Search papers, labs, and topics across Lattice.
University of California, San Diego
1
0
2
13
Stop hand-crafting hints for RL agents: HiLL learns to generate adaptive hints that actually improve the agent's performance on the original task, not just the hinted one.
