Ask a Strong LLM Judge when Your Reward Model is Uncertain

• URL: http://arxiv.org/abs/2510.20369v1
• Date: Thu, 23 Oct 2025 09:09:13 GMT
• Title: Ask a Strong LLM Judge when Your Reward Model is Uncertain
• Authors: Zhenghao Xu, Qin Lu, Qingru Zhang, Liang Qiu, Ilgee Hong, Changlong Yu, Wenlin Yao, Yao Liu, Haoming Jiang, Lihong Li, Hyokun Yun, Tuo Zhao,
• Abstract summary: We propose an uncertainty-based routing framework that efficiently complements a fast RM with a strong but costly LLM judge.<n>Our approach formulates advantage estimation in policy gradient (PG) methods as pairwise preference classification.<n> Experiments on RM benchmarks demonstrate that our uncertainty-based routing strategy significantly outperforms random judge calling at the same cost.
• Score: 46.41334493044746
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reward model (RM) plays a pivotal role in reinforcement learning with human feedback (RLHF) for aligning large language models (LLMs). However, classical RMs trained on human preferences are vulnerable to reward hacking and generalize poorly to out-of-distribution (OOD) inputs. By contrast, strong LLM judges equipped with reasoning capabilities demonstrate superior generalization, even without additional training, but incur significantly higher inference costs, limiting their applicability in online RLHF. In this work, we propose an uncertainty-based routing framework that efficiently complements a fast RM with a strong but costly LLM judge. Our approach formulates advantage estimation in policy gradient (PG) methods as pairwise preference classification, enabling principled uncertainty quantification to guide routing. Uncertain pairs are forwarded to the LLM judge, while confident ones are evaluated by the RM. Experiments on RM benchmarks demonstrate that our uncertainty-based routing strategy significantly outperforms random judge calling at the same cost, and downstream alignment results showcase its effectiveness in improving online RLHF.

Related papers

• GDRO: Group-level Reward Post-training Suitable for Diffusion Models [55.948229011478304]
  Group-level rewards successfully align the model with the targeted reward.<n>Group-level Direct Reward Optimization (GDRO) is a new post-training paradigm for group-level reward alignment.<n>GDRO supports full offline training that saves the large time cost for image rollout sampling.<n>It is diffusion-sampler-independent, which eliminates the need for the ODE-to-SDE approximation to obtainity.
  arXiv  Detail & Related papers  (2026-01-05T11:47:18Z)
• Confidence as a Reward: Transforming LLMs into Reward Models [54.98336080630691]
  Confidence-as-a-Reward (CRew) is a training-free method that utilizes token-level confidence in the model's final answers as a proxy for reward.<n>We show that CRew outperforms existing training-free reward approaches on the MATH500 and RewardMATH benchmarks.<n>We propose CRew-DPO, a training strategy that constructs preference data from confidence scores combined with correctness signals.
  arXiv  Detail & Related papers  (2025-10-15T12:51:47Z)
• APLOT: Robust Reward Modeling via Adaptive Preference Learning with Optimal Transport [37.21695864040979]
  The reward model (RM) plays a crucial role in aligning Large Language Models (LLMs) with human preferences through Reinforcement Learning.<n>This paper introduces an effective enhancement to BT-based RMs through an adaptive margin mechanism.
  arXiv  Detail & Related papers  (2025-10-13T03:13:28Z)
• Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains [9.917318870162365]
  Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective for complex reasoning tasks with clear correctness signals such as math and coding.<n>Rubrics have recently been used in evaluation benchmarks to capture such judgments, but their potential as reward signals for on-policy post-training remains underexplored.<n>We introduce RaR, an on-policy reinforcement learning method that extends RLVR beyond verifiable domains by using rubric-based feedback.
  arXiv  Detail & Related papers  (2025-07-23T17:57:55Z)
• Off-Policy Corrected Reward Modeling for Reinforcement Learning from Human Feedback [52.1410307583181]
  We useReinforcement Learning from Human Feedback to train language models (LMs) to follow complex human preferences.<n>As training progresses, the responses generated by the LM no longer resemble the responses seen by the reward model (RM)<n>We propose Off-Policy Corrected Reward Modeling to correct the RM using importance weighting, without requiring new labels or samples.
  arXiv  Detail & Related papers  (2025-07-21T11:19:04Z)
• Reverse Engineering Human Preferences with Reinforcement Learning [14.508050809497847]
  Large Language Models (LLMs) are routinely evaluated by other LLMs trained to predict human preferences.<n>Previous work shows that the answers generated by a candidate-LLM can be edited post hoc to maximise the score assigned to them by a judge-LLM.<n>We adopt a different approach and use the signal provided by judge-LLMs as a reward to adversarially tune models.
  arXiv  Detail & Related papers  (2025-05-21T17:48:16Z)
• Adversarial Training of Reward Models [74.17196154247964]
  We introduce Adv-RM, a novel adversarial training framework that automatically identifies adversarial examples.<n>By leveraging reinforcement learning, Adv-RM trains a policy to expose vulnerabilities in large state-of-the-art reward models.<n>We demonstrate that Adv-RM significantly outperforms conventional reward training.
  arXiv  Detail & Related papers  (2025-04-08T15:38:25Z)
• RMB: Comprehensively Benchmarking Reward Models in LLM Alignment [44.84304822376291]
  Reward models (RMs) guide the alignment of large language models (LLMs)<n>We propose RMB, a comprehensive RM benchmark that covers over 49 real-world scenarios.<n>Based on our benchmark, we conduct extensive analysis on the state-of-the-art RMs.
  arXiv  Detail & Related papers  (2024-10-13T16:06:54Z)
• WARM: On the Benefits of Weight Averaged Reward Models [63.08179139233774]
  We propose Weight Averaged Reward Models (WARM) to mitigate reward hacking. Experiments on summarization tasks, using best-of-N and RL methods, shows that WARM improves the overall quality and alignment of LLM predictions.
  arXiv  Detail & Related papers  (2024-01-22T18:27:08Z)
• RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback [5.3113139864044046]
  Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models (LLMs) with human preferences, but gathering high-quality preference labels is expensive. RLAIF offers a promising alternative that trains the reward model (RM) on preferences generated by an off-the-shelf LLM. Our results suggest that RLAIF can achieve performance on-par with using human feedback, offering a potential solution to the scalability limitations of RLHF.
  arXiv  Detail & Related papers  (2023-09-01T05:53:33Z)

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.