Search papers, labs, and topics across Lattice.
1
0
3
Even state-of-the-art MLLMs fail to grasp intuitive visual reasoning, scoring less than 60% on tasks humans solve effortlessly at a glance.
