Accelerating Speculative Decoding with Block Diffusion Draft Trees Paper • 2604.12989 • Published 5 days ago • 6
KnowRL: Boosting LLM Reasoning via Reinforcement Learning with Minimal-Sufficient Knowledge Guidance Paper • 2604.12627 • Published 5 days ago • 96
Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe Paper • 2604.13016 • Published 5 days ago • 77