Search papers, labs, and topics across Lattice.
1
0
2
4
Forget text-dominance: Today's Omni-modal LLMs surprisingly favor visual inputs, creating new challenges for cross-modal reasoning.
