Research map / Graph learning / Graph self-supervised learning
Ultrahyperbolic Neural Networks
Marc T Law
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 |
|---|---|---|---|
| ultrahyperbolic P^{1,3}_1 instead of Euclidean R^4 (ablation) | GCN · Cora node classification d=4 | accuracy: 53.5 -> 63.2 | Improved |
| ultrahyperbolic P^{1,3}_1 instead of Euclidean R^4 (ablation) | GCN · Citeseer node classification d=4 | accuracy: 44.5 -> 51.8 | Improved |
| ultrahyperbolic P^{1,3}_1 instead of Euclidean R^4 (ablation) | GCN · Pubmed node classification d=4 | accuracy: 66.9 -> 73.1 | Improved |
| ultrahyperbolic P^{1,3}_1 instead of hyperbolic P^{0,4}_1 (ablation) | GCN · Cora node classification d=4 | accuracy: 56.2 -> 63.2 | Improved |
| elliptic P^{4,0}_1 instead of ultrahyperbolic P^{1,3}_1 (ablation) | GCN · Citeseer node classification d=4 | accuracy: 51.8 -> 47.2 | Hurt |
| ultrahyperbolic instead of Euclidean (ablation) | GCN · D&D graph classification | accuracy: 76.93 -> 81.97 | Improved |
| ultrahyperbolic instead of Lorentz hyperbolic (ablation) | GCN · Enzymes graph classification | accuracy: 44.83 -> 50.50 | Improved |
| ultrahyperbolic instead of Euclidean (ablation) | GCN · Collab graph classification | accuracy: 81.88 -> 82.26 | No clear effect |
| ultrahyperbolic instead of Euclidean (ablation) | GCN · Reddit-multi-12K graph classification | accuracy: 45.65 -> 47.08 | No clear effect |
| ultrahyperbolic instead of Euclidean (ablation) | GCN · Proteins graph classification | accuracy: 75.46 -> 76.56 | No clear effect |
| increase layer dimensionality to d=600 (ablation) | Euclidean, hyperbolic and ultrahyperbolic GCNs · node classification | test accuracy | No clear effect |
| ultrahyperbolic P^{1,3}_1 instead of Euclidean R^4 (ablation) | MLP on karate club · rank of first leader | rank: 4.6 -> 1.2 | Improved |
Closest papers by meaning
- Ultrahyperbolic Representation Learning (NeurIPS 2020 · similarity 0.93)
- Siegel Neural Networks (NeurIPS 2025 · similarity 0.87)
- Neural networks on Symmetric Spaces of Noncompact Type (ICLR 2025 · similarity 0.87)
- Weighted Embeddings for Low-Dimensional Graph Representation (AAAI 2025 · similarity 0.86)
- Hyperbolic-PDE GNN: Spectral Graph Neural Networks in the Perspective of A System of Hyperbolic Partial Differential Equations (ICML 2025 · similarity 0.86)
- Hyperbolic Embeddings of Supervised Models (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.
