Research map / Graph learning / GNN architectures
Invariant and Equivariant Graph 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 |
|---|---|---|---|
| Use the maximal permutation-equivariant basis instead of the Hartford et al. basis | Equivariant graph-function networks · Symmetric projection | MSE: 2.09 -> 1E-05 | Helped |
| Use the maximal permutation-equivariant basis instead of the Hartford et al. basis | Equivariant graph-function networks · Diagonal extraction | MSE: 0.81 -> 8E-06 | Helped |
| Use the maximal permutation-equivariant basis instead of the Hartford et al. basis | Equivariant graph-function networks · Max singular vector | cosine loss: 0.043 -> 0.015 | Helped |
| Use the maximal permutation-invariant basis instead of the Hartford et al. basis | Invariant graph-function networks · Trace | MSE: 316.22 -> 0.005 | Helped |
| Use the proposed graph network for graph classification | Graph classification model · Social datasets (COLLAB, IMDB-B, IMDB-M) | accuracy | Mixed |
| Use the proposed graph network for graph classification | Graph classification model · Bioinformatics datasets (MUTAG, PTC, PROTEINS, NCI1, NCI109) | accuracy | Hurt |
Closest papers by meaning
- A Characterization Theorem for Equivariant Networks with Point-wise Activations (ICLR 2024 · similarity 0.91)
- REVISITING MULTI-PERMUTATION EQUIVARIANCE THROUGH THE LENS OF IRREDUCIBLE REPRESENTATIONS (ICLR 2025 · similarity 0.91)
- Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks? (NeurIPS 2024 · similarity 0.90)
- Universally Invariant Learning in Equivariant GNNs (NeurIPS 2025 · similarity 0.90)
- Schur Nets: exploiting local structure for equivariance in higher order graph neural networks (NeurIPS 2024 · similarity 0.90)
- On the Generalization of Equivariant Graph Neural Networks (ICML 2024 · 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.