Research map / Architectures and efficiency / Neural architecture search
AdvantageNAS: Efficient Neural Architecture Search with Credit Assignment
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 |
|---|---|---|---|
| Use per-edge advantage estimates instead of one scalar reward for architecture updates | REINFORCE-style sparse propagation NAS · CIFAR-10 | accuracy (argmax): 93.01 ± 0.47 -> 93.76 ± 0.00 | Helped |
| Use per-edge advantage estimates instead of one scalar reward for architecture updates | REINFORCE-style sparse propagation NAS · CIFAR-10 | accuracy (process best): 92.69 ± 0.52 -> 93.29 ± 0.21 | Helped |
| Use per-edge advantage estimates for architecture updates | AdvantageNAS · PTB | validation perplexity: 95.12 ± 0.22 -> 94.90 ± 0.18 | Helped |
| Use per-edge advantage estimates for architecture updates | AdvantageNAS · PTB | test perplexity: 90.76 ± 0.23 -> 91.18 ± 0.13 | Hurt |
| Include a zero operation in the search space (ablation) | AdvantageNAS · CIFAR-10 | accuracy (argmax) | No clear effect |
| Include a zero operation in the search space (ablation) | other SP-NAS methods · CIFAR-10 | accuracy | Hurt |
| Select the final architecture by argmax rather than process best (ablation) | AdvantageNAS · CIFAR-10 | accuracy: 93.29 ± 0.21 -> 93.76 ± 0.00 | Helped |
Closest papers by meaning
- ISTA-NAS: Efficient and Consistent Neural Architecture Search by Sparse Coding (NeurIPS 2020 · similarity 0.92)
- Unifying and Boosting Gradient-Based Training-Free Neural Architecture Search (NeurIPS 2022 · similarity 0.92)
- MiLeNAS: Efficient Neural Architecture Search via Mixed-Level Reformulation (CVPR 2020 · similarity 0.91)
- Efficient Few-Shot Neural Architecture Search by Counting the Number of Nonlinear Functions (AAAI 2025 · similarity 0.90)
- EG-NAS: Neural Architecture Search with Fast Evolutionary Exploration (AAAI 2024 · similarity 0.90)
- Encodings for Prediction-based Neural Architecture Search (ICML 2024 · 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.