Research map / Graph learning / GNN architectures
Graph Representational Learning: When Does More Expressivity Hurt Generalization?
Sohir Maskey, Raffaele Paolino, Fabian Jogl, Gitta Kutyniok, Johannes F. Lutzeyer
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 |
|---|---|---|---|
| add cycle-count features up to length 4 (F4-MPNN) to the MPNN (ablation) | MPNN · Task 1 cycle-count labels | test accuracy: 0.8490 -> 0.9793 | Improved |
| use more expressive Local Folklore 2-GNN instead of F4-MPNN (ablation) | F4-MPNN · Task 1 cycle-count labels | test accuracy: 0.9793 -> 0.8450 | Hurt |
| extend cycle features from length 4 to length 7 (ablation) | F4-MPNN · Task 1 cycle-count labels | test accuracy: 0.9793 -> 0.9660 | Hurt |
| use fewer cycle features (length 3 only) than task requires (ablation) | F4-MPNN · Task 1 cycle-count labels | test accuracy: 0.9793 -> 0.8657 | Hurt |
| increase distance of test graphs from the training set (zeta-TMD) (ablation) | GIN on real-world datasets · BZR, Mutagenicity, NCI109 | accuracy | Hurt |
| destroy structure-label correlation via random label resampling (ablation) | MPNN · BZR, Mutagenicity, NCI109 | test error | Hurt |
Closest papers by meaning
- Towards Bridging Generalization and Expressivity of Graph Neural Networks (ICLR 2025 · similarity 0.96)
- Beyond Weisfeiler-Lehman: A Quantitative Framework for GNN Expressiveness (ICLR 2024 · similarity 0.93)
- On the Rademacher Complexity of Graph Neural Networks: Unifying Expressivity and Geometry (NeurIPS 2026 · similarity 0.92)
- Generalization, Expressivity, and Universality of Graph Neural Networks on Attributed Graphs (ICLR 2025 · similarity 0.91)
- Covered Forest: Fine-grained generalization analysis of graph neural networks (ICML 2025 · similarity 0.90)
- An Empirical Study of Realized GNN Expressiveness (ICML 2024 · 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.
