Search papers, labs, and topics across Lattice.
3
0
5
TokenMinds reveals that combining discrete SID-based user tokens with dense embeddings can significantly enhance user modeling in recommender systems at scale.
Humanoids can now traverse complex real-world terrain, including handrails and cluttered environments, with zero-shot transfer from simulation, thanks to a novel depth-conditioned locomotion framework.
Legged robots can now nimbly walk backwards on challenging terrain using only a forward-facing camera, thanks to a clever memory system that recalls the path they already traversed.
