Search papers, labs, and topics across Lattice.
3
0
4
Adversaries can hijack robotic policies in real-time, achieving complete control over their actions with a 100% success rate across multiple tasks.
AetheRock not only streamlines data collection for robot learning but also pairs it with a novel framework that enhances tactile learning in the face of sensor inconsistencies.
Over-eagerly adapting VLMs for robot control can actually hurt performance, suggesting the original VLM representation already encodes surprisingly useful action priors.
