Search papers, labs, and topics across Lattice.
Affiliation:
2
0
5
TailSieve achieves up to 2.59x speedup in LLM rollouts by intelligently routing long-tail requests, transforming how we handle high-concurrency decoding.
DiT-Pruning achieves unprecedented image quality retention in Diffusion Transformers, maintaining a CLIP score loss of only 0.001 at 50% sparsity.
