Research map / Graph learning / Graph self-supervised learning
FEW-SHOT LEARNING ON GRAPHS VIA SUPER-CLASSES BASED ON GRAPH SPECTRAL MEASURES
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 spectral super-class and super-graph GAT framework instead of the strongest listed baseline | few-shot graph classification · TRIANGLES 20-shot | accuracy: 72.55 $\pm$ 1.97 -> 80.04 $\pm$ 2.20 | Helped |
| Use the spectral super-class and super-graph GAT framework instead of the strongest listed baseline | few-shot graph classification · ENZYMES 5-shot | accuracy: 57.24 $\pm$ 7.06 -> 55.42 $\pm$ 5.74 | Hurt |
| Replace the GAT classifier with GCN (ablation) | OurMethod-GAT · TRIANGLES 20-shot | accuracy: 80.04 $\pm$ 2.20 -> 77.86 $\pm$ 2.84 | Hurt |
| Replace the super-graph classifier with feature-space k-NN (ablation) | OurMethod-GAT · TRIANGLES 20-shot | accuracy: 80.04 $\pm$ 2.20 -> 63.45 $\pm$ 2.76 | Hurt |
| Add the super-class classifier to the method (ablation) | OurMethod-GAT · Letter-High 20-shot | accuracy: 75.23 $\pm$ 2.48 -> 77.42 $\pm$ 1.47 | Helped |
| Add the super-class classifier to the method (ablation) | OurMethod-GAT · TRIANGLES 20-shot | accuracy: 76.89 $\pm$ 2.63 -> 80.14 $\pm$ 1.88 | Helped |
| Add the super-class classifier to the method (ablation) | OurMethod-GAT · Reddit-12K 20-shot | accuracy: 48.19 $\pm$ 4.01 -> 50.36 $\pm$ 3.04 | Helped |
| Add the super-class classifier to the method (ablation) | OurMethod-GAT · ENZYMES 20-shot | accuracy: 62.7 $\pm$ 3.63 -> 62.39 $\pm$ 3.48 | No clear effect |
| Use 3 rather than 1 super-class (ablation) | OurMethod-GAT · TRIANGLES 20-shot | accuracy: 76.43 $\pm$ 2.87 -> 80.51 $\pm$ 1.72 | Helped |
| Use 3 rather than 1 super-class (ablation) | OurMethod-GAT · Letter-High 20-shot | accuracy: 74.43 $\pm$ 2.61 -> 77.51 $\pm$ 1.49 | Helped |
| Use 2 rather than 3 super-classes (ablation) | OurMethod-GAT · Reddit-12K 20-shot | accuracy: 50.10 $\pm$ 3.02 -> 50.67 $\pm$ 2.94 | Helped |
| Increase super-graph neighbors from 2 to 8 (ablation) | OurMethod-GAT · Letter-High 20-shot | accuracy: 77.33 $\pm$ 1.71 -> 74.66 $\pm$ 2.61 | Hurt |
Closest papers by meaning
- Unlocking the Potential of Black-box Pre-trained GNNs for Graph Few-shot Learning (AAAI 2025 · similarity 0.88)
- SpeAr: A Spectral Approach for Zero-Shot Node Classification (NeurIPS 2024 · similarity 0.88)
- Fast Graph Sharpness-Aware Minimization for Enhancing and Accelerating Few-Shot Node Classification (NeurIPS 2024 · similarity 0.88)
- Graph Few-Shot Learning via Adaptive Spectrum Experts and Cross-Set Distribution Calibration (NeurIPS 2025 · similarity 0.88)
- Normalize Then Propagate: Efficient Homophilous Regularization for Few-Shot Semi-Supervised Node Classification (AAAI 2025 · similarity 0.87)
- Spectral Graph Neural Networks are Incomplete on Graphs with a Simple Spectrum (NeurIPS 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.