Search papers, labs, and topics across Lattice.
1
0
3
LVLMs can be trained more effectively by focusing only on the most visually informative samples and tokens, as measured by a new Visual Information Gain metric.
