Search papers, labs, and topics across Lattice.
1
0
3
ActionCache can accelerate inference for VLA models by up to 34.43脳 without compromising task success rates, revolutionizing real-time robotic manipulation.