Graph Anomaly Detection on Weibo
98.8AUROCSAGAD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SAGAD2026.03 | 98.8 | — | |
| XGBGraph2026.03 | 96.4 | — | |
| SAWGAD2026.05 | 96.2 | 93.6 | |
| GATSep2026.05 | 94.7 | 92.6 | |
| SpaceGNN2026.03 | 94.4 | — | |
| NK-GAD2026.04 | 93.7 | — | |
| BWGNN2026.03 | 93.6 | — | |
| GCN2026.03 | 93.5 | — | |
| GraphSAGE2026.05 | 93.5 | 91.2 | |
| EvoFGCategory=Generalist method, Setting=Zero-shot2026.02 | 93.05 | 69.8 | |
| GCN2026.05 | 93 | 89.8 | |
| REFI-GADTraining setup=Models Trained on Group 2, k=502026.05 | 92.66 | — | |
| REFI-GADTraining Protocol=Training on Group 22026.05 | 92.66 | — | |
| ACM2026.03 | 92.5 | — | |
| ADA-GAD2026.04 | 92.44 | — | |
| AnomalyGFMCategory=Generalist method, Setting=Zero-shot2026.02 | 92.09 | 59.62 | |
| ProMoSTraining/Inference Setting=Unsupervised - Pre-Train Only2026.05 | 91.74 | — | |
| IA-GGADTraining setup=Models Trained on Group 22026.05 | 91.7 | — | |
| IA-GGADTraining Protocol=Training on Group 22026.05 | 91.7 | — | |
| GHRN2026.03 | 91.6 | — | |
| GIN2026.05 | 91.3 | 89.4 | |
| GAD-MoRELearning Paradigm=Generalist, Evaluation Setting=Zero-shot cross-domain2026.02 | 91.03 | — | |
| GAD-MoREShot count=0-shot2026.02 | 91.03 | — | |
| AnomalyDAE2026.04 | 89.88 | — | |
| GAAN2026.04 | 89.82 | — | |
| ARCCategory=Generalist method, Setting=Zero-shot2026.02 | 89.76 | 69.15 | |
| DOMINANT2026.04 | 89.45 | — | |
| DCI2026.03 | 89.3 | — | |
| ARCTraining/Inference Setting=Supervised - Pre-Train & Few-Shot Inference, Few-Shot=10-shot2026.05 | 88.85 | — | |
| ARCShot count=10-shot2026.02 | 88.65 | — | |
| H2GCN2026.03 | 87.8 | — | |
| IA-GGADLearning Paradigm=Generalist, Evaluation Setting=Zero-shot cross-domain2026.02 | 87.35 | — | |
| GAT2026.05 | 86.9 | 85.6 | |
| GAT2026.03 | 86.4 | — | |
| CAREGNN2026.05 | 86.1 | 85.3 | |
| SmoothGNNTraining setup=Models Trained on Group 22026.05 | 85.96 | — | |
| SmoothGNNTraining Protocol=Training on Group 22026.05 | 85.96 | — | |
| ARCTraining setup=Models Trained on Group 22026.05 | 85.54 | — | |
| ARCTraining Protocol=Training on Group 22026.05 | 85.54 | — | |
| ConsisGAD2026.03 | 85 | — | |
| DONE2026.04 | 84.2 | — | |
| PCGNN2026.03 | 83.9 | — | |
| GIN2026.03 | 83.8 | — | |
| AdONE2026.04 | 83.5 | — | |
| FAGCN2026.03 | 83.4 | — | |
| AMNet2026.03 | 82.4 | — | |
| BWGNN2026.05 | 82.1 | 81 | |
| SAGE2026.03 | 81.8 | — | |
| GAS2026.03 | 81.8 | — | |
| BWGNNTraining setup=Models Trained on Group 22026.05 | 81.62 | — | |
| BWGNNTraining Protocol=Training on Group 22026.05 | 81.62 | — | |
| H2FD2026.05 | 81.3 | 78.9 | |
| BernNet2026.03 | 80.1 | — | |
| GHRN2026.05 | 79.5 | 77.8 | |
| BGNNTraining setup=Models Trained on Group 22026.05 | 77.47 | — | |
| BGNNTraining Protocol=Training on Group 22026.05 | 77.47 | — | |
| GAD-NR2026.04 | 73.81 | — | |
| GATTraining setup=Models Trained on Group 22026.05 | 73.51 | — | |
| GATTraining Protocol=Training on Group 22026.05 | 73.51 | — | |
