CLMay 20, 2025

Reward Reasoning Model

arXiv:2505.14674v133 citationsh-index: 41Has Code
Originality Incremental advance
AI Analysis

This addresses a critical problem in aligning large language models with human expectations, offering an incremental improvement in reward modeling through test-time reasoning.

The paper tackles the challenge of improving reward model performance by utilizing test-time compute, introducing Reward Reasoning Models (RRMs) that use chain-of-thought reasoning to generate rewards. Experimental results show RRMs achieve superior performance on reward modeling benchmarks and can adaptively exploit test-time compute to improve reward accuracy.

Reward models play a critical role in guiding large language models toward outputs that align with human expectations. However, an open challenge remains in effectively utilizing test-time compute to enhance reward model performance. In this work, we introduce Reward Reasoning Models (RRMs), which are specifically designed to execute a deliberate reasoning process before generating final rewards. Through chain-of-thought reasoning, RRMs leverage additional test-time compute for complex queries where appropriate rewards are not immediately apparent. To develop RRMs, we implement a reinforcement learning framework that fosters self-evolved reward reasoning capabilities without requiring explicit reasoning traces as training data. Experimental results demonstrate that RRMs achieve superior performance on reward modeling benchmarks across diverse domains. Notably, we show that RRMs can adaptively exploit test-time compute to further improve reward accuracy. The pretrained reward reasoning models are available at https://huggingface.co/Reward-Reasoning.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes