Search papers, labs, and topics across Lattice.
1
0
2
Current multimodal LLMs struggle with UI-based reasoning, but the new UI-UX model achieves a remarkable 0.7963 accuracy on the UXBench benchmark, setting a new standard.
