Search papers, labs, and topics across Lattice.
5
0
10
4
SciDSK transforms how AI agents interact with scientific datasets, enabling more effective discovery and interpretation through a structured, reusable skill representation.
SFF-CLIP boosts fine-grained feature representation in CLIP without the need for cumbersome region annotations, achieving significant performance gains.
Local Branch Routing enables language models to leverage contextual evidence for decision-making without the computational burden of full solution searches, leading to substantial improvements in reasoning accuracy.
Sparse prefilling can dramatically accelerate long-context inference in diffusion language models, achieving up to 28x speedup without sacrificing quality.
Forget coarse sequence-level hacks: LenVM lets you precisely dial in token generation length, boosting a 7B model's length accuracy from 30.9 to 64.8 and crushing closed-source rivals.
