Search papers, labs, and topics across Lattice.
1
0
3
Omni-LLMs struggle to identify the same objects across different modalities, but a new dataset and training strategies can significantly improve their cross-modal reasoning.
