# Optimal Transport for LLM Reward Modeling from Noisy Preference

arXiv: https://arxiv.org/abs/2605.06036 · HTML: https://mlsanity.com/paper/2605.06036/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 34 of its 48 references that are in the map.
- First posted on arXiv 7 May 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 30 of 34)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels](https://mlsanity.com/paper/1804.06872/) (2018)
- [Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach](https://mlsanity.com/paper/1609.03683/) (2016)
- [Learning from Noisy Labels with Deep Neural Networks: A Survey](https://mlsanity.com/paper/2007.08199/) (2020)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [Confident Learning: Estimating Uncertainty in Dataset Labels](https://mlsanity.com/paper/1911.00068/) (2019)
- [UltraFeedback: Boosting Language Models with Scaled AI Feedback](https://mlsanity.com/paper/2310.01377/) (2023)
- [Joint Distribution Optimal Transportation for Domain Adaptation](https://mlsanity.com/paper/1705.08848/) (2017)
- [RLHF Workflow: From Reward Modeling to Online RLHF](https://mlsanity.com/paper/2405.07863/) (2024)
- [PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference](https://mlsanity.com/paper/2406.15513/) (2024)
- [HelpSteer: Multi-attribute Helpfulness Dataset for SteerLM](https://mlsanity.com/paper/2311.09528/) (2023)
- [Unbalanced minibatch Optimal Transport; applications to Domain Adaptation](https://mlsanity.com/paper/2103.03606/) (2021)
- [Reward Shaping to Mitigate Reward Hacking in RLHF](https://mlsanity.com/paper/2502.18770/) (2025)
- [Sample Selection with Uncertainty of Losses for Learning with Noisy Labels](https://mlsanity.com/paper/2106.00445/) (2021)
- [InfoRM: Mitigating Reward Hacking in RLHF via Information-Theoretic Reward Modeling](https://mlsanity.com/paper/2402.09345/) (2024)
- [To Smooth or Not? When Label Smoothing Meets Noisy Labels](https://mlsanity.com/paper/2106.04149/) (2021)
- [Clusterability as an Alternative to Anchor Points When Learning with Noisy Labels](https://mlsanity.com/paper/2102.05291/) (2021)
- [Instance-Dependent Label-Noise Learning with Manifold-Regularized Transition Matrix Estimation](https://mlsanity.com/paper/2206.02791/) (2022)
- [Optimal Transport for Treatment Effect Estimation](https://mlsanity.com/paper/2310.18286/) (2023)
- [FFT: Towards Harmlessness Evaluation and Analysis for LLMs with Factuality, Fairness, Toxicity](https://mlsanity.com/paper/2311.18580/) (2023)
- [Early Stopping Against Label Noise Without Validation Data](https://mlsanity.com/paper/2502.07551/) (2025)
- [BadLabel: A Robust Perspective on Evaluating and Enhancing Label-noise Learning](https://mlsanity.com/paper/2305.18377/) (2023)
- [DistDF: Time-Series Forecasting Needs Joint-Distribution Wasserstein Alignment](https://mlsanity.com/paper/2510.24574/) (2025)
- [Instance-dependent Label Distribution Estimation for Learning with Label Noise](https://mlsanity.com/paper/2212.08380/) (2022)

## Built on by (most cited first, 3 of 3)

- [Diagnosis Before Recovery: Turning Agent Failures into Selective Self-Correction](https://mlsanity.com/paper/2608.11772/) (2026)
- [Rubrics on Trial: Evolving Rubrics from a Single Query via Synthetic Pairwise Evidence](https://mlsanity.com/paper/2607.15092/) (2026)
- [ARC: Fair Relative Advantage Comparison in Open-Ended Real-World Interaction](https://mlsanity.com/paper/2608.13622/) (2026)

## Abstract (arXiv)

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.