SOURCE-LINKED INTELLIGENCE

GMTS: Gradient Magnitude-based Token Selection Improves RLVR Training for LLM Reasoning

arXiv · AI, language, vision and robotics · article · Aug 31, 2026 · UTC

Reinforcement learning (RL), particularly RL with Verifiable Rewards (RLVR), has recently emerged as a central paradigm for enhancing large language models' (LLMs) reasoning abilities, demonstrating remarkable effectiveness across reasoning tasks. Recent studies suggest that high-entropy tokens play an exceptionally important role in model training, since training with only the highest 20% entropy tokens yields significant performance gains. However, why such high-entropy tokens are beneficial remains insufficiently understood. In this work, we find that although high-entropy tokens within one

Read original source ↗ Open in workspace

recordType

paper

region

Global