Search papers, labs, and topics across Lattice.
1
0
3
Sharper text-to-image alignment is now possible in diffusion models by explicitly aggregating related attention and isolating unrelated attention.
