Research map / Graph learning / GNN architectures
How hard is to distinguish graphs with graph neural networks?
Andreas Loukas
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 communication capacity | MPNN graph-isomorphism classifiers · 12 graph and tree isomorphism tasks | test accuracy | Improved |
| Increase graph size from 6 to 12 nodes | MPNN graph-isomorphism classifiers · X_graph^6, X_graph^8, X_graph^10, X_graph^12 | capacity required for 99% accuracy: 4, 8, 24, and 112 | Mixed |
| Use sufficient communication capacity | MPNN graph-isomorphism classifiers · Graph and tree isomorphism tasks | classification accuracy | Improved |
| Keep capacity below the theoretical lower-bound region | MPNN graph-isomorphism classifiers · Graph and tree isomorphism tasks | test accuracy | Hurt |
| Increase depth and width while keeping capacity comparable (ablation) | MPNN graph-isomorphism classifiers · Graph and tree isomorphism tasks | test accuracy | No clear effect |
| Make the network deep enough for its receptive field to cover the graph (ablation) | MPNN graph-isomorphism classifiers · Graph and tree isomorphism tasks | test accuracy | Improved |
| Use a large MPNN with one-hot node-order features | GIN0 graph-isomorphism classifier · X_tree^22 test | accuracy: 82.92% | Improved |
Closest papers by meaning
- Fine-grained Expressivity of Graph Neural Networks (NeurIPS 2023 · similarity 0.89)
- Let’s Agree to Degree: Comparing Graph Convolutional Networks in the Message-Passing Framework (ICML 2021 · similarity 0.89)
- A New Perspective on "How Graph Neural Networks Go Beyond Weisfeiler-Lehman?" (ICLR 2022 · similarity 0.89)
- Covered Forest: Fine-grained generalization analysis of graph neural networks (ICML 2025 · similarity 0.87)
- What do Graph Neural Networks learn? Insights from Tropical Geometry (NeurIPS 2024 · similarity 0.87)
- On the Expressivity and Sample Complexity of Node-Individualized Graph Neural Networks (NeurIPS 2024 · 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.
