AI Paper Atlas

Research map / Label-efficient and robust learning / Noisy labels

The Curious Case of Benign Memorization

Sotiris Anagnostidis, Gregor Bachmann, Lorenzo Noci, Thomas Hofmann

ICLR 2023 · Label-efficient and robust learning › Noisy labels · approach: Loss correction and noise robustness 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 standard data augmentation during training on random labelsResNet18 · CIFAR10K-NN probing accuracy (%): 12.4 -> 76.2Improved
Add standard data augmentation during training on random labelsVGG11 · CIFAR10K-NN probing accuracy (%): 15.0 -> 71.9Improved
Add standard data augmentation during training on random labelsResNet18 · CIFAR100K-NN probing accuracy (%): 5.0 -> 42.7Improved
Add standard data augmentation during training on random labelsVGG11 · CIFAR100K-NN probing accuracy (%): 6.0 -> 46.4Improved
Add standard data augmentation during training on random labelsResNet18 · TinyImageNetK-NN probing accuracy (%): 2.4 -> 33.4Improved
Add standard data augmentation during training on random labelsVGG11 · TinyImageNetK-NN probing accuracy (%): 1.2 -> 29.4Improved
Use independent same-class samples as augmentations instead of correlated standard augmentations (ablation)ResNet18 trained on random labels · CIFAR10K-NN probing accuracyHurt
Increase the number of label-preserving augmentations (ablation)ResNet18 trained on random labels · CIFAR10K-NN probing accuracyImproved
Increase augmentation strength (ablation)ResNet18 trained on random labels · Not specifiedK-NN probing accuracyImproved
Increase the number of random classes (ablation)ResNet18 trained on random labels · Not specifiedK-NN probing accuracyImproved
Increase projector hidden-layer size (ablation)ResNet18 trained on random labels · Not specifiedK-NN probing accuracyImproved

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