Search papers, labs, and topics across Lattice.
Xiaohongshu Inc.
3
0
7
Hypic slashes time-to-first-token by 2.45x and doubles throughput for hybrid-attention LLMs, all while preserving near-full accuracy.
LLMs can be sped up by over 2x without sacrificing accuracy, by compressing the input and predicting multiple output tokens at once using a unified framework.
Omni-modal RL post-training just got a whole lot faster: Relax delivers up to 2x speedups over existing systems, even for massive MoE models, without sacrificing reward convergence.