| GINCategory=Supervised method, Setting=Zero-shot2026.02 | 73.31 | 20.44 | |
| ConsisGAD2026.05 | 73 | 64.3 | |
| MLP2026.05 | 71.8 | 73.8 | |
| GHRNCategory=Supervised method, Setting=Zero-shot2026.02 | 71.52 | 21.78 | |
| GGADTraining setup=Models Trained on Group 22026.05 | 70.23 | — | |
| GGADTraining Protocol=Training on Group 22026.05 | 70.23 | — | |
| AdaGNN2026.03 | 69.5 | — | |
| PCGNN2026.05 | 69 | 65.2 | |
| CGNN2026.05 | 66.5 | 52.6 | |
| UNPromptLearning Paradigm=Generalist, Evaluation Setting=Zero-shot cross-domain2026.02 | 66.29 | — | |
| BWGNNCategory=Supervised method, Setting=Zero-shot2026.02 | 63.24 | 13.61 | |
| AHFANTraining setup=Models Trained on Group 22026.05 | 63.03 | — | |
| AHFANTraining Protocol=Training on Group 22026.05 | 63.03 | — | |
| SAGECategory=Supervised method, Setting=Zero-shot2026.02 | 61.06 | 36.2 | |
| AMNetCategory=Supervised method, Setting=Zero-shot2026.02 | 59.76 | 13.28 | |
| CHRNTraining setup=Models Trained on Group 22026.05 | 57.77 | — | |
| CHRNTraining Protocol=Training on Group 22026.05 | 57.77 | — | |
| SmoothGNN2026.04 | 56.97 | — | |
| AnomalyGFMTraining/Inference Setting=Supervised - Pre-Train & Few-Shot Inference, Few-Shot=10-shot2026.05 | 56.91 | — | |
| UNPromptCategory=Generalist method, Setting=Zero-shot2026.02 | 53.88 | 26.89 | |
| BWGNNLearning Paradigm=Supervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 50.08 | — | |
| AnomalyGFMShot count=10-shot2026.02 | 49.68 | — | |
| UNPromptTraining setup=Models Trained on Group 22026.05 | 47.21 | — | |
| UNPromptTraining Protocol=Training on Group 22026.05 | 47.21 | — | |
| GCNTraining setup=Models Trained on Group 22026.05 | 46.4 | — | |
| GCNTraining Protocol=Training on Group 22026.05 | 46.4 | — | |
| CoLALearning Paradigm=Unsupervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 45.25 | — | |
| AnomalyGFMLearning Paradigm=Generalist, Evaluation Setting=Zero-shot cross-domain2026.02 | 44.43 | — | |
| AnomalyGFMTraining setup=Models Trained on Group 22026.05 | 43.16 | — | |
| AnomalyGFMTraining Protocol=Training on Group 22026.05 | 43.16 | — | |
| CoLATraining setup=Models Trained on Group 22026.05 | 41.08 | — | |
| CoLATraining Protocol=Training on Group 22026.05 | 41.08 | — | |
| GCNLearning Paradigm=Supervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 37.64 | — | |
| ANEMONETraining setup=Models Trained on Group 22026.05 | 35.57 | — | |
| ANEMONETraining Protocol=Training on Group 22026.05 | 35.57 | — | |
| GATLearning Paradigm=Supervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 35.5 | — | |
| GADAM2026.04 | 25.54 | — | |
| TAMLearning Paradigm=Unsupervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 21.5 | — | |
| AnomalyDAELearning Paradigm=Unsupervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 19.77 | — | |
| GHRNLearning Paradigm=Supervised, Evaluation Setting=Zero-shot cross-domain2026.02 | 17.79 | — | |
| ACM2026.03 | — | 66 |