Search papers, labs, and topics across Lattice.
2
0
6
0
Robot RL training can be dramatically sped up (3-10x) by decoupling CPU-based simulation from GPU-based learning, challenging the assumption that GPU-resident physics is essential for efficiency.
Surprisingly, high sparsity in video diffusion models doesn't degrade generation quality if the sparse mask accurately mimics the tile-wise geometry of full attention.
