MP1: MeanFlow Tames Policy Learning in 1-step for Robotic Manipulation
MP1: 均值流使政策学习在1步中适用于机器人操作
专题命中 机器人学习 :manipulation(title,abstract);robotic(title);robot learning(abstract);分类 cs.RO
AI总结 MP1通过MeanFlow范式和轻量级分散损失,在1-NFE中实现更精确且可控的动作轨迹生成,优于现有方法。
Comments This paper has been accepted by AAAI 2026