Research map / Architectures and efficiency / Neural architecture search
AttentiveNAS: Improving Neural Architecture Search via Attentive Sampling
Dilin Wang, Meng Li, Chengyue Gong, Vikas Chandra
Check 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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use WorstUp-50 accuracy-predictor sampling instead of uniform sampling (ablation) | Two-stage NAS · ImageNet, 200±10 MFLOPs | Top-1 accuracy: around 0.3% | Improved |
| Focus training on the worst Pareto architectures rather than uniform sampling (ablation) | Two-stage NAS · ImageNet, low-FLOPs regimes | Minimum and first-quartile accuracy | Improved |
| Use WorstUp-1M accuracy-predictor sampling rather than BestUp-1M (ablation) | Two-stage NAS · ImageNet | Top-1 accuracy across FLOPs regimes | Improved |
| Use BestUp-50 accuracy-predictor sampling rather than uniform sampling (ablation) | Two-stage NAS · ImageNet, medium-FLOPs regimes above 400 MFLOPs | Maximum top-1 accuracy | Improved |
| Use BestUp-3 minibatch-loss sampling rather than uniform sampling (ablation) | Two-stage NAS · ImageNet | Maximum top-1 accuracy | Improved |
| Use WorstUp-3 minibatch-loss sampling rather than uniform sampling (ablation) | Two-stage NAS · ImageNet | Maximum top-1 accuracy | Improved |
| Balance exploration and exploitation with BestUp-3 minibatch-loss sampling rather than accuracy-predictor-based BestUp (ablation) | Two-stage NAS · ImageNet | Maximum top-1 accuracy | Improved |
| Increase the number of sampled candidate architectures from BestUp-50 to offline BestUp-1M (ablation) | BestUp accuracy-predictor sampling · ImageNet | Maximum top-1 accuracy | Mixed |
| Use accuracy predictors trained early to rank candidate subnetworks (ablation) | AttentiveNAS performance estimation · ImageNet subnetworks | Kendall's τ: 0.89; 0.87; 0.88 | Improved |
Closest papers by meaning
- BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage Models (ECCV 2020 · similarity 0.92)
- TransNAS-Bench-101: Improving Transferability and Generalizability of Cross-Task Neural Architecture Search (CVPR 2021 · similarity 0.90)
- Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective (ICLR 2021 · similarity 0.90)
- Understanding and Accelerating Neural Architecture Search With Training-Free and Theory-Grounded Metrics (TPAMI 2023 · similarity 0.90)
- GreedyNAS: Towards Fast One-Shot NAS With Greedy Supernet (CVPR 2020 · similarity 0.90)
- Efficient Few-Shot Neural Architecture Search by Counting the Number of Nonlinear Functions (AAAI 2025 · similarity 0.89)
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.
