AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Learning Contrastive Embedding in Low-Dimensional Space

Shuo Chen, Chen Gong, Jun Li, Jian Yang, Gang Niu, Masashi Sugiyama

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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
add nuclear-norm low-rank reconstruction (256-dim) vs no regularizer (ablation)SimCLR · STL-10 400 epochsaccuracy (%): 74.1 -> 79.2Improved
add nuclear-norm low-rank reconstruction (256-dim) vs no regularizer (ablation)SimCLR · CIFAR-10 400 epochsaccuracy (%): 88.4 -> 93.1Improved
add l2,1-norm regularizer to 512-dim embedding (ablation)SimCLR · CIFAR-10 400 epochsaccuracy (%): 90.1 -> 92.5Improved
increase embedding dimension from 512 to 4096 without regularizer (ablation)SimCLR · CIFAR-10 400 epochsaccuracy (%): 90.3 -> 85.4Hurt
reduce embedding dimension naively from 2048 to 256 without regularizer (ablation)SimCLR · STL-10 400 epochsaccuracy (%): 76.2 -> 74.2No clear effect
low-rank regularizer at early training (100 epochs) (ablation)CLLR · STL-10 100 epochsaccuracy (%): 56.4 -> 56.3No clear effect
low-rank regularizer at early training (100 epochs) (ablation)CLLR · CIFAR-10 100 epochsaccuracy (%): 66.4 -> 67.5No clear effect
add CLLR (nuclear-norm) to DCLDCL · BookCorpus MPQAaccuracy (%): 93.7 -> 95.8Improved
add CLLR (nuclear-norm) to DCLDCL · BookCorpus TRECaccuracy (%): 89.1 -> 91.2Improved
add CLLR to CMCCMC · ImageNet-100 neg. 4096Top-1 accuracy (%): 73.58 -> 76.91Improved
add CLLR (nuclear-norm) to HCLHCL · ImageNet-100 neg. 1024Top-1 accuracy (%): 60.89 -> 62.43Improved
add CLLR (nuclear-norm) to CMC at small negative countCMC · ImageNet-100 neg. 1024Top-1 accuracy (%): 60.23 -> 61.23Improved

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