Research map / Architectures and efficiency / Neural architecture search
ReNAS: Relativistic Evaluation of Neural Architecture Search
Check 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 pairwise ranking loss instead of MSE (ablation) | ReNAS predictor with type-matrix features · NAS-Bench-101, 0.1% predictor training samples | Kendall's Tau: 0.3465 -> 0.6039 | Helped |
| Use pairwise ranking loss instead of MSE (ablation) | ReNAS predictor with feature-tensor inputs · NAS-Bench-101, 0.1% predictor training samples | Kendall's Tau: 0.4856 -> 0.6335 | Helped |
| Add FLOPs and parameter information in the feature tensor (ablation) | ReNAS predictor with MSE loss · NAS-Bench-101, 0.1% predictor training samples | Kendall's Tau: 0.3465 -> 0.4856 | Helped |
| Add continuity loss to pairwise ranking loss (ablation) | ReNAS predictor with type-matrix inputs · NAS-Bench-101, 0.1% predictor training samples | Kendall's Tau: 0.6039 -> 0.6096 | Helped |
| Add continuity loss to pairwise ranking loss (ablation) | ReNAS predictor with feature-tensor inputs · NAS-Bench-101, 0.1% predictor training samples | Kendall's Tau: 0.6335 -> 0.6574 | Helped |
| Use random rather than parameter-count-based sample selection | ReNAS predictor training · NAS-Bench-101, CIFAR-10 test | accuracy (%): 93.84 ± 0.21 -> 93.95 ± 0.11 | Helped |
| Use random rather than FLOPs-based sample selection | ReNAS predictor training · NAS-Bench-101, CIFAR-10 test | accuracy (%): 93.76 ± 0.13 -> 93.95 ± 0.11 | Helped |
| Use ReNAS predictor-guided search instead of REA | NAS-Bench-201 search · CIFAR-10 test | accuracy (%): 93.92±0.30 -> 93.99±0.25 | Helped |
Closest papers by meaning
- Speedy Performance Estimation for Neural Architecture Search (NeurIPS 2021 · similarity 0.91)
- NAS-Bench-201: Extending the Scope of Reproducible Neural Architecture Search (ICLR 2020 · similarity 0.90)
- NAS-Bench-360: Benchmarking Neural Architecture Search on Diverse Tasks (NeurIPS 2022 · similarity 0.90)
- How Powerful are Performance Predictors in Neural Architecture Search? (NeurIPS 2021 · similarity 0.90)
- ISTA-NAS: Efficient and Consistent Neural Architecture Search by Sparse Coding (NeurIPS 2020 · similarity 0.90)
- Contrastive Neural Architecture Search With Neural Architecture Comparators (CVPR 2021 · 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.