Asynchronous RL trains on stale weights. Align rollout and training kernels bit for bit, and unclipped importance sampling corrects that lag. Plus zero-mismatch RL for DeepSeek-V4.1-Flash, Kimi K3.1 and AMD MI355X.
Zero Mismatch, Simple Async RL: REINFORCE Zero, DeepSeek-V4.1 and Kimi K3.1
calendar_today
November 6, 2026
person
Ao Shen (Inferact), Jian Hu (NVIDIA Research), Yan Bai (NVIDIA DevTech), Shekhar Pandey (AMD), and the REINFORCE Zero authors
domain
vllm