# VRPO: Rethinking Value Modeling for Robust RL under Noisy Supervision in LLM Post-Training

arXiv: https://arxiv.org/abs/2508.03058 · HTML: https://mlsanity.com/paper/2508.03058/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning.
- Builds on 34 of its 39 references that are in the map.
- First posted on arXiv 5 Aug 2025.

## Placed in

- LLM reasoning / [Reinforcement learning for reasoning](https://mlsanity.com/lr/reinforcement-learning-for-reasoning/)

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

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Solving Quantitative Reasoning Problems with Language Models](https://mlsanity.com/paper/2206.14858/) (2022)
- [s1: Simple test-time scaling](https://mlsanity.com/paper/2501.19393/) (2025)
- [Scaling Laws for Reward Model Overoptimization](https://mlsanity.com/paper/2210.10760/) (2022)
- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [Humanity's Last Exam](https://mlsanity.com/paper/2501.14249/) (2025)
- [Measuring short-form factuality in large language models](https://mlsanity.com/paper/2411.04368/) (2024)
- [Reward learning from human preferences and demonstrations in Atari](https://mlsanity.com/paper/1811.06521/) (2018)
- [Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyond](https://mlsanity.com/paper/2503.10460/) (2025)
- [Secrets of RLHF in Large Language Models Part II: Reward Modeling](https://mlsanity.com/paper/2401.06080/) (2024)
- [SMARTS: Scalable Multi-Agent Reinforcement Learning Training School for Autonomous Driving](https://mlsanity.com/paper/2010.09776/) (2020)
- [Towards Human-Level Bimanual Dexterous Manipulation with Reinforcement Learning](https://mlsanity.com/paper/2206.08686/) (2022)
- [RLHF Deciphered: A Critical Analysis of Reinforcement Learning from Human Feedback for LLMs](https://mlsanity.com/paper/2404.08555/) (2024)
- [Reasoning or Memorization? Unreliable Results of Reinforcement Learning Due to Data Contamination](https://mlsanity.com/paper/2507.10532/) (2025)
- [InfoRM: Mitigating Reward Hacking in RLHF via Information-Theoretic Reward Modeling](https://mlsanity.com/paper/2402.09345/) (2024)
- [SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines](https://mlsanity.com/paper/2502.14739/) (2025)
- [InfoBot: Transfer and Exploration via the Information Bottleneck](https://mlsanity.com/paper/1901.10902/) (2019)
- [Compressing Neural Networks using the Variational Information Bottleneck](https://mlsanity.com/paper/1802.10399/) (2018)
- [Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization](https://mlsanity.com/paper/2407.07880/) (2024)
- [Correlated Proxies: A New Definition and Improved Mitigation for Reward Hacking](https://mlsanity.com/paper/2403.03185/) (2024)
- [RIME: Robust Preference-based Reinforcement Learning with Noisy Preferences](https://mlsanity.com/paper/2402.17257/) (2024)
- [Inverse Reinforcement Learning with Dynamic Reward Scaling for LLM Alignment](https://mlsanity.com/paper/2503.18991/) (2025)
- [KTAE: A Model-Free Algorithm to Key-Tokens Advantage Estimation in Mathematical Reasoning](https://mlsanity.com/paper/2505.16826/) (2025)
- [MINER: Improving Out-of-Vocabulary Named Entity Recognition from an Information Theoretic Perspective](https://mlsanity.com/paper/2204.04391/) (2022)
- [ROPO: Robust Preference Optimization for Large Language Models](https://mlsanity.com/paper/2404.04102/) (2024)
- [RoVRM: A Robust Visual Reward Model Optimized via Auxiliary Textual Preference Data](https://mlsanity.com/paper/2408.12109/) (2024)

## Built on by (most cited first, 9 of 9)

- [A Survey of Reinforcement Learning for Large Reasoning Models](https://mlsanity.com/paper/2509.08827/) (2025)
- [Asymmetric Proximal Policy Optimization: mini-critics boost LLM reasoning](https://mlsanity.com/paper/2510.01656/) (2025)
- [DiRL: An Efficient Post-Training Framework for Diffusion Language Models](https://mlsanity.com/paper/2512.22234/) (2025)
- [V0: A Generalist Value Model for Any Policy at State Zero](https://mlsanity.com/paper/2602.03584/) (2026)
- [SHAPE: Stage-aware Hierarchical Advantage via Potential Estimation for LLM Reasoning](https://mlsanity.com/paper/2604.06636/) (2026)
- [IACM-RL: Intent-Aware Context Management and Reinforcement Learning for Complex Tool Invocation under Dynamic Intent Fluctuations](https://mlsanity.com/paper/2608.02110/) (2026)
- [DFPO: Scaling Value Modeling via Distributional Flow towards Robust and Generalizable LLM Post-Training](https://mlsanity.com/paper/2602.05890/) (2026)
- [FlexRec: Adapting LLM-based Recommenders for Flexible Needs via Reinforcement Learning](https://mlsanity.com/paper/2603.11901/) (2026)
- [Explicit Critic Guidance for Aligning Diffusion Models](https://mlsanity.com/paper/2605.27736/) (2026)

## Abstract (arXiv)

Reinforcement Learning (RL) in real-world environments often suffers from ambiguous or incomplete reward supervision, which undermines policy stability and generalization. Such noise may cause models to ignore key information or even collapse in advantage estimation. We find that a strong value model is essential for absorbing unstable signals and producing reliable advantages, offering denser and more robust supervision than the reward model. To better optimize noisy supervision, we propose VRPO, a framework that enhances value modeling for robust RL in LLM post-training. VRPO integrates (1) auxiliary losses guided by entropy and perplexity from a frozen language model, and (2) a variational information bottleneck, enabling the value model to filter noise and capture key words. This design allows the value model to correct noise rewards and generate more reliable advantage estimates, transforming it from a passive predictor into an active noise regulator. Experiments on multi-turn dialogue, math reasoning, and science QA with both rule-based and model-based rewards show that VRPO consistently outperforms baselines such as PPO and GRPO. Our work highlight the central role of the value model in Robust RL and provide a principled and practical approach to policy optimization under noisy supervision.