Research map / Architectures and efficiency / Pruning and efficient tuning
Dataset Pruning: Reducing Training Data by Examining Generalization Influence
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 |
|---|---|---|---|
| Optimization-based dataset pruning (group-effect constrained selection) instead of score-based selection for building NAS proxy set | 100 images/class CIFAR10 proxy set · CIFAR10 NAS over 720 ConvNets | Performance (%) of selected architecture on whole dataset: 85.7 vs 82.5 (forgetting), 80.1 (herding), 79.4 (random) | Helped |
| Optimization-based dataset pruning proxy set instead of random proxy set | NAS proxy-set ranking · CIFAR10 NAS over 720 ConvNets | Spearman correlation with whole-dataset ranking: 0.94 vs 0.21 | Helped |
| Proxy-set search via pruned dataset instead of full dataset | NAS architecture search · CIFAR10 NAS over 720 ConvNets | Search time (min): 113 vs 3029 minutes | Helped |
Closest papers by meaning
- Lightweight Dataset Pruning without Full Training via Example Difficulty and Prediction Uncertainty (ICML 2025 · similarity 0.90)
- $\mathbb{D}^2$ Pruning: Message Passing for Balancing Diversity & Difficulty in Data Pruning (ICLR 2024 · similarity 0.88)
- GDeR: Safeguarding Efficiency, Balancing, and Robustness via Prototypical Graph Pruning (NeurIPS 2024 · similarity 0.88)
- DRoP: Distributionally Robust Data Pruning (ICLR 2025 · similarity 0.88)
- Exploring Learning Complexity for Efficient Downstream Dataset Pruning (ICLR 2025 · similarity 0.87)
- Repeated Random Sampling for Minimizing the Time-to-Accuracy of Learning (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.