Beyond High-Entropy Exploration: Correctness-Aware Low-Entropy Segment-Based Advantage Shaping for Reasoning LLMs
This work addresses the challenge of enhancing reasoning abilities in large language models for tasks like math problem-solving, representing an incremental advancement by refining existing RLVR methods with a novel segment-based approach.
The paper tackles the problem of improving reasoning in large language models by focusing on low-entropy segments in reasoning trajectories, finding that overlaps of these segments in correct responses correlate with accuracy, and proposes LESS, a correctness-aware reinforcement framework that modulates advantages over low-entropy segments, resulting in consistent accuracy improvements over strong RL baselines across three backbones and six math benchmarks.
Reinforcement Learning with Verifiable Rewards (RLVR) has become a central approach for improving the reasoning ability of large language models. Recent work studies RLVR through token entropy, arguing that high-entropy tokens drive exploration and should receive stronger updates. However, they overlook the fact that most of a reasoning trajectory consists of low-entropy segments that encode stable and reusable structural patterns. Through qualitative and quantitative analyses, we find that the overlap of low-entropy segments across correct responses strongly correlates with model accuracy, while overlaps involving incorrect responses exhibit stable but unproductive patterns. Motivated by these findings, we propose LESS, a correctness-aware reinforcement framework that performs fine-grained advantage modulation over low-entropy segments. LESS amplifies segments unique to correct responses, suppresses those unique to incorrect ones, and neutralizes segments shared by both, while preserving high-entropy exploration in the underlying RL algorithm. Instantiated on top of the popular GRPO, LESS consistently improves accuracy over strong RL baselines across three backbones and six math benchmarks, achieves stronger robustness of the performance floor.