Search papers, labs, and topics across Lattice.
2
0
6
0
Omnidirectional imagery + language unlocks robust multi-object tracking that overcomes the field-of-view limitations plaguing conventional video datasets.
Forget full attention: a hybrid sparse-linear attention model, MiniCPM-SALA, achieves 3.5x faster inference and supports 1M context length on a single GPU, all while maintaining comparable performance.
