Search papers, labs, and topics across Lattice.
1
0
GORDON achieves a remarkable 74.4% success rate in long-horizon manipulation tasks by automatically discovering subtasks and learning dense object-centric rewards from video demonstrations.