Search papers, labs, and topics across Lattice.
1
0
3
By disentangling structure and motion in the latent space, CoWVLA achieves superior visuomotor learning compared to standard world-model and latent-action approaches.
