arXivZhuguanyu Wu, Congjia Chen, Ruihao Gong, Yang Yong, Shiqiao Gu, Yi Xu, Xianglong LiuFri, Oct 30, 2026, 9:05 AM PDT
score 15.3
Method Makes Few-Step Image and Video Models Trainable With Reinforcement Learning
Original: Optimizing Stochastic Policy Geometry for Few-Step Flow Reinforcement Learning
Source: arxiv.org ↗
Writing ELI5 summary…