Search papers, labs, and topics across Lattice.
4
0
7
Video generative priors only translate into robust physical control when paired with explicit world-to-action information routing and synchronized joint denoising rather than standard monolithic fine-tuning.
Sparse keypoints combined with dense correspondences enable robots to master complex rigid-deformable interactions with minimal demonstrations.
Ditching caches for compiler-managed data streams, Li Auto's M100 architecture achieves higher utilization than GPUs on autonomous driving tasks, hinting at a new path for efficient AI inference.
VLMs can achieve better multimodal reasoning simply by dynamically rescaling positional indices based on information density, without any training or architectural changes.
