2025
Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
NeurIPS 2025poster
Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a powerful approach to enhancing the reasoning capabilities of Large Language Models (LLMs), yet its underlying mechanisms remain insufficiently understood. In this work, we undertake a pioneering exploration of RLVR through the no…