📝 SummarXiv

Abstract

Reinforcement Learning from Human Feedback (RLHF) reward models exhibit systematic failures on longtail distributions, leading to reward hacking and misalignment. We propose a mechanistic interpretability framework that identifies specialized neural circuits responsible for rare-event processing in reward models. Drawing from recent advances showing distributed specialization for rare tokens in language models\citep{liu2025no, liu2025emergent}, we hypothesize that reward models also develop functionally distinct circuits for longtail scenarios. Our theoretical framework establishes formal connections between circuit specialization, reward generalization bounds, and longtail performance. We introduce \textbf{Circuit-Aware Reward Training (CART)}, which uses circuit analysis to guide data augmentation, regularization, and ensemble strategies. This approach provides both theoretical insights into reward model failures and practical interventions for improving longtail robustness.