Search papers, labs, and topics across Lattice.
1
0
3
Encoding "what-is-where" at the pixel level in a single token unlocks SOTA performance in vision-based robot policy learning.
