AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

The Power of Contrast for Feature Learning: A Theoretical Analysis

Wenlong Ji, Zhun Deng, Ryumei Nakada, James Zou, Linjun Zhang

JMLR 2023 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Inductive bias and minimal representations · 27 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
use linear contrastive loss instead of InfoNCEResNet-18 contrastive learning · CIFAR-10test accuracy: 65.11 -> 66.07Improved
use linear contrastive loss instead of InfoNCEResNet-18 contrastive learning · STL-10test accuracy: 71.02 -> 70.30Hurt
add quadratic regularization instead of weight decay on fully connected layers (ablation)SimCLR ResNet-18 · CIFAR-10test accuracy: 65.11 -> 65.54Improved
add quadratic regularization instead of weight decay on fully connected layers (ablation)SimCLR ResNet-18 · STL-10test accuracy: 71.02 -> 71.39Improved

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