Search papers, labs, and topics across Lattice.
1
0
2
Spatial perception in VLA models can be dramatically enhanced by integrating learnable tokens that rectify attention artifacts and refine local information.