Uncertainty Quantification on Uniform Synthetic
10.49SharpnessCF-GNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CF-GNN2026.05 | 10.49 | 3.07 | |
| BayesianNN2026.05 | 9.02 | 3 | |
| QpiGNNlambda=0.1, Backbone=GraphSAGE2026.05 | 0.79 | 0.89 | |
| RQRadj.-GNNBackbone=GraphSAGE2026.05 | 0.46 | 0.68 | |
| QpiGNNlambda=optimal, Backbone=GraphSAGE2026.05 | 0.39 | 0.62 | |
| SQR-GNNBackbone=GraphSAGE2026.05 | 0.27 | 0.52 | |
| QpiGNNlambda=0.5, Backbone=GraphSAGE2026.05 | 0.19 | 0.44 | |
| MC dropout2026.05 | 0.07 | 0.36 |