Anomaly Detection on Amazon
60.46AUPRCOWLEYE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OWLEYEEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 60.46 | — | |
| CAREEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 56.76 | — | |
| TAMEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 45.26 | — | |
| ARCLearning setup=Ours2024.05 | 44.25 | 1.9 | |
| DOMINANTEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 36.8 | — | |
| ARCEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 27.47 | — | |
| LUNAR2026.03 | 19.36 | — | |
| GATLearning setup=Supervised, Protocol=Pre-Train Only2024.05 | 15.74 | 7.3 | |
| GHRNEvaluation Protocol=Supervised (10-shot)2026.01 | 13.84 | — | |
| BWGNNLearning setup=Supervised, Protocol=Pre-Train Only2024.05 | 13.12 | 8.6 | |
| BWGNNEvaluation Protocol=Supervised (10-shot)2026.01 | 12.4 | — | |
| DisentAD2026.03 | 11.82 | — | |
| TAMLearning setup=Unsupervised, Protocol=Pre-Train & Fine-Tune2024.05 | 11.56 | 4.1 | |
| CoLALearning setup=Unsupervised, Protocol=Pre-Train Only2024.05 | 11.06 | 9 | |
| TAMLearning setup=Unsupervised, Protocol=Pre-Train Only2024.05 | 10.75 | 5.8 | |
| OFA-TAD2026.03 | 10.28 | — | |
| LOF2026.03 | 10.1 | — | |
| KNN2026.03 | 9.99 | — | |
| AE2026.03 | 9.87 | — | |
| UNPromptEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 9.74 | — | |
| MCM2026.03 | 9.74 | — | |
| DRL2026.03 | 9.61 | — | |
| DSVDD2026.03 | 9.52 | — | |
| iForest2026.03 | 9.47 | — | |
| GHRNLearning setup=Supervised, Protocol=Pre-Train Only2024.05 | 7.54 | 8.4 | |
| BGNNLearning setup=Supervised, Protocol=Pre-Train Only2024.05 | 7.51 | 10.5 | |
| DOMINANTLearning setup=Unsupervised, Protocol=Pre-Train & Fine-Tune2024.05 | 7.48 | 4 | |
| CoLALearning setup=Unsupervised, Protocol=Pre-Train & Fine-Tune2024.05 | 7.27 | 5.8 | |
| GCNLearning setup=Supervised, Protocol=Pre-Train Only2024.05 | 6.96 | 9.6 | |
| DOMINANTLearning setup=Unsupervised, Protocol=Pre-Train Only2024.05 | 6.11 | 6.1 | |
| HCM-ALearning setup=Unsupervised, Protocol=Pre-Train Only2024.05 | 5.87 | 12.1 | |
| HCM-ALearning setup=Unsupervised, Protocol=Pre-Train & Fine-Tune2024.05 | 5.64 | 11.9 | |
| SLGADEvaluation Protocol=Unsupervised (zero-shot)2026.01 | 5.33 | — |