Research map / Graph learning / GNN architectures
One-shot Graph Neural Architecture Search with Dynamic Search Space
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 a fixed search space with a dynamically updated search space (ablation) | DSS · Cora semi-supervised | accuracy: 83.4 ± 0.2% -> 83.9 ± 0.3% | Helped |
| replace a fixed search space with a dynamically updated search space (ablation) | DSS · Citeseer semi-supervised | accuracy: 71.8 ± 0.2% -> 72.2 ± 0.3% | Helped |
| replace a fixed search space with a dynamically updated search space (ablation) | DSS · Pubmed semi-supervised | accuracy: 78.9 ± 0.3% -> 80.3 ± 0.2% | Helped |
| replace a random one-iteration search space with a dynamically updated search space (ablation) | DSS · Cora semi-supervised | accuracy: 82.7 ± 0.9% -> 83.9 ± 0.3% | Helped |
| replace a random one-iteration search space with a dynamically updated search space (ablation) | DSS · Citeseer semi-supervised | accuracy: 71.2 ± 0.2% -> 72.2 ± 0.3% | Helped |
| replace a random one-iteration search space with a dynamically updated search space (ablation) | DSS · Pubmed semi-supervised | accuracy: 79.2 ± 0.2% -> 80.3 ± 0.2% | Helped |
| increase retained operations from K=1 to K=3 (M=10) (ablation) | DSS · Cora semi-supervised | accuracy: 82.6 ± 0.4% -> 83.9 ± 0.3% | Helped |
| increase retained operations from K=1 to K=3 (M=10) (ablation) | DSS · Cora full-supervised | accuracy: 90.8 ± 0.3% -> 91.0 ± 0.2% | No clear effect |
| increase subset size from M=8 to M=10 (K=3) (ablation) | DSS · Cora semi-supervised | accuracy: 82.8 ± 0.3% -> 83.9 ± 0.3% | Helped |
| increase subset size from M=8 to M=10 (K=3) (ablation) | DSS · Cora full-supervised | accuracy: 91.1 ± 0.1% -> 91.0 ± 0.2% | No clear effect |
| use DSS instead of GraphNAS | GNN architecture search · Cora semi-supervised | accuracy: 83.7 ± 0.4% -> 83.9 ± 0.3% | Helped |
| use DSS instead of GraphNAS | GNN architecture search · Cora full-supervised | accuracy: 90.6 ± 0.3% -> 91.0 ± 0.2% | Helped |
Closest papers by meaning
- Rethinking Graph Neural Architecture Search From Message-Passing (CVPR 2021 · similarity 0.92)
- Graph HyperNetworks for Neural Architecture Search (ICLR 2019 · similarity 0.90)
- NAS-Bench-Graph: Benchmarking Graph Neural Architecture Search (NeurIPS 2022 · similarity 0.90)
- Can Classic GNNs Be Strong Baselines for Graph-level Tasks? Simple Architectures Meet Excellence (ICML 2025 · similarity 0.88)
- How Graph Neural Networks Learn: Lessons from Training Dynamics (ICML 2024 · similarity 0.87)
- Diffusing to the Top: Boost Graph Neural Networks with Minimal Hyperparameter Tuning (ICLR 2025 · 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.