Search papers, labs, and topics across Lattice.
1
0
3
Forget scaling data – scaling the *number of tasks* unlocks surprisingly sample-efficient humanoid control via multi-task model-based RL.
