Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning Paper • 2510.19338 • Published 11 days ago • 101
Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model Paper • 2510.18855 • Published 12 days ago • 62
Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning Paper • 2510.19338 • Published 11 days ago • 101
WSM: Decay-Free Learning Rate Schedule via Checkpoint Merging for LLM Pre-training Paper • 2507.17634 • Published Jul 23 • 2
Ring Collection Ring is a reasoning MoE LLM provided and open-sourced by InclusionAI, derived from Ling. • 5 items • Updated 20 days ago • 19