Search papers, labs, and topics across Lattice.
1
0
3
Models trained on the new VideoKR dataset achieve superior performance in knowledge-intensive video reasoning, setting a new benchmark for the field.
