Research map / Graph learning / Graph self-supervised learning
NetGAN: Generating Graphs via Random Walks
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 |
|---|---|---|---|
| Increase generated random walks from 500K to 100M (ablation) | NetGAN · Cora-ML, Cora, Citeseer, DBLP, Pubmed, PolBlogs link prediction | AUC: 94.00 -> 95.19; 82.31 -> 84.82; 95.18 -> 96.30; 82.45 -> 86.61; 87.39 -> 93.41; 95.06 -> 95.51 | Helped |
| Increase generated random walks from 500K to 100M (ablation) | NetGAN · Cora-ML, Cora, Citeseer, DBLP, Pubmed, PolBlogs link prediction | AP: 92.32 -> 95.24; 68.47 -> 88.04; 91.93 -> 96.89; 70.28 -> 89.21; 76.55 -> 94.59; 94.61 -> 94.83 | Helped |
| Use longer random walks rather than only edges (T=2) (ablation) | NetGAN · Cora-ML link prediction | link prediction performance | Helped |
| Increase walk length from T=16 to T=20 (ablation) | NetGAN · Cora-ML link prediction | link prediction performance | No clear effect |
| Use edge-overlap stopping instead of validation-criterion stopping (ablation) | NetGAN · Cora-ML graph generation | average rank: 3.00 -> 1.75 | Helped |
| Use validation-criterion stopping instead of edge-overlap stopping (ablation) | NetGAN · Citeseer graph generation | edge overlap: 0.76 -> 0.42 | Hurt |
Closest papers by meaning
- Geometric Generative Modeling with Noise-Conditioned Graph Networks (ICML 2025 · similarity 0.87)
- Revisiting Random Walks for Learning on Graphs (ICLR 2025 · similarity 0.86)
- Generative Graph Pattern Machine (NeurIPS 2025 · similarity 0.86)
- Efficient and Scalable Graph Generation through Iterative Local Expansion (ICLR 2024 · similarity 0.86)
- Learning Long Range Dependencies on Graphs via Random Walks (ICLR 2025 · similarity 0.86)
- Towards Addressing Frontiers in Graph Generation (AAAI 2025 · similarity 0.86)
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.