Lattice AI Research

Research focus

Architecture Design (Transformers, SSMs, MoE) (1)Scaling Laws & Emergent Abilities (1)Computer Vision (1)Distributed Systems & Hardware (1)

Frequent co-authors

Ziyan Gan (1)Fangxin Liu (1)Chenyang Guan (1)Ning Yang (1)

Papers (2)

Sep 1, 2026

Affiliation:2w ago

PCoMoE: Shifting MoE Inference from Monolithic Expert Selection to Fine-Grained Path Composition

Shifting from monolithic expert selection to fine-grained path composition, PCoMoE boosts inference speed by 1.31x while enhancing accuracy by 10%.

Ziyan Gan, Fangxin Liu, Chenyang Guan +8

Architecture Design (Transformers, SSMs, MoE)Scaling Laws & Emergent Abilities

Apr 2, 2026

Apr 2, 2026·also Affiliation:

GEMM-GS: Accelerating 3D Gaussian Splatting on Tensor Cores with GEMM-Compatible Blending

Unlock the full potential of your GPU's Tensor Cores for 3D Gaussian Splatting with a GEMM-friendly blending transformation that delivers up to 2x speedups.

Haomin Li, Bowen Zhu, Fangxin Liu +4

Computer Vision Distributed Systems & Hardware Inference & Quantization

Zongwu Wang

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)

Search

Zongwu Wang

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)