Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Out-of-Distribution Detection research benchmarks – Page 32 · SOTA2 Research

Out-of-Distribution Detection

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

FOOD101 SVHN iD OoD (test)

0.971Log Loss

2

Feb 26, 2026

FOOD101 iD PLACES365 OoD (test)

0.996Log-Likelihood

2

Feb 26, 2026

FOOD101 (iD) CIFAR-100 (OoD) (test)

0.998Log-Likelihood

2

Feb 26, 2026

99.9OOD Score (SVHN)

2

Feb 26, 2026

ImageNet-O (val)

22.6AUPR

2

Feb 26, 2026

29.92FPR95

2

Feb 26, 2026

KNN (non-parametric)

56.46FPR@95

2

Feb 26, 2026

41.06FPR@95

2

Feb 26, 2026

22.22FPR@95

2

Feb 26, 2026

Genomics (test)

BERT pre-train and fine-tune

77.49Mahalanobis AUROC

2

Feb 26, 2026

WMT English-German Permuted input '17 newstest14 (test)

83.7EoE

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs Gaussian Noise (out-of-distribution) Dense-BC (test)

Baseline (Hendrycks & Gimpel, 2017)

100FPR @ 95% TPR

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs Uniform Noise (out-of-distribution) Dense-BC (test)

Baseline (Hendrycks & Gimpel, 2017)

100FPR (95% TPR)

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs LSUN resize (out-of-distribution) Dense-BC (test)

58FPR (95% TPR)

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs LSUN crop (out-of-distribution) Dense-BC (test)

18.6FPR (95% TPR)

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs TinyImageNet resize (out-of-distribution) Dense-BC (test)

57FPR (95% TPR)

2

Feb 26, 2026

CIFAR-100 (in-distribution) vs TinyImageNet crop (out-of-distribution) Dense-BC (test)

26.9FPR @ 95% TPR

2

Feb 26, 2026

CIFAR-10 (in-distribution) vs Gaussian Noise (out-of-distribution) Dense-BC (test)

0FPR (95% TPR)

2

Feb 26, 2026

CIFAR-10 (in-distribution) vs Uniform Noise (out-of-distribution) Dense-BC (test)

0FPR @ 95% TPR

2

Feb 26, 2026

CIFAR-10 (in-distribution) vs LSUN crop (out-of-distribution) Dense-BC (test)

11.4FPR @ 95% TPR

2

Feb 26, 2026

CIFAR-10 (in-distribution) vs TinyImageNet resize (out-of-distribution) Dense-BC (test)

7.5FPR @ 95% TPR

2

Feb 26, 2026

CIFAR-10 LSUN OOD (unlabeled data stream)

Multi-Task Curriculum Framework

99.48Detection Recall

2

Feb 26, 2026

CIFAR-10 with TinyImageNet (TIN) as OOD (unlabeled data stream)

Multi-Task Curriculum Framework

99.22Detection Recall

2

Feb 26, 2026

IMDB vs Movie Reviews (test)

Softmax Baseline

0.94AUROC

2

Feb 26, 2026

non-face images n=176

Epistemic Uncertainty

75.4AUROC

1

Jun 23, 2026

Page 32 of 34

Popular datasets

CIFAR ImageNet CIFAR-100 SVHN MNIST

Follow for updates

Join SOTA2 Contact