Search papers, labs, and topics across Lattice.
1
0
2
MLLMs struggle to "read" missing text directly from visual context, even when they possess the necessary visual grounding and layout understanding.
