Junpeng Yue*, Boyuan Li*, Yuxuan Wang*, Zepeng Wang, Yuhui Fu, Feiyang Xie, Yu Zhang, Jing Zhang, Jiangxing Wang, Zongqing Lu
Technical Report, 2026 · Co-first author
Overview
Being-M0.7 learns world-action priors through joint video-motion pretraining and transfers them to humanoid robot control. I contributed the model design, training, and evaluation of its Video-Motion pretraining module.