Research map / Graph learning / Graph self-supervised learning
Does Graph Distillation See Like Vision Dataset Counterpart?
Beining Yang, Kai Wang, Qingyun Sun, Cheng Ji, Xingcheng Fu, Hao Tang, Yang You, Jianxin Li
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 |
|---|---|---|---|
| broadcast original graph structure into condensed structure generation (SGDD vs GCond) | GCond · YelpChi anomaly detection | F1-macro: +9.6% | Improved |
| same, on Amazon | GCond · Amazon anomaly detection | F1-macro: +7.7% | Improved |
| SGDD instead of GCond at 0.90% ratio | GCond · Citeseer node classification | accuracy: 70.5 -> 69.5 | Hurt |
| SGDD instead of GCond at 0.25% ratio | GCond · Ogbn-arxiv node classification | accuracy: 63.2 -> 65.8 | Improved |
| remove original adjacency condition A from SGDD (ablation) | SGDD · Ogbn-arxiv and YelpChi | accuracy / F1 | Hurt |
| use both X' and A conditions vs GCond (ablation) | SGDD · Ogbn-arxiv | accuracy: +3.1% | Improved |
| use both X' and A conditions vs GCond (ablation) | SGDD · YelpChi | F1-macro: +8.0% | Improved |
| cross-architecture: SGDD vs GCond on Reddit | SGDD · Reddit cross-architecture | average accuracy: 76.3 -> 80.9 | Improved |
| cross-architecture: SGDD vs GCond on Cora | SGDD · Cora cross-architecture | average accuracy: 75.9 -> 78.3 | Improved |
| cross-architecture: SGDD vs GCond on YelpChi | SGDD · YelpChi cross-architecture | average accuracy: 45.3 -> 56.6 | Improved |
| SGDD vs GCond NAS on YelpChi | SGDD · NAS YelpChi | test accuracy: 58.4 -> 60.6 | Improved |
| set structure weight alpha to nonzero (ablation) | SGDD · YelpChi, Cora, Ogbn-arxiv | performance | Improved |
Closest papers by meaning
- Graph Distillation with Eigenbasis Matching (ICML 2024 · similarity 0.87)
- VQGraph: Rethinking Graph Representation Space for Bridging GNNs and MLPs (ICLR 2024 · similarity 0.87)
- Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs (NeurIPS 2026 · similarity 0.87)
- StructComp: Substituting propagation with Structural Compression in Training Graph Contrastive Learning (ICLR 2024 · similarity 0.86)
- Graph Structure Refinement with Energy-based Contrastive Learning (AAAI 2025 · similarity 0.86)
- Efficient Graph Continual Learning via Lightweight Graph Neural Tangent Kernels-based Dataset Distillation (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.
