Optimal Transport for LLM Reward Modeling from Noisy Preference

• URL: http://arxiv.org/abs/2605.06036v1
• Date: Thu, 07 May 2026 11:26:04 GMT
• Title: Optimal Transport for LLM Reward Modeling from Noisy Preference
• Abstract summary: We propose SelectiveRM, a framework grounded in optimal transport.<n>We first devise a Joint Consistency Discrepancy to align the distribution of model predictions with preference data.<n>We then incorporate a Mass Relaxation mechanism via partial transport.<n>This enables the autonomous exclusion of samples with noisy preference that contradict semantic consistency.
• Score: 21.518348751984433
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reward models are fundamental to Reinforcement Learning from Human Feedback (RLHF), yet real-world datasets are inevitably corrupted by noisy preference. Conventional training objectives tend to overfit these errors, while existing denoising approaches often rely on homogeneous noise assumptions that fail to capture the complexity of linguistic preferences. To handle these challenges, we propose SelectiveRM, a framework grounded in optimal transport. We first devise a Joint Consistency Discrepancy to align the distribution of model predictions with preference data. Furthermore, to address the limitation of strict mass conservation which compels the model to fit outliers, we incorporate a Mass Relaxation mechanism via partial transport. This enables the autonomous exclusion of samples with noisy preference that contradict semantic consistency. Theoretically, we demonstrate that SelectiveRM optimizes a tighter upper bound on the unobserved clean risk. Extensive experiments validate that our approach significantly outperforms state-of-the-art baselines across diverse benchmarks.

Related papers

• Preference Data Selection for Mitigating the Alignment Tax in Large Language Models [52.18310687523229]
  BALIGN is a balanced data selection strategy that mitigates catastrophic forgetting while optimizing alignment efficacy.<n>By aggregating three key data-centric features into a unified composite risk score, BALIGN systematically filters out high-risk preference samples.<n>Experiments on standard human preference datasets demonstrate that BALIGN strongly preserves foundational capabilities without compromising alignment gains.
  arXiv  Detail & Related papers  (2026-08-25T07:58:42Z)
• Unbiased Alignment for Large Language Models with Noisy Preferences [74.69408994904965]
  We present a theoretical framework for unbiased alignment, introducing the Unbiased Reward Model (URM) loss and the Unbiased Direct Preference Optimization (UDPO) loss.<n>By mathematically correcting the distortion induced by preference noise, our novel objectives enable unbiased model training directly from noisy datasets.
  arXiv  Detail & Related papers  (2026-07-03T12:04:43Z)
• Joint Model and Data Sparsification via the Marginal Likelihood [53.29070892356214]
  We propose jointly learning individual feature and sample relevancies, enabling simultaneous model and data sparsification.<n>This symmetric pruning of model and data offers a natural extension that preserves conjugacy.<n> Empirical results across diverse regression tasks affirm that a joint ARD approach consistently yields both sparse and robust prediction models.
  arXiv  Detail & Related papers  (2026-05-28T13:26:53Z)
• Anomaly-Preference Image Generation [41.96735569500412]
  We introduce Anomaly Preference Optimization, a novel paradigm that reformulates anomaly generation as a preference learning problem.<n>We show that our approach significantly outperforms existing baselines,achieving state-of-the-art performance in both realism and diversity.
  arXiv  Detail & Related papers  (2026-05-04T10:37:09Z)
• Too Correct to Learn: Reinforcement Learning on Saturated Reasoning Data [55.84428098924793]
  We propose Constrained Uniform Top-K Sampling (CUTS), a parameter-free decoding strategy enforcing structure-preserving exploration.<n>We integrate this into Mixed-CUTS, a training framework synergizing exploitative and exploratory rollouts to amplify intra-group advantage variance.<n> Notably, Mixed-CUTS improves Pass@1 accuracy on the challenging AIME25 benchmark by up to 15.1% over standard GRPO.
  arXiv  Detail & Related papers  (2026-04-20T16:43:28Z)
• CDRRM: Contrast-Driven Rubric Generation for Reliable and Interpretable Reward Modeling [61.75914342638658]
  CDRRM (Contrast-Driven Reward Model) is a framework for high-quality rubric generation and guided preference judgment.<n>This work offers a scalable, interpretable, and data-efficient path for reward modeling.
  arXiv  Detail & Related papers  (2026-03-09T07:15:23Z)
• Mitigating Reward Hacking in RLHF via Bayesian Non-negative Reward Modeling [49.41422138354821]
  We propose a principled reward modeling framework that integrates non-negative factor analysis into the Bradley-Terry preference model.<n>BNRM represents rewards through a sparse, non-negative latent factor generative process.<n>We show that BNRM substantially mitigates reward over-optimization, improves robustness under distribution shifts, and yields more interpretable reward decompositions than strong baselines.
  arXiv  Detail & Related papers  (2026-02-11T08:14:11Z)
• How Well Can Preference Optimization Generalize Under Noisy Feedback? [7.374590753074647]
  Preference optimization trains models to distinguish between preferred and non-preferred responses based on human feedback.<n>Most existing works assume noise-free feedback, which is unrealistic due to the inherent errors and inconsistencies in human judgments.<n>This paper addresses the impact of noisy feedback on preference optimization, providing generalization guarantees under these conditions.
  arXiv  Detail & Related papers  (2025-10-01T20:56:31Z)
• Self-Evolutionary Large Language Models through Uncertainty-Enhanced Preference Optimization [9.618391485742968]
  Iterative preference optimization has recently become one of the de-facto training paradigms for large language models (LLMs) We present an uncertainty-enhanced textbfPreference textbfOptimization framework to make the LLM self-evolve with reliable feedback. Our framework substantially alleviates the noisy problem and improves the performance of iterative preference optimization.
  arXiv  Detail & Related papers  (2024-09-17T14:05:58Z)
• ROPO: Robust Preference Optimization for Large Language Models [59.10763211091664]
  We propose an iterative alignment approach that integrates noise-tolerance and filtering of noisy samples without the aid of external models. Experiments on three widely-used datasets with Mistral-7B and Llama-2-7B demonstrate that ROPO significantly outperforms existing preference alignment methods.
  arXiv  Detail & Related papers  (2024-04-05T13:58:51Z)
• Direct Preference Optimization: Your Language Model is Secretly a Reward Model [119.65409513119963]
  We introduce a new parameterization of the reward model in RLHF that enables extraction of the corresponding optimal policy in closed form. The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight. Our experiments show that DPO can fine-tune LMs to align with human preferences as well as or better than existing methods.
  arXiv  Detail & Related papers  (2023-05-29T17:57:46Z)

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.