Research map / Graph learning / GNN architectures
Graphon Neural Networks and the Transferability of 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 |
|---|---|---|---|
| increase training subgraph size n from 271 to 1354 before transfer | 1-layer GNN · Cora node classification | relative accuracy difference (%): 23.1% -> 15.4% | Helped |
| increase training subnetwork size n from 100 to 900 users | 1-layer GNN · MovieLens 100k | relative test RMSE difference | Helped |
| increase number of layers L (ablation) | GNN · Consensus on SBM graphs | transfer rRMSE difference variability | Hurt |
| increase number of features F from 4 to 8 or 16 (ablation) | GNN · Consensus on SBM graphs | transfer rRMSE difference variability | Hurt |
| increase number of filter taps K (ablation) | GNN · Consensus on SBM graphs | transfer rRMSE difference | No clear effect |
| increase training graph size n from 50 to 1000 | GNN · Consensus on SBM graphs | relative rRMSE difference | Helped |
Closest papers by meaning
- Graph neural networks and non-commuting operators (NeurIPS 2024 · similarity 0.92)
- Size Transferability of Graph Convolutional Networks across Sparsity: A Generalized Graphon Perspective (ICML 2026 · similarity 0.91)
- A Manifold Perspective on the Statistical Generalization of Graph Neural Networks (ICML 2025 · similarity 0.91)
- Higher-Order Graphon Neural Networks: Approximation and Cut Distance (ICLR 2025 · similarity 0.90)
- Almost Surely Asymptotically Constant Graph Neural Networks (NeurIPS 2024 · similarity 0.88)
- Graph Neural Networks Are More Than Filters: Revisiting and Benchmarking from A Spectral Perspective (ICLR 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.