Search papers, labs, and topics across Lattice.
1
0
2
Forget hand-tuning reward functions: this framework learns to control a 700-muscle human musculoskeletal model well enough to do cartwheels and backflips, just from adversarial imitation.
