Research map / Graph learning / GNN architectures
Geom-GCN: Geometric Graph Convolutional 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 Geom-GCN-I instead of GCN | Geom-GCN-I · Citeseer | accuracy: 73.68 -> 77.99 | Helped |
| Use Geom-GCN-P instead of GAT | Geom-GCN-P · Chameleon | accuracy: 42.93 -> 60.90 | Helped |
| Use Geom-GCN-P instead of GAT | Geom-GCN-P · Squirrel | accuracy: 30.03 -> 38.14 | Helped |
| Use Geom-GCN-P instead of GAT | Geom-GCN-P · Texas | accuracy: 58.38 -> 67.57 | Helped |
| Add the graph neighborhood to Geom-GCN-I (ablation) | Geom-GCN-I · Cora | accuracy: +0.48 | Helped |
| Add the latent-space neighborhood to Geom-GCN-I (ablation) | Geom-GCN-I · Cora | accuracy: ↓8.34 | Hurt |
| Add the graph neighborhood to Geom-GCN-P (ablation) | Geom-GCN-P · Chameleon | accuracy: +34.89 | Helped |
| Add the latent-space neighborhood to Geom-GCN-P (ablation) | Geom-GCN-P · Cornell | accuracy: +22.70 | Helped |
| Use graph-only neighborhoods rather than both neighborhoods (ablation) | Geom-GCN variants · Cora | accuracy: 86.26; 86.30; 87.00 | Mixed |
| Combine different embedding spaces for neighborhoods (ablation) | Geom-GCN variants · Cora | accuracy: 85.13; 85.09; 84.51; 85.31; 85.65; 85.43 | Mixed |
Closest papers by meaning
- Going Deep: Graph Convolutional Ladder-Shape Networks (AAAI 2020 · similarity 0.90)
- Learning Discrete Structures for Graph Neural Networks (ICML 2019 · similarity 0.90)
- L2-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks (CVPR 2020 · similarity 0.89)
- Principal Neighbourhood Aggregation for Graph Nets (NeurIPS 2020 · similarity 0.89)
- Deeper Insights Into Graph Convolutional Networks for Semi-Supervised Learning (AAAI 2018 · similarity 0.89)
- A Manifold Perspective on the Statistical Generalization of Graph Neural Networks (ICML 2025 · similarity 0.88)
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.