Papers › Ordinal Preference Optimization: Aligning Human Preferences via NDCG

Ordinal Preference Optimization: Aligning Human Preferences via NDCG

6 Oct 2024arXiv:2410.04346archive 2025-07-28

Yang Zhao, Yixin Wang, Mingzhang Yin

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.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

zhaoyang02/ordinal-preference-optimization officialmentioned in paperpytorchnot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-ready report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Information Retrieval

Datasets

Introduced by this paper, per the archive.

ListUltraFeedback

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections