AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Is Attention Better Than Matrix Decomposition?

Zhengyang Geng, Meng-Hao Guo, Hongxu Chen, Xia Li, Ke Wei, Zhouchen Lin

ICLR 2021 · Architectures and efficiency › Attention and transformers · approach: Sparse and structured attention · 51 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Replace self-attention with HamburgerHamNet · PASCAL VOC test, w/o COCO pretrainingmIoU(%): 84.4 -> 85.9Improved
Replace self-attention with HamburgerHamNet · PASCAL Context ValmIoU(%): 54.8 -> 55.2Improved
Replace self-attention with HamburgerHamGAN · ImageNet 128×128FID: 18.28 -> 16.05Improved
Replace YLG with HamburgerHamGAN-strong · ImageNet 128×128FID: 15.94 -> 14.77Improved
Add matrix decomposition ham (ablation)Hamburger with NMF · PASCAL VOC valmIoU(%): 75.8(75.6) -> 78.3(77.8)Improved
Add upper bread (ablation)Hamburger with NMF · PASCAL VOC valmIoU(%): 77.0(76.8) -> 78.3(77.8)Improved
Add lower bread (ablation)Hamburger with NMF · PASCAL VOC valmIoU(%): 77.3(77.2) -> 78.3(77.8)Improved
Add both linear bread transformations around ham (ablation)Hamburger with NMF · PASCAL VOC valmIoU(%): 77.0(76.8) -> 78.3(77.8)Improved
Use one-step gradient instead of BPTT (ablation)VQ ham · PASCAL VOC valmIoU: 76.6(76.3) -> 77.7(77.4)Improved
Use one-step gradient instead of BPTT (ablation)CD ham · PASCAL VOC valmIoU: 75.0(74.6) -> 78.1(77.5)Improved
Use one-step gradient instead of BPTT (ablation)NMF ham · PASCAL VOC valmIoU: 77.4(77.0) -> 78.3(77.8)Improved
Use random rather than learned dictionary initialization (ablation)NMF ham · PASCAL VOC valmIoU: 76.8(76.5) -> 78.3(77.8)Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy