Research map / Architectures and efficiency / Neural architecture search
Topology-Weighted Effective Rank: A Zero-Cost Proxy for Training Dynamics Stability in Neural Architecture Search
Haojie Zhang, Yeming Yang, Songbai Liu, Lijia Ma, Ka-Chun Wong, Qiuzhen Lin
Read the paperCheck 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 |
|---|---|---|---|
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 C100 (AutoFormer space) | Spearman correlation (%): 63.87 -> 95.36 | Improved |
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 Flowers (AutoFormer space) | Spearman correlation (%): 83.65 -> 89.17 | Improved |
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 Chaoyang (AutoFormer space) | Spearman correlation (%): 41.76 -> 53.58 | Improved |
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 C100 (PiT space) | Spearman correlation (%): 95.12 -> 95.90 | Improved |
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 Flowers (PiT space) | Spearman correlation (%): 92.94 -> 96.33 | Improved |
| integrate ER-Score into the Auto-Prox search framework | Auto-Prox · ViT-Bench-101 Chaoyang (PiT space) | Spearman correlation (%): 55.09 -> 76.14 | Improved |
| add topology weighting (PageRank) to ER-Score, giving TER-Score (ablation) | ER-Score · NAS-Bench-301 elite architecture nDCG | nDCG | Improved |
| add degree or PageRank topology weighting to ER-Score (ablation) | ER-Score · NAS-Bench-201/301, TransNAS-Bench-101 | Spearman correlation | Improved |
| use Gaussian or uniform random inputs instead of real images for the mini-batch (ablation) | ER-Score and TER-Score · Mini-batch distribution ablation | Spearman correlation | Improved |
| use all-one input instead of random input (ablation) | ER-Score and TER-Score · Mini-batch distribution ablation | Spearman correlation | Hurt |
| use medium input resolution instead of low or high resolution (ablation) | ER-Score and TER-Score · Mini-batch resolution ablation | Spearman correlation | Improved |
| use TER-Score as the guidance signal in evolutionary search instead of other ZCPs | Evolutionary search · NAS-Bench-301 surrogate (DARTS space) | predicted accuracy | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- AZ-NAS — AZ-NAS: Assembling Zero-Cost Proxies for Network Architecture Search (CVPR 2024) · better on 8 of 8 shared results
- Auto-Prox — Auto-Prox: Training-Free Vision Transformer Architecture Search via Automatic Proxy Discovery (AAAI 2024) · better on 6 of 6 shared results
- SWAP-NAS — SWAP-NAS: Sample-Wise Activation Patterns for Ultra-fast NAS (ICLR 2024) · better on 2 of 2 shared results
- DARTS — DARTS-: Robustly Stepping out of Performance Collapse Without Indicators (ICLR 2021) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EZNAS: Evolving Zero-Cost Proxies For Neural Architecture Scoring (NeurIPS 2022 · similarity 0.89)
- Per-Architecture Training-Free Metric Optimization for Neural Architecture Search (NeurIPS 2025 · similarity 0.88)
- Understanding and Accelerating Neural Architecture Search With Training-Free and Theory-Grounded Metrics (TPAMI 2023 · similarity 0.88)
- Zero-Cost Operation Scoring in Differentiable Architecture Search (AAAI 2023 · similarity 0.88)
- Robustifying and Boosting Training-Free Neural Architecture Search (ICLR 2024 · similarity 0.88)
- Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approach (CVPR 2024 · similarity 0.87)
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.
