2026
Rectify Evaluation Preference: Improving LLMs’ Critique on Math Reasoning via Perplexity-aware Reinforcement Learning
AAAI 2026technical
To improve Multi-step Mathematical Reasoning (MsMR) of Large Language Models (LLMs), it is crucial to obtain scalable supervision from the corpus by automatically critiquing mistakes in the reasoning process of MsMR and rendering a final verdict of the problem-solution. Most existing methods rely on