Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning Paper • 2506.01939 • Published 20 days ago • 160
Recitation over Reasoning: How Cutting-Edge Language Models Can Fail on Elementary School-Level Reasoning Problems? Paper • 2504.00509 • Published Apr 1 • 22
view article Article Welcome Gemma 3: Google's all new multimodal, multilingual, long context open LLM By ariG23498 and 3 others • Mar 12 • 435
Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models Paper • 2503.16419 • Published Mar 20 • 74
SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines Paper • 2502.14739 • Published Feb 20 • 104