Search papers, labs, and topics across Lattice.
University of California
1
0
2
Cloning synthetic demonstrations alone caps performance, but combining it with reinforcement learning unlocks superior real-world manipulation capabilities.
