VCRL: Variance-based Curriculum Reinforcement Learning for Large Language Models Paper • 2509.19803 • Published Sep 24 • 120 • 2
PVPO: Pre-Estimated Value-Based Policy Optimization for Agentic Reasoning Paper • 2508.21104 • Published Aug 28 • 35 • 2
FlashThink: An Early Exit Method For Efficient Reasoning Paper • 2505.13949 • Published May 20 • 1 • 2