Search papers, labs, and topics across Lattice.
2
0
3
7
Visual Pretraining outperforms text-only methods, revealing that rich visual cues can enhance language model performance in ways previously underestimated.
Current MLLMs still struggle to connect the dots between images and text when they're interleaved, highlighting a critical gap in real-world multimodal understanding.
