Search papers, labs, and topics across Lattice.
1
0
Learned driving planners can be dramatically improved in closed-loop by a simple, non-learned refinement layer that prioritizes safety and interaction over raw route progress.
