Research map / Graph learning / Graph self-supervised learning
GOOD: A Graph Out-of-Distribution Benchmark
Shurui Gui, Xiner Li, Limei 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 |
|---|---|---|---|
| evaluate under OOD split instead of ID split | ERM · GOOD-HIV size covariate | ROC-AUC (%): 83.72 -> 59.94 | Hurt |
| evaluate under OOD split instead of ID split | ERM · GOOD-CMNIST color covariate | accuracy (%): 77.96 -> 28.60 | Hurt |
| use DIR instead of ERM on concept shift | DIR · GOOD-HIV size concept | ROC-AUC (%): 63.26 -> 72.61 | Improved |
| use DIR instead of ERM on covariate shift | DIR · GOOD-HIV size covariate | ROC-AUC (%): 59.94 -> 57.11 | Hurt |
| use DIR instead of ERM on regression | DIR · GOOD-ZINC scaffold covariate | MAE: 0.1995 -> 0.6493 | Hurt |
| use VREx instead of ERM | VREx · GOOD-Motif base covariate | accuracy (%): 68.66 -> 71.47 | Improved |
| use GroupDRO instead of ERM | GroupDRO · GOOD-ZINC scaffold covariate | MAE: 0.1995 -> 0.1934 | Improved |
| use Mixup instead of ERM on node tasks | Mixup · GOOD-Cora degree concept | accuracy (%): 60.54 -> 63.65 | Improved |
| use Mixup instead of ERM on graph tasks | Mixup · GOOD-HIV scaffold covariate | ROC-AUC (%): 69.58 -> 68.88 | No clear effect |
| use IRM instead of ERM | IRM · GOOD-HIV scaffold covariate | ROC-AUC (%): 69.58 -> 67.97 | No clear effect |
| use EERM instead of ERM on WebKB | EERM · GOOD-WebKB university covariate | accuracy (%): 14.29 -> 24.61 | Improved |
| use OOD validation set instead of ID validation set for model selection | ERM · GOOD splits | OOD test score | Improved |
Closest papers by meaning
- GOODAT: Towards Test-Time Graph Out-of-Distribution Detection (AAAI 2024 · similarity 0.90)
- No Metric to Rule Them All: Toward Principled Evaluations of Graph-Learning Datasets (ICML 2025 · similarity 0.90)
- ML-GOOD: Towards Multi-Label Graph Out-Of-Distribution Detection (AAAI 2025 · similarity 0.88)
- Bridging OOD Detection and Generalization: A Graph-Theoretic View (NeurIPS 2024 · similarity 0.88)
- Revisiting Score Propagation in Graph Out-of-Distribution Detection (NeurIPS 2024 · similarity 0.87)
- Graph Structure Extrapolation for Out-of-Distribution Generalization (ICML 2024 · 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.
