Research map / Graph learning / GNN architectures
Geometric Knowledge Distillation: Topology Compression for Graph Neural Networks
Chenxiao Yang, Qitian Wu, Junchi Yan
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 |
|---|---|---|---|
| Align teacher and student neural heat kernels instead of using vanilla KD | 3-layer GCN · OGBN-Arxiv edge-aware | accuracy: 68.63 -> 70.90 | Improved |
| Use parametric neural heat-kernel alignment instead of vanilla KD | 3-layer GCN · OGBN-Arxiv edge-aware | accuracy: 68.63 -> 71.38 | Improved |
| Align teacher and student neural heat kernels instead of using vanilla KD | 3-layer GCN · OGBN-Arxiv node-aware | accuracy: 68.86 -> 71.31 | Improved |
| Use parametric neural heat-kernel alignment instead of vanilla KD | 3-layer GCN · OGBN-Arxiv node-aware | accuracy: 68.86 -> 71.27 | Improved |
| Apply Gaussian neural heat-kernel distillation rather than using the student alone | Student GNN · Cora edge-aware offline distillation | accuracy: 83.84 -> 87.68 | Improved |
| Apply sigmoid neural heat-kernel distillation rather than using the student alone | Student GNN · Cora edge-aware offline distillation | accuracy: 83.84 -> 88.01 | Improved |
| Apply randomized neural heat-kernel distillation rather than using the student alone | Student GNN · Cora edge-aware offline distillation | accuracy: 83.84 -> 88.48 | Improved |
| Apply parametric neural heat-kernel distillation rather than using the student alone | Student GNN · Cora edge-aware offline distillation | accuracy: 83.84 -> 88.41 | Improved |
Closest papers by meaning
- Geometry Awakening: Cross-Geometry Learning Exhibits Superiority over Individual Structures (NeurIPS 2024 · similarity 0.89)
- Toward Data-centric Directed Graph Learning: An Entropy-driven Approach (ICML 2025 · similarity 0.89)
- Topological Blindspots: Understanding and Extending Topological Deep Learning Through the Lens of Expressivity (ICLR 2025 · similarity 0.88)
- Do Topological Characteristics Help in Knowledge Distillation? (ICML 2024 · similarity 0.88)
- VQGraph: Rethinking Graph Representation Space for Bridging GNNs and MLPs (ICLR 2024 · similarity 0.88)
- Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs (NeurIPS 2026 · similarity 0.88)
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.
