Research map / Architectures and efficiency / Neural architecture search
NAS-Bench-101: Towards Reproducible 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 |
|---|---|---|---|
| replace one 3×3 convolution with a 1×1 convolution | NAS-Bench-101 cells · CIFAR-10 validation | absolute validation accuracy: 1.16% | Hurt |
| replace one 3×3 convolution with 3×3 max-pooling | NAS-Bench-101 cells · CIFAR-10 validation | absolute validation accuracy: 1.99% | Hurt |
| replace one 3×3 convolution with a 1×1 convolution | NAS-Bench-101 cells · CIFAR-10 training | training time: 14.11% | Helped |
| replace one 3×3 convolution with 3×3 max-pooling | NAS-Bench-101 cells · CIFAR-10 training | training time: 9.84% | Helped |
| use 3×3 max-pooling instead of 1×1 convolution | NAS-Bench-101 cells · CIFAR-10 training | training time: +5.04% | Hurt |
| use 3×3 max-pooling instead of 1×1 convolution | NAS-Bench-101 cells · CIFAR-10 validation | absolute validation accuracy: 0.81% | Hurt |
| increase cell depth up to 3 | NAS-Bench-101 cells · CIFAR-10 validation | mean validation accuracy | Helped |
| increase cell width up to 5 | NAS-Bench-101 cells · CIFAR-10 validation | mean validation accuracy | Helped |
| use regularized evolution instead of random search | NAS-Bench-101 architecture search · CIFAR-10 test regret over search time | test regret: about 5 times faster | Helped |
| use BOHB instead of random search | NAS-Bench-101 architecture search · CIFAR-10 test regret over search time | test regret: about 5 times faster | Helped |
| use SMAC instead of random search | NAS-Bench-101 architecture search · CIFAR-10 test regret over search time | test regret: about 5 times faster | Helped |
| use TPE instead of random search | NAS-Bench-101 architecture search · CIFAR-10 test regret | test regret | No clear effect |
Closest papers by meaning
- NAS-Bench-201: Extending the Scope of Reproducible Neural Architecture Search (ICLR 2020 · similarity 0.95)
- NAS-Bench-360: Benchmarking Neural Architecture Search on Diverse Tasks (NeurIPS 2022 · similarity 0.94)
- Insights from the Use of Previously Unseen Neural Architecture Search Datasets (CVPR 2024 · similarity 0.92)
- einspace: Searching for Neural Architectures from Fundamental Operations (NeurIPS 2024 · similarity 0.92)
- Efficient Few-Shot Neural Architecture Search by Counting the Number of Nonlinear Functions (AAAI 2025 · similarity 0.90)
- Towards Neural Architecture Search through Hierarchical Generative Modeling (ICML 2024 · 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.