Research map / Graph learning / GNN architectures
Learning to Pre-train Graph Neural Networks
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 L2P-GNN pre-training instead of the strongest baseline | GCN · Biology | ROC-AUC: 64.72±1.06 -> 66.48±1.59 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GraphSAGE · Biology | ROC-AUC: 67.39±1.54 -> 69.89±1.63 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GAT · Biology | ROC-AUC: 68.37±0.54 -> 69.15±1.86 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GIN · Biology | ROC-AUC: 65.99±1.22 -> 70.13±0.95 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GCN · PreDBLP | Micro-F1: 66.11±0.16 -> 66.58±0.28 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GraphSAGE · PreDBLP | Micro-F1: 63.60±0.21 -> 65.84±0.37 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GAT · PreDBLP | Micro-F1: 61.30±2.17 -> 62.24±1.89 | Helped |
| Use L2P-GNN pre-training instead of the strongest baseline | GIN · PreDBLP | Micro-F1: 69.37±0.21 -> 70.79±0.17 | Helped |
| Add graph-level adaptation to node-level-only L2P-GNN (ablation) | L2P-GNN · Biology and PreDBLP | ROC-AUC and Micro-F1 | Helped |
| Add node-level adaptation to graph-level-only L2P-GNN (ablation) | L2P-GNN · Biology and PreDBLP | ROC-AUC and Micro-F1 | Helped |
| Vary node- and graph-level adaptation steps from 0 to 3 (ablation) | L2P-GNN with GIN · Biology and PreDBLP | ROC-AUC and Micro-F1 | Mixed |
| Use one gradient update for each adaptation (ablation) | L2P-GNN with GIN · Biology and PreDBLP | ROC-AUC and Micro-F1 | Helped |
Closest papers by meaning
- Edge Prompt Tuning for Graph Neural Networks (ICLR 2025 · similarity 0.91)
- Generative Graph Pattern Machine (NeurIPS 2025 · similarity 0.90)
- How Graph Neural Networks Learn: Lessons from Training Dynamics (ICML 2024 · similarity 0.90)
- Towards Pre-trained Graph Condensation via Optimal Transport (NeurIPS 2025 · similarity 0.90)
- Diffusing to the Top: Boost Graph Neural Networks with Minimal Hyperparameter Tuning (ICLR 2025 · similarity 0.90)
- GraphTOP: Graph Topology-Oriented Prompting for Graph Neural Networks (NeurIPS 2025 · 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.