Search papers, labs, and topics across Lattice.
1
0
3
Forget pick-and-place: RuleSafe, a new benchmark featuring LLM-generated safe-cracking tasks, exposes the long-horizon planning weaknesses of current robot learning methods.
