MNIST
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Mnist
89.9SMD
6
MNIST Input-Aware Attack (test)
99.41Acc
6
MNIST BadNets (test)
99.06Acc
6
MNIST Wide first 1000 images (val)
-4.2Relative Bound
6
MNIST FFNet
1.7Time (s)
6
MNIST (evaluation)
1.68IM1
6
MNIST
4.33Avg Latency (ms)
6
MNIST Overlap
0.035Absolute Estimation Error
6
MNIST Overlap
79Accuracy
6
MNIST 17
98.7Accuracy
6
Binarized MNIST
96.3Accuracy
6
MNIST (inlier) vs. CelebA (outlier) (test)
0.943ROC AUC
6
MNIST (val)
99.1Acc
6
MNIST (test)
1.07Standard Error
6
Disjoint MNIST (test)
96.59Accuracy
6
MNIST Scattered
0.99Accuracy
6
MNIST Asymmetry-40% noise (test)
95.24Test Accuracy
6
MNIST Symmetry-80% noise (test)
84.89Accuracy
6
MNIST (test)
99.1Accuracy (Test)
6
Permuted MNIST (test)
98.13Accuracy
6
Sequential MNIST (test)
0.992Accuracy
6
Split MNIST
99Avg Acc (A<=2)
6
MNIST half matching (test)
49.15Sum of Correlations
6
MNIST
17.18Time Cost (s)
6
Permuted MNIST (test)
1.7Error Rate
6