Search papers, labs, and topics across Lattice.
University of California
2
0
1
Agents can now actively seek out sparse audio-visual evidence across long contexts, rather than just passively processing it.
Iterative local refinement in Mask Diffusion Models can dramatically enhance reasoning capabilities, outperforming traditional methods across diverse tasks.
