Search papers, labs, and topics across Lattice.
1
0
2
LLMs can transform sparse reward signals into structured learning paths, leading to better performance in complex manipulation tasks than traditional dense reward systems.
