Research map / Graph learning / GNN architectures
AdS-GNN - a Conformally Equivariant Graph Neural Network
Maksim Zhdanov, Nabil Iqbal, Erik J Bekkers, Patrick Forré
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 |
|---|---|---|---|
| replace Euclidean distance with AdS proper distance in EGNN messages (conformal equivariance) | EGNN · PascalVOC-SP (LRGB) | test F1: 27.80 -> 28.07 | No clear effect |
| use AdS-GNN instead of EGNN on SuperPixel MNIST with rotated and scaled test data | EGNN · SuperPixel MNIST (augmented test) | test error | Improved |
| use AdS-GNN instead of EGNN on shape segmentation, in distribution | EGNN · Shape segmentation | accuracy | Improved |
| use AdS-GNN instead of MPNN (translation-only) on shape segmentation | MPNN · Shape segmentation | accuracy | Hurt |
| use AdS-GNN instead of EGNN for Ising N-point correlation prediction | EGNN · 2d Ising N-point functions | relative L2 | Improved |
| reduce graph connectivity k_con from fully connected to 2 (ablation) | AdS-GNN and EGNN · Ising generalization across N | generalization and in-distribution performance | Mixed |
| compare AdS-GNN to P-Theta-NITA with orientation information on SuperPixel MNIST in-distribution | AdS-GNN · SuperPixel MNIST | classification error | Hurt |
Closest papers by meaning
- Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks? (NeurIPS 2024 · similarity 0.88)
- Lie Neurons: Adjoint-Equivariant Neural Networks for Semisimple Lie Algebras (ICML 2024 · similarity 0.87)
- A Characterization Theorem for Equivariant Networks with Point-wise Activations (ICLR 2024 · similarity 0.86)
- Lie Group Decompositions for Equivariant Neural Networks (ICLR 2024 · similarity 0.86)
- Universally Invariant Learning in Equivariant GNNs (NeurIPS 2025 · similarity 0.85)
- Weisfeiler Leman for Euclidean Equivariant Machine Learning (ICML 2024 · similarity 0.85)
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.
