Research map / Graph learning / GNN architectures
MagNet: A Neural Network for Directed Graphs
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 magnetic Laplacian to encode directed edge information rather than the classical graph Laplacian | MagNet versus ChebNet and GCN · Cornell, Texas, Wisconsin, and Telegram node classification | accuracy: +4 percentage point | Helped |
| Tune q to zero, treating directional information as noise | MagNet · Cora-ML and CiteSeer node classification | accuracy | No clear effect |
| Use nonzero q to incorporate directional information | MagNet · Link prediction, including Cora-ML and CiteSeer | accuracy | Helped |
| Use a single nonzero q for the combined three-class link prediction task | MagNet · Three-class link prediction | accuracy | Mixed |
Closest papers by meaning
- Improving Graph Neural Networks by Learning Continuous Edge Directions (ICLR 2025 · similarity 0.88)
- What Are Good Positional Encodings for Directed Graphs? (ICLR 2025 · similarity 0.88)
- Toward Data-centric Directed Graph Learning: An Entropy-driven Approach (ICML 2025 · similarity 0.86)
- Enhancing Spectral GNNs: From Topology and Perturbation Perspectives (ICML 2025 · similarity 0.86)
- AutoSGNN: Automatic Propagation Mechanism Discovery for Spectral Graph Neural Networks (AAAI 2025 · similarity 0.86)
- Beyond Message Passing: Neural Graph Pattern Machine (ICML 2025 · 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.