Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

MNIST

Benchmarks

Task NameDataset NameSOTA ResultTrend
Image ClassificationMNIST (test)
Accuracy99.35
894
Image ClassificationMNIST
Accuracy99.87
417
Image ClassificationMNIST
Accuracy100
398
Image ClassificationMNIST (test)
Accuracy99.6
196
Image ClassificationMNIST (test)
Test Accuracy99.67
189
Image ClassificationMNIST (test)
Accuracy98.15
177
ClusteringMNIST (test)
NMI0.97
132
ClusteringMNIST
NMI0.978
113
Backdoor CertificationMNIST (test)
AER1.81
112
Graph ClassificationMNIST (test)
Accuracy98.8
110
Image ClassificationMNIST rotated (test)
Test Error (%)0.59
105
Image GenerationMNIST Binary (test)
FID4.69
98
ClusteringMNIST (full)
Accuracy98.7
98
Graph classificationMNIST
Accuracy99.2
95
Digit ClassificationMNIST (test)
Error Rate0.21
94
Anomaly DetectionMNIST
AUC100
90
Federated Learning ClassificationMNIST FedAvg 0.962 (test)
Accuracy96.2
88
Anomaly DetectionMNIST
AUC99.35
87
Image GenerationMNIST
FID1.02
85
Image ClassificationMNIST 0.5 (test)
Accuracy96.28
80
Pixel-by-pixel Image ClassificationPermuted Sequential MNIST (pMNIST) (test)
Accuracy98.76
79
Image ClassificationMNIST -> USPS (test)
Accuracy99.5
79
Image ClassificationMNIST
Clean Accuracy99.64
71
Image Classificationpermuted MNIST (pMNIST) (test)
Accuracy97.87
69
Anomaly DetectionMNIST (test)
AUC99.6
65
Showing 25 of 1026 rows
...