, UAL, London, UK · Department of Engineering Science, University of Oxford, Oxford, UK

Abstract

This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many algorithms are highly sensitive to misspecifications in the underlying preference model (e.g., the Bradley-Terry model), the reference policy, or the reward function, resulting in undesirable fine-tuning. To address model misspecification, we propose a doubly robust preference optimization algorithm that remains consistent when either the preference model or the reference policy is correctly specified (without requiring both). Our proposal demonstrates superior and more robust performance than state-of-the-art algorithms, both in theory and in practice. The code is available at https://github.com/DRPO4LLM/DRPO4LLM

Figures & tables

Explore similar work

CardsList
  1. A Regret Minimization Framework on Preference Learning in Large Language Models

    Jun 8, 2026Suhwan Kim, Taehyun Cho, Geon-Hyeong Kim +4Regret Minimization in RLLLM Alignment

  2. Unbiased Alignment for Large Language Models with Noisy Preferences

    Jul 3, 2026Jialiang Wang, Xianming Liu, Xiong Zhou +2Pairwise Preference LearningLLM Alignment

  3. Distortion of AI Alignment Revisited: RLHF is a Decent Utilitarian Aligner

    Sep 11, 2026Kazusato Oko, Annie Ulichney, Nika Haghtalab +1LLM AlignmentPreference Alignment