Search papers, labs, and topics across Lattice.
1
0
2
Forget painstakingly tuning reward weights: this two-stage curriculum trains robots more effectively by first focusing on task completion, then layering in desired behaviors.
