Ordinal Preference Optimization: Aligning Human Preferences via NDCG

2024-10-06

arXiv.org

AI Summary

Existing Bradley–Terry–based alignment methods (e.g., DPO, RLHF) rely on pairwise comparisons, which inadequately capture the full ranking structure among multiple responses and thus limit alignment with diverse human preferences. This work introduces, for the first time, the information retrieval metric Normalized Discounted Cumulative Gain (NDCG) into LLM alignment, proposing a differentiable ordinal preference optimization framework. We construct a surrogate loss via a differentiable NDCG approximation to enable end-to-end multi-response ranking optimization. Furthermore, we integrate ordinal preference modeling with a negative-sample pool expansion strategy to mitigate interference from trivial negatives. Extensive evaluation on benchmarks—including AlpacaEval—demonstrates significant improvements over DPO and RLHF, validating the method’s effectiveness in enhancing both response quality and fidelity to human ordinal preferences.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Search and Optimization: Learning to Search
• Humans and AI: Learning Human Values and Preferences

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Aligning Large Language Models (LLMs) with diverse human preferences is a pivotal technique for controlling model behaviors and enhancing generation quality. Reinforcement Learning from Human Feedback (RLHF), Direct Preference Optimization (DPO), and their variants optimize language models by pairwise comparisons. However, when multiple responses are available, these approaches fall short of leveraging the extensive information in the ranking given by the reward models or human feedback. In this work, we propose a novel listwise approach named Ordinal Preference Optimization (OPO), which employs the Normalized Discounted Cumulative Gain (NDCG), a widely-used ranking metric, to better utilize relative proximity within ordinal multiple responses. We develop an end-to-end preference optimization algorithm by approximating NDCG with a differentiable surrogate loss. This approach builds a connection between ranking models in information retrieval and the alignment problem. In aligning multi-response datasets assigned with ordinal rewards, OPO outperforms existing pairwise and listwise approaches on evaluation sets and general benchmarks like AlpacaEval. Moreover, we demonstrate that increasing the pool of negative samples can enhance model performance by reducing the adverse effects of trivial negatives.

Problem

• Aligning LLMs with human preferences for controllable behaviors
• Addressing limitations of pairwise methods in listwise ranking tasks
• Improving ranking accuracy using NDCG-based alignment approach

Innovation

• Uses listwise ranking with NDCG metric
• Develops differentiable surrogate loss approximation
• Outperforms pairwise methods on benchmarks