Research map / Graph learning / GNN architectures
Going beyond persistent homology using persistent homology
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 |
|---|---|---|---|
| Replace vertex-color PH with RePHINE diagrams | GCN · NCI1 | accuracy: 79.08 -> 80.44 | Helped |
| Replace vertex-color PH with RePHINE diagrams | GCN · IMDB-B | accuracy: 64.80 -> 69.40 | Helped |
| Replace vertex-color PH with RePHINE diagrams | GCN · ZINC | MAE (lower better): 0.513 -> 0.468 | Helped |
| Replace vertex-color PH with RePHINE diagrams | GIN · IMDB-B | accuracy: 69.80 -> 72.80 | Helped |
| Replace vertex-color PH with RePHINE diagrams | GIN · PROTEINS | accuracy: 69.46 -> 72.32 | Helped |
| Replace vertex-color PH with RePHINE diagrams | GIN · MOLHIV | AUC: 73.37 -> 73.71 | No clear effect |
| Add vertex-color PH diagrams to the GNN | GCN · MOLHIV | AUC: 74.99 -> 73.64 | Hurt |
| Add vertex-color PH diagrams to the GNN | GIN · PROTEINS | accuracy: 72.50 -> 69.46 | Hurt |
| Add vertex-color PH diagrams to the GNN | GIN · ZINC | MAE (lower better): 0.621 -> 0.440 | Helped |
| Add RePHINE diagrams to the GNN | GIN · IMDB-B | accuracy: 74.20 -> 72.80 | Hurt |
| Add RePHINE diagrams to the GNN | GCN · ZINC | MAE (lower better): 0.875 -> 0.468 | Helped |
| Use RePHINE+Linear instead of PersLay | PersLay · NCI1 | accuracy: 68.86 -> 74.94 | Helped |
Closest papers by meaning
- Graphcode: Learning from multiparameter persistent homology using graph neural networks (NeurIPS 2024 · similarity 0.92)
- Graph Persistence goes Spectral (NeurIPS 2025 · similarity 0.91)
- Boosting Graph Pooling with Persistent Homology (NeurIPS 2024 · similarity 0.90)
- Positional Encoding meets Persistent Homology on Graphs (ICML 2025 · similarity 0.90)
- On topological descriptors for graph products (NeurIPS 2025 · similarity 0.90)
- TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration (ICML 2025 · similarity 0.89)
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.