AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Dataset Distillation via Factorization

Songhua Liu, Kai Wang, Xingyi Yang, Jingwen Ye, Xinchao Wang

NeurIPS 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Self and data-free distillation · 58 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
Factorize distilled data into shared hallucinators and basesMTT dataset distillation · SVHN, CIFAR10, CIFAR100test accuracyImproved
Add hallucinator-basis factorizationDC, DM, and MTT · CIFAR10, ConvNet; 2, 11, and 51 IPCtest accuracy: +2.75, +4.59, +4.67; +2.75, +6.48, +2.41; +6.17, +5.58, +3.44Improved
Add hallucinator-basis factorizationDD distilled datasets · Cross-architecture CIFAR10test accuracy: +1.40 to +17.57Improved
Remove the cosine-similarity loss L_cos. (ablation)HaBa · CIFAR10test accuracyHurt
Remove the contrastive loss L_con. (ablation)HaBa · CIFAR10test accuracyHurt
Remove the task-specific loss L_task (ablation)HaBa · CIFAR10test accuracyHurt
Use L_con. during downstream training (ablation)HaBa · CIFAR10test accuracyImproved
Use class-independent hallucinators instead of shared hallucinators (ablation)HaBa · CIFAR10, low BPCtest accuracyImproved
Use class-independent hallucinators instead of shared hallucinators (ablation)HaBa · CIFAR10, higher BPCtest accuracyHurt
Use single-channel bases and increase BPC to use the same storage (ablation)HaBa · CIFAR10, small BPCtest accuracyImproved
Increase hallucinator count above 10 (ablation)HaBa · CIFAR10, 10 or 50 BPCtest accuracyNo clear effect
Remove DSA data augmentation (ablation)HaBa · CIFAR10, 10 BPCtest accuracyHurt

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