Search papers, labs, and topics across Lattice.
2
0
6
9
Qualitative examples show the PhysBrain 1.5 model's ability to produce end-effector trajectories and predict future scenes through spatially aligned RGB, depth, and robot-mask outputs.
Train LLMs with RL on a budget: SparrowRL achieves near-RDMA throughput using commodity networks by exploiting the sparsity of RL updates, slashing transfer payloads by 79x.
