Multiplayer Nash Preference Optimization

• URL: http://arxiv.org/abs/2509.23102v1
• Date: Sat, 27 Sep 2025 04:18:33 GMT
• Title: Multiplayer Nash Preference Optimization
• Authors: Fang Wu, Xu Huang, Weihao Xuan, Zhiwei Zhang, Yijia Xiao, Guancheng Wan, Xiaomin Li, Bing Hu, Peng Xia, Jure Leskovec, Yejin Choi,
• Abstract summary: Reinforcement learning from human feedback (RLHF) has emerged as the standard paradigm for aligning large language models (LLMs) with human preferences.<n>Recent studies have reframed alignment as a two-player Nash game, giving rise to Nash learning from human feedback (NLHF)<n>We introduce Multiplayer Nash Preference Optimization (MNPO), a novel framework that generalizes NLHF to the multiplayer regime.
• Score: 79.15013211640566
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement learning from human feedback (RLHF) has emerged as the standard paradigm for aligning large language models (LLMs) with human preferences. However, reward-based methods built on the Bradley-Terry assumption struggle to capture the non-transitive and heterogeneous nature of real-world preferences. To address this, recent studies have reframed alignment as a two-player Nash game, giving rise to Nash learning from human feedback (NLHF). While this perspective has inspired algorithms such as INPO, ONPO, and EGPO with strong theoretical and empirical guarantees, they remain fundamentally restricted to two-player interactions, creating a single-opponent bias that fails to capture the full complexity of realistic preference structures. In this work, we introduce Multiplayer Nash Preference Optimization (MNPO), a novel framework that generalizes NLHF to the multiplayer regime. It formulates alignment as an $n$-player game, where each policy competes against a population of opponents while being regularized toward a reference model. Our framework establishes well-defined Nash equilibria in multiplayer settings and extends the concept of duality gap to quantify approximation quality. We demonstrate that MNPO inherits the equilibrium guarantees of two-player methods while enabling richer competitive dynamics and improved coverage of diverse preference structures. Through comprehensive empirical evaluation, we show that MNPO consistently outperforms existing NLHF baselines on instruction-following benchmarks, achieving superior alignment quality under heterogeneous annotator conditions and mixed-policy evaluation scenarios. Together, these results establish MNPO as a principled and scalable framework for aligning LLMs with complex, non-transitive human preferences. Code is available at https://github.com/smiles724/MNPO.

Related papers

• Fundamental Limits of Game-Theoretic LLM Alignment: Smith Consistency and Preference Matching [23.0436612817548]
  Nash Learning from Human Feedback is a framework for aligning large language models with human preferences by modeling learning as a zero-sum game.<n>In this paper, we study using what choices of payoff based on the pairwise human preferences can yield desirable alignment properties.
  arXiv  Detail & Related papers  (2025-05-27T02:07:35Z)
• Improving LLM General Preference Alignment via Optimistic Online Mirror Descent [57.622821649679786]
  Reinforcement learning from human feedback (RLHF) has demonstrated remarkable effectiveness in aligning large language models (LLMs) with human preferences.<n>In this paper, we drop the Bradley-Terry (BT) model assumption and study LLM alignment under general preferences, formulated as a two-player game.<n>We show that our approach achieves an $O(T-1)$ bound on the duality gap, improving upon the previous $O(T-1/2)$ result.
  arXiv  Detail & Related papers  (2025-02-24T05:24:52Z)
• Multi-Step Alignment as Markov Games: An Optimistic Online Gradient Descent Approach with Convergence Guarantees [91.88803125231189]
  Reinforcement Learning from Human Feedback (RLHF) has been highly successful in aligning large language models with human preferences.<n>While prevalent methods like DPO have demonstrated strong performance, they frame interactions with the language model as a bandit problem.<n>In this paper, we address these challenges by modeling the alignment problem as a two-player constant-sum Markov game.
  arXiv  Detail & Related papers  (2025-02-18T09:33:48Z)
• COMAL: A Convergent Meta-Algorithm for Aligning LLMs with General Preferences [57.70902561665269]
  We propose a meta-algorithm, Convergent Meta Alignment Algorithm (COMAL), for language model alignment with general preferences.<n>We provide theoretical analysis that our meta-algorithm converges to an exact Nash policy in the last iterate and demonstrate its effectiveness on a range of synthetic and preference optimization datasets.
  arXiv  Detail & Related papers  (2024-10-30T17:13:02Z)
• Iterative Nash Policy Optimization: Aligning LLMs with General Preferences via No-Regret Learning [55.65738319966385]
  We propose a novel online algorithm, iterative Nash policy optimization (INPO)<n>Unlike previous methods, INPO bypasses the need for estimating the expected win rate for individual responses.<n>With an LLaMA-3-8B-based SFT model, INPO achieves a 42.6% length-controlled win rate on AlpacaEval 2.0 and a 37.8% win rate on Arena-Hard.
  arXiv  Detail & Related papers  (2024-06-30T08:00:34Z)
• Nash Learning from Human Feedback [86.09617990412941]
  We introduce an alternative pipeline for the fine-tuning of large language models using pairwise human feedback. We term this approach Nash learning from human feedback (NLHF) We present a novel algorithmic solution, Nash-MD, founded on the principles of mirror descent.
  arXiv  Detail & Related papers  (2023-12-01T19:26:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.