CIFAR-10
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
CIFAR-10 50% Random Forgetting (test)
84.8Accuracy
8
CIFAR-10 50% Random Forgetting Forget D_f^train (train)
92.5Accuracy
8
CIFAR-10 30% Random Forgetting Forget set D_f^train (train)
92.9Accuracy
8
CIFAR-10 50% class-wise forgetting (train test)
99.9Utility Retention (Train, D^r)
8
CIFAR-10 alpha=0.5 (test)
36.8Accuracy
8
CIFAR-10 multi-label 28 (beta=0.5, gamma=0.5)
92.2macro AUC
8
CIFAR-10 (retention set (Dr))
99.6Accuracy
8
CIFAR-10 Dir(0.5) (New Clients)
72.94Accuracy (Clean)
8
CIFAR-10 (val)
87.8Accuracy
8
CIFAR-10-LT (gamma=50, beta=10%) (test)
81.5Accuracy
8
CIFAR-10 gamma=20, beta=4% long-tailed variant (test)
85.6Accuracy (%)
8
CIFAR-10 r=0.2 (test)
100Precision
8
CIFAR-10 Type
0.959RI
8
CIFAR-10 Environment
60.75NMI
8
CIFAR-10 (test)
87.06Accuracy
8
CIFAR-10 10% Adversarial deletion (test)
0.745AUC
8
CIFAR-10 Influence-inspired deletion (test)
67.91Accuracy
8
CIFAR-10 Low-margin deletion (test)
68.12Accuracy
8
CIFAR-10 OODRobustBench
76.34C-OODd
8
CIFAR-10 Asymmetric 30% noise (test)
96.3Accuracy
8
CIFAR-10 (test)
94.81Baseline Accuracy
8
CIFAR-10 WaNet attack
93.67Backdoor Accuracy
8
CIFAR-10 (test)
0.031Test Loss Interpolation Barrier
8
CIFAR-10 Automobile Erased
99.72Accuracy (Erase)
8
CIFAR-10 all-to-all setting, Blend attack (test)
94.89Accuracy
8