Graph Anomaly Detection on Facebook (test)
17.53AUPRCTA-GGAD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TA-GGADMethod Category=GGAD methods2026.03 | 17.53 | — | |
| GCTAMMethod Category=GAD methods2026.03 | 9.61 | — | |
| GAD-MoREMethod Category=Generalist, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 9 | — | |
| IA-GGADMethod Category=Generalist, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 8.65 | — | |
| TAMMethod Category=GAD methods2026.03 | 8.4 | — | |
| TAMParadigm=Unsupervised - Pre-Train Only, Setting=Zero-shot2026.05 | 8.4 | — | |
| ARCMethod Category=GGAD methods2026.03 | 8.38 | — | |
| ProMoSParadigm=Ours, Setting=Zero-shot2026.05 | 6.42 | — | |
| GATMethod Category=Supervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 6.18 | — | |
| AnomalyGFMParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 6.02 | — | |
| AnomalyGFMMethod Category=Generalist, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 6 | — | |
| BWGNNMethod Category=Supervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 5.71 | — | |
| AnomalyGFMMethod Category=GGAD methods2026.03 | 5.48 | — | |
| TAMMethod Category=Unsupervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 5.37 | — | |
| GCNMethod Category=Supervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 4.55 | — | |
| BGNNMethod Category=GAD methods2026.03 | 3.81 | — | |
| BGNNParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 3.81 | — | |
| SmoothGNNMethod Category=GAD methods2026.03 | 3.76 | — | |
| GHRNMethod Category=Supervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 3.61 | — | |
| UNPromptMethod Category=Generalist, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 3.44 | — | |
| CoLAMethod Category=Unsupervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 3.41 | — | |
| UNPromptParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 3.17 | — | |
| GATMethod Category=GAD methods2026.03 | 3.14 | — | |
| GATParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 3.14 | — | |
| DOMINANTMethod Category=GAD methods2026.03 | 2.95 | — | |
| DOMINANTParadigm=Unsupervised - Pre-Train Only, Setting=Zero-shot2026.05 | 2.95 | — | |
| CONSISGADMethod Category=GAD methods2026.03 | 2.72 | — | |
| CAGADMethod Category=GAD methods2026.03 | 2.61 | — | |
| UNPromptMethod Category=GGAD methods2026.03 | 2.61 | — | |
| BWGNNMethod Category=GAD methods2026.03 | 2.54 | — | |
| BWGNNParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 2.54 | — | |
| GHRNMethod Category=GAD methods2026.03 | 2.41 | — | |
| GHRNParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 2.41 | — | |
| HCM-AMethod Category=GAD methods2026.03 | 2.08 | — | |
| HCM-AParadigm=Unsupervised - Pre-Train Only, Setting=Zero-shot2026.05 | 2.08 | — | |
| CoLAMethod Category=GAD methods2026.03 | 1.9 | — | |
| CoLAParadigm=Unsupervised - Pre-Train Only, Setting=Zero-shot2026.05 | 1.9 | — | |
| S-GADMethod Category=GAD methods2026.03 | 1.82 | — | |
| SpaceGNNMethod Category=GAD methods2026.03 | 1.65 | — | |
| GCNMethod Category=GAD methods2026.03 | 1.59 | — | |
| GCNParadigm=Supervised - Pre-Train Only, Setting=Zero-shot2026.05 | 1.59 | — | |
| AnomalyDAEMethod Category=Unsupervised, Zero-shot cross-domain setting=true, Training Source Graphs=PubMed, Flickr, Reddit, YelpChi2026.02 | 1.32 | — | |
| ARCEvaluation Protocol=In-context Learning2024.05 | — | 0.6756 | |
| ARCEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.6728 | |
| ARCSupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.6503 | |
| BGNNLearning Supervision=Supervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.5474 | |
| BWGNNLearning Supervision=Supervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.4584 | |
| BWGNNEvaluation Protocol=Supervised, Shot-count=10-shot2026.01 | — | 0.6774 | |
| BWGNNSupervision Type=Supervised, Evaluation Setting=10-shot2026.01 | — | 0.6774 | |
| CAREEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.7994 | |
| CARESupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.7918 | |
| COLALearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.1299 | |
| COLALearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train & Fine-Tune2024.05 | — | 0.647 | |
| DOMINANTLearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.5101 | |
| DOMINANTLearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train & Fine-Tune2024.05 | — | 0.4992 | |
| DOMINANTEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.5122 | |
| DOMINANTSupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.5231 | |
| GATLearning Supervision=Supervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.5188 | |
| GCNLearning Supervision=Supervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.2951 | |
| GHRNLearning Supervision=Supervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.4481 | |
| GHRNEvaluation Protocol=Supervised, Shot-count=10-shot2026.01 | — | 0.7598 | |
| GHRNSupervision Type=Supervised, Evaluation Setting=10-shot2026.01 | — | 0.7598 | |
| HCM-ALearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.3544 | |
| HCM-ALearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train & Fine-Tune2024.05 | — | 0.3657 | |
| OWLEYEEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.6694 | |
| OWLEYESupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.6485 | |
| SLGADEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.514 | |
| SLGADSupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.4557 | |
| TAMLearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train Only2024.05 | — | 0.6588 | |
| TAMLearning Supervision=Unsupervised, Evaluation Protocol=Pre-Train & Fine-Tune2024.05 | — | 0.7626 | |
| TAMEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.7626 | |
| TAMSupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.633 | |
| UNPromptEvaluation Protocol=Unsupervised & Finetune, Shot-count=10-shot2026.01 | — | 0.6713 | |
| UNPromptSupervision Type=Unsupervised, Evaluation Setting=zero-shot2026.01 | — | 0.6387 |