Attribute Ranking on ARTIFACTBENCH (transductive)
0.57Kendall's TauGATv2Conv
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GATv2ConvMethod Category=GNN-based, Joint Training=Yes2026.05 | 0.57 | 0.655 | 56.4 | 86.2 | |
| NeoGNNMethod Category=GNN-based, Joint Training=Yes2026.05 | 0.533 | 0.634 | 52.6 | 85 | |
| NCNMethod Category=GNN-based, Joint Training=Yes2026.05 | 0.533 | 0.619 | 55.1 | 86.6 | |
| BUDDYMethod Category=GNN-based, Joint Training=Yes2026.05 | 0.516 | 0.611 | 60.9 | 86.8 | |
| NCNCMethod Category=GNN-based, Joint Training=Yes2026.05 | 0.492 | 0.592 | 48.7 | 81.9 | |
| GPT-5.2Method Category=LLM-based, Graph Info=Yes2026.05 | 0.464 | 0.518 | 59.1 | 94.5 | |
| GPT-5.2Method Category=LLM-based, Graph Info=No2026.05 | 0.367 | 0.413 | 51.9 | 92.3 | |
| Qwen3-8BMethod Category=LLM-based, Graph Info=Yes2026.05 | 0.207 | 0.246 | 45.2 | 89.8 | |
| Model meanMethod Category=Heuristic2026.05 | 0.206 | 0.233 | 39.9 | 88.6 | |
| Qwen3-8BMethod Category=LLM-based, Graph Info=No2026.05 | 0.174 | 0.193 | 40.4 | 87.3 | |
| Global meanMethod Category=Heuristic2026.05 | 0 | 0 | 2.9 | 80.3 | |
| Dataset meanMethod Category=Heuristic2026.05 | 0 | 0 | 2.9 | 80.3 |