Iterative Nash Policy Optimization: Aligning LLMs with General Preferences via No-Regret Learning

TL;DR

This paper introduces INPO, a novel no-regret learning algorithm for aligning large language models with human preferences by framing RLHF as a game and avoiding costly response evaluations.

Contribution

It proposes a game-theoretic approach with an online algorithm that bypasses traditional reward estimation, improving efficiency and effectiveness in LLM alignment.

Findings

Achieves 42.6% win rate on AlpacaEval 2.0

Attains 37.8% win rate on Arena-Hard

Outperforms existing online RLHF methods

Abstract

Reinforcement Learning with Human Feedback (RLHF) has achieved great success in aligning large language models (LLMs) with human preferences. Prevalent RLHF approaches are reward-based, following the Bradley-Terry (BT) model assumption, which may not fully capture the complexity of human preferences. In this paper, we explore RLHF under a general preference framework and approach it from a game-theoretic perspective. Specifically, we formulate the problem as a two-player game and propose a novel online algorithm, iterative Nash policy optimization (INPO). The key idea is to let the policy play against itself via no-regret learning, thereby approximating the Nash policy. Unlike previous methods, INPO bypasses the need for estimating the expected win rate for individual responses, which typically incurs high computational or annotation costs. Instead, we introduce a new loss objective that is directly minimized over a preference dataset. We provide theoretical analysis for our approach and demonstrate its effectiveness through experiments on various representative benchmarks. 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, showing substantial improvement over the state-of-the-art online RLHF algorithms.