Research map / Learning theory and algorithms / Combinatorial and search optimization
Slimmable Dataset Condensation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use first-order and infinity-order parameter matching for successive condensation | Slimmable dataset condensation · FashionMNIST, CIFAR10, and CIFAR100 | accuracy | Helped |
| Add first-order parameter matching (ablation) | Slimmable dataset condensation · CIFAR10 | accuracy | Helped |
| Add infinity-order parameter matching instead of performance matching (ablation) | Slimmable dataset condensation · CIFAR10 | accuracy | Helped |
| Add skewness loss (ablation) | Significance-aware learning-free slimming · CIFAR10 | accuracy | Helped |
| Add orthogonality loss (ablation) | Significance-aware learning-free slimming · CIFAR10 | accuracy | Helped |
| Use significance-aware parameterization (ablation) | Slimmable dataset condensation · CIFAR10 | accuracy | Helped |
| Use distribution matching rather than performance matching for successive condensation | DM · FashionMNIST, CIFAR10, and CIFAR100 | performance gap to retraining | Mixed |
Closest papers by meaning
- Elucidating the Design Space of Dataset Condensation (NeurIPS 2024 · similarity 0.90)
- Diversity-Driven Synthesis: Enhancing Dataset Distillation through Directed Weight Adjustment (NeurIPS 2024 · similarity 0.88)
- Large Scale Dataset Distillation with Domain Shift (ICML 2024 · similarity 0.88)
- Embarrassingly Simple Dataset Distillation (ICLR 2024 · similarity 0.87)
- CondTSF: One-line Plugin of Dataset Condensation for Time Series Forecasting (NeurIPS 2024 · similarity 0.87)
- Multisize Dataset Condensation (ICLR 2024 · similarity 0.87)
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.