Research map / Graph learning / GNN architectures
Fine-grained Expressivity 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 |
|---|---|---|---|
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · Mutag | accuracy: 79.01 ± 2.24 -> 82.56 ± 3.12 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · IMDB-Binary | accuracy: 69.96 ± 1.43 -> 70.70 ± 0.60 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · IMDB-Multi | accuracy: 46.29 ± 0.76 -> 47.59 ± 0.95 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · NCI1 | accuracy: 78.61 ± 0.34 -> 77.82 ± 0.55 | Hurt |
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · Proteins | accuracy: 73.51 ± 0.47 -> 73.45 ± 0.30 | Hurt |
| Freeze randomly initialized input and hidden weights; train only the output layer | GIN-m · Reddit-Binary | accuracy: 89.73 ± 0.37 -> 82.32 ± 0.45 | Hurt |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · Mutag | accuracy: 81.62 ± 2.08 -> 78.03 ± 1.57 | Hurt |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · IMDB-Binary | accuracy: 59.14 ± 1.93 -> 65.77 ± 1.32 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · IMDB-Multi | accuracy: 38.75 ± 1.62 -> 43.29 ± 0.96 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · NCI1 | accuracy: 63.28 ± 0.6 -> 62.36 ± 0.45 | Hurt |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · Proteins | accuracy: 71.49 ± 0.67 -> 71.83 ± 0.42 | Helped |
| Freeze randomly initialized input and hidden weights; train only the output layer | GraphConv-m · Reddit-Binary | accuracy: 82.4 ± 0.19 -> 77.15 ± 0.29 | Hurt |
Closest papers by meaning
- Homomorphism Expressivity of Spectral Invariant Graph Neural Networks (ICLR 2025 · similarity 0.92)
- Beyond Weisfeiler-Lehman: A Quantitative Framework for GNN Expressiveness (ICLR 2024 · similarity 0.91)
- Weisfeiler-Leman at the margin: When more expressivity matters (ICML 2024 · similarity 0.91)
- Covered Forest: Fine-grained generalization analysis of graph neural networks (ICML 2025 · similarity 0.91)
- Higher-Order Graphon Neural Networks: Approximation and Cut Distance (ICLR 2025 · similarity 0.90)
- Generalization, Expressivity, and Universality of Graph Neural Networks on Attributed Graphs (ICLR 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.