Search papers, labs, and topics across Lattice.
Westlake University
1
0
0
Direct reinforcement learning can master coupled whole-body locomotion and multi-fingered dexterous manipulation from human data without relying on teacher-student distillation or residual policy staging.