Research map / Architectures and efficiency / Neural architecture search
Understanding Architectures Learnt by Cell-based Neural Architecture Search
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 |
|---|---|---|---|
| Adapt NASNet cells to the widest, shallowest topology | NASNet · CIFAR-10, CIFAR-100, Tiny-ImageNet-200 | test error (%): 2.65 -> 2.80; 17.06 -> 16.86; 31.88 -> 32.05 | Mixed |
| Adapt AmoebaNet cells to the widest, shallowest topology | AmoebaNet · CIFAR-10, CIFAR-100, Tiny-ImageNet-200 | test error (%): 2.76 -> 2.91; 17.55 -> 17.28; 32.22 -> 33.16 | Mixed |
| Adapt ENAS cells to the widest, shallowest topology | ENAS · CIFAR-10, CIFAR-100, Tiny-ImageNet-200 | test error (%): 2.64 -> 2.76; 16.67 -> 16.04; 30.68 -> 31.36 | Mixed |
| Adapt DARTS cells to the widest, shallowest topology | DARTS · CIFAR-10, CIFAR-100, Tiny-ImageNet-200 | test error (%): 2.67 -> 2.73; 16.41 -> 16.15; 30.58 -> 31.33 | Mixed |
| Adapt SNAS cells to the widest, shallowest topology | SNAS · CIFAR-10, CIFAR-100, Tiny-ImageNet-200 | test error (%): 2.88 -> 2.69; 17.78 -> 17.20; 32.40 -> 32.61 | Mixed |
| Increase cell width and decrease cell depth (ablation) | DARTS and random-connection variants · CIFAR-10 and CIFAR-100 | convergence rate and stability | Helped |
| Replace the operations while preserving NAS cell connections (ablation) | DARTS, ENAS, AmoebaNet, NASNet · CIFAR-10 | convergence | No clear effect |
| Use wider and shallower cells rather than narrower and deeper cells (ablation) | DARTS and random-connection variants · CIFAR-10 and CIFAR-100 | test accuracy at convergence | Mixed |
Closest papers by meaning
- On Redundancy and Diversity in Cell-based Neural Architecture Search (ICLR 2022 · similarity 0.93)
- einspace: Searching for Neural Architectures from Fundamental Operations (NeurIPS 2024 · similarity 0.92)
- Towards Neural Architecture Search through Hierarchical Generative Modeling (ICML 2024 · similarity 0.92)
- EG-NAS: Neural Architecture Search with Fast Evolutionary Exploration (AAAI 2024 · similarity 0.92)
- Insights from the Use of Previously Unseen Neural Architecture Search Datasets (CVPR 2024 · similarity 0.91)
- Efficient Few-Shot Neural Architecture Search by Counting the Number of Nonlinear Functions (AAAI 2025 · similarity 0.90)
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.