Search papers, labs, and topics across Lattice.
Affiliation:
1
0
3
MLLMs exhibit substantial reasoning deficits in text-rich image understanding, failing to apply visible rules and often providing incorrect answers despite compliant rationales.
