| Joint contrastive and supervised training | - | 78.3 | Contrastive Training for Improved Out-of-Distribution Detection | - |
| ViT_B-16 finetuned on CIFAR-100 | 91.89 | 95.53 | Exploring the Limits of Out-of-Distribution Detection | |
| MLP-Mixer_B-16 finetuned on CIFAR-100 | 90.22 | 95.31 | Exploring the Limits of Out-of-Distribution Detection | |
| ViT-L_16 finetuned on CIFAR-100 | - | 97.98 | Exploring the Limits of Out-of-Distribution Detection | |
| R50+ViT_B-16 finetuned on CIFAR-100 | 92.08 | 96.23 | Exploring the Limits of Out-of-Distribution Detection | |
| CLIP using class name words describing the two distributions | - | 94.68 | Exploring the Limits of Out-of-Distribution Detection | |
| ResNet + diagonal elements of Gram matrix | - | 76.6 | Detecting Out-of-Distribution Examples with In-distribution Examples and Gram Matrices | |
| DenseNet + diagonal elements of Gram matrix | - | 70.1 | Detecting Out-of-Distribution Examples with In-distribution Examples and Gram Matrices | |