Search papers, labs, and topics across Lattice.
4
0
7
3
LEEVLA reveals that effectively guiding attention to task-critical evidence can dramatically enhance performance in vision-language-action tasks.
Binarization methods that ignore weight significance can lead to substantial performance losses, but SAB-LVLM optimizes this process, achieving superior efficiency without sacrificing accuracy.
ViGOS reveals that decoupling perception from reasoning can significantly enhance multimodal model robustness against shortcut learning.
SkillsCrafter enables robots to learn new manipulation skills without forgetting old ones by distilling knowledge into a shared semantic space, paving the way for more adaptable and robust robotic systems.
