Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning Paper • 2506.01939 • Published 6 days ago • 132
ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models Paper • 2505.24864 • Published 9 days ago • 115
ZeroGUI: Automating Online GUI Learning at Zero Human Cost Paper • 2505.23762 • Published 10 days ago • 45
ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows Paper • 2505.19897 • Published 13 days ago • 101