Research map / Graph learning / GNN architectures
Non-Local Graph Neural Networks
Meng Liu, Zhengyang Wang, Shuiwang Ji
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 |
|---|---|---|---|
| add non-local aggregation via attention-guided sorting to MLP | MLP · Actor | accuracy (%): 35.1 -> 37.9 | Improved |
| add non-local aggregation to MLP | MLP · Cornell | accuracy (%): 81.6 -> 84.9 | Improved |
| add non-local aggregation to MLP | MLP · Texas | accuracy (%): 81.3 -> 85.4 | Improved |
| add non-local aggregation to GCN | GCN · snap-patents | accuracy (%): 45.7 -> 50.8 | Improved |
| add non-local aggregation to GCN | GCN · Squirrel | accuracy (%): 54.9 -> 59.0 | Improved |
| add non-local aggregation to GCN | GCN · Chameleon | accuracy (%): 67.6 -> 70.1 | Improved |
| add non-local aggregation to GAT | GAT · Squirrel | accuracy (%): 51.3 -> 56.8 | Improved |
| add non-local aggregation to GAT | GAT · Chameleon | accuracy (%): 65.0 -> 65.7 | Improved |
| add non-local aggregation to GCN on assortative graph | GCN · Cora | accuracy (%): 88.2 -> 88.1 | No clear effect |
| add local aggregation (GCN instead of MLP) on disassortative graph | MLP · Cornell | accuracy (%): 83.1 -> 58.5 | Hurt |
| add local aggregation (GCN instead of MLP) on Texas | MLP · Texas | accuracy (%): 81.5 -> 68.1 | Hurt |
| add non-local aggregation to GCN on running time | GCN · Chameleon | accuracy (%): 22.2 -> 26.3 | Hurt |
Closest papers by meaning
- A New Perspective on "How Graph Neural Networks Go Beyond Weisfeiler-Lehman?" (ICLR 2022 · similarity 0.89)
- EdgeNets: Edge Varying Graph Neural Networks (TPAMI 2021 · similarity 0.88)
- Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods (NeurIPS 2021 · similarity 0.88)
- Graph Neural Networks Are More Than Filters: Revisiting and Benchmarking from A Spectral Perspective (ICLR 2025 · similarity 0.87)
- Improving Graph Neural Networks by Learning Continuous Edge Directions (ICLR 2025 · similarity 0.87)
- Integrating Co-Training with Edge Discrimination to Enhance Graph Neural Networks Under Heterophily (AAAI 2025 · similarity 0.87)
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.
