# Out-of-Distribution Detection on CIFAR100 (test) (OOD Targets: SVHN, TinyImageNet, LSUN, Places, Textures)

Source: [Out-of-Distribution Detection on CIFAR100 (test) (OOD Targets: SVHN, TinyImageNet, LSUN, Places, Textures)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:36:11.449Z

Evaluation of Out-of-Distribution Detection performance using the CIFAR-100 test set as in-distribution data, reporting FPR95 and AUROC against a specific set of OOD targets: SVHN, TinyImageNet, LSUN, Places, and Textures.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Out-of-Distribution Detection |
| Dataset (including any stated split) | CIFAR100 (test) |
| Metric | Avg FPR95 (Overall) |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 46 of 46 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Avg FPR95 (Overall)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=Avg%20FPR95%20(Overall)>): Lower is better; unit: PERCENT.

- [FPR95 (SVHN)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=FPR95%20(SVHN)>): Lower is better; unit: PERCENT.

- [AUROC (SVHN)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=AUROC%20(SVHN)>): Higher is better; unit: PERCENT.

- [FPR95 (TinyImageNet)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=FPR95%20(TinyImageNet)>): Lower is better; unit: PERCENT.

- [AUROC (TinyImageNet)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=AUROC%20(TinyImageNet)>): Higher is better; unit: PERCENT.

- [FPR95 (LSUN)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=FPR95%20(LSUN)>): Lower is better; unit: PERCENT.

- [AUROC (LSUN)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=AUROC%20(LSUN)>): Higher is better; unit: PERCENT.

- [FPR95 (Places)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=FPR95%20(Places)>): Lower is better; unit: PERCENT.

- [AUROC (Places)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=AUROC%20(Places)>): Higher is better; unit: PERCENT.

- [FPR95 (Textures)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=FPR95%20(Textures)>): Lower is better; unit: PERCENT.

- [AUROC (Textures)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=AUROC%20(Textures)>): Higher is better; unit: PERCENT.

- [Avg AUROC (Overall)](<https://www.sota2.com/research/sota/out-of-distribution-detection-on-cifar100-test-ood-targets-svhn-tinyimagenet-lsun-places-textures.md?metric=Avg%20AUROC%20(Overall)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Avg FPR95 (Overall) | Result date | Report paper |
| --- | --- | --- | --- |
| GAIA-Z | 29.1 | Not recorded | 68fa71ce1e209131cea530aa |
| GAIA-A | 36.29 | Not recorded | 68fa71ce1e209131cea530aa |
| GAIA-Z | 38.63 | Not recorded | 68fa71ce1e209131cea530aa |
| WeiPer+KLD | 48.78 | Not recorded | 68fa73631e209131cea5c1b0 |
| WeiPer+MSP | 49.28 | Not recorded | 68fa73631e209131cea5c1b0 |
| RMDS | 49.56 | Not recorded | 68fa73631e209131cea5c1b0 |
| KNN | 49.65 | Not recorded | 68fa73631e209131cea5c1b0 |
| WeiPer+ReAct | 49.65 | Not recorded | 68fa73631e209131cea5c1b0 |
| GEN | 49.98 | Not recorded | 68fa73631e209131cea5c1b0 |
| TempScale | 50.26 | Not recorded | 68fa73631e209131cea5c1b0 |
| MSP | 50.7 | Not recorded | 68fa73631e209131cea5c1b0 |
| ReAct | 51.47 | Not recorded | 68fa73631e209131cea5c1b0 |
| MLS | 51.83 | Not recorded | 68fa73631e209131cea5c1b0 |
| EBO | 52.03 | Not recorded | 68fa73631e209131cea5c1b0 |
| OpenMax | 52.99 | Not recorded | 68fa73631e209131cea5c1b0 |
| VIM | 54.66 | Not recorded | 68fa73631e209131cea5c1b0 |
| DICE | 54.93 | Not recorded | 68fa73631e209131cea5c1b0 |
| ODIN | 55.19 | Not recorded | 68fa73631e209131cea5c1b0 |
| SHE | 57.74 | Not recorded | 68fa73631e209131cea5c1b0 |
| ASH | 63.35 | Not recorded | 68fa73631e209131cea5c1b0 |
| KNN | 66.13 | Not recorded | 68fa71ce1e209131cea530aa |
| GAIA-A | 68.97 | Not recorded | 68fa71ce1e209131cea530aa |
| KLM | 71.07 | Not recorded | 68fa73631e209131cea5c1b0 |
| OpenGAN | 74.21 | Not recorded | 68fa73631e209131cea5c1b0 |
| ReAct | 74.51 | Not recorded | 68fa71ce1e209131cea530aa |
| ReAct | 74.67 | Not recorded | 68fa71ce1e209131cea530aa |
| KNN | 77.25 | Not recorded | 68fa71ce1e209131cea530aa |
| RankFeat | 78.4 | Not recorded | 68fa73631e209131cea5c1b0 |
| MDS | 79.05 | Not recorded | 68fa73631e209131cea5c1b0 |
| ASH-P@70 | 79.8 | Not recorded | 68fa71ce1e209131cea530aa |
| ODIN | 80.1 | Not recorded | 68fa71ce1e209131cea530aa |
| ASH-P@65 | 80.7 | Not recorded | 68fa71ce1e209131cea530aa |
| Mahalanobis | 80.97 | Not recorded | 68fa71ce1e209131cea530aa |
| MSP | 81.26 | Not recorded | 68fa71ce1e209131cea530aa |
| Mahalanobis | 81.29 | Not recorded | 68fa71ce1e209131cea530aa |
| Energy | 81.55 | Not recorded | 68fa71ce1e209131cea530aa |
| ODIN | 82.61 | Not recorded | 68fa71ce1e209131cea530aa |
| MSP | 82.84 | Not recorded | 68fa71ce1e209131cea530aa |
| Energy | 83.29 | Not recorded | 68fa71ce1e209131cea530aa |
| GradNorm | 85.89 | Not recorded | 68fa71ce1e209131cea530aa |
| GradNorm | 86.32 | Not recorded | 68fa71ce1e209131cea530aa |
| GradNorm | 86.85 | Not recorded | 68fa73631e209131cea5c1b0 |
| Rankfeat | 88.29 | Not recorded | 68fa71ce1e209131cea530aa |
| Rankfeat | 89.75 | Not recorded | 68fa71ce1e209131cea530aa |
| Gram | 91.85 | Not recorded | 68fa73631e209131cea5c1b0 |
| MDSEns | 95.82 | Not recorded | 68fa73631e209131cea5c1b0 |

### GAIA-Z

Avg FPR95 (Overall): 29.1. Result ID: 6991e5683a6c3148490b4a5b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Implementation](<https://github.com/JGEthanChen/GAIA-OOD>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GAIA-A

Avg FPR95 (Overall): 36.29. Result ID: 6991e5683a6c3148490b4a67.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Implementation](<https://github.com/JGEthanChen/GAIA-OOD>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GAIA-Z

Avg FPR95 (Overall): 38.63. Result ID: 6991e5683a6c3148490b4a66.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Implementation](<https://github.com/JGEthanChen/GAIA-OOD>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### WeiPer+KLD

Avg FPR95 (Overall): 48.78. Result ID: 6992d72f3a6c31484911d966.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### WeiPer+MSP

Avg FPR95 (Overall): 49.28. Result ID: 6992d72f3a6c31484911d964.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### RMDS

Avg FPR95 (Overall): 49.56. Result ID: 6992d72f3a6c31484911d955.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a271e209131cea2666a.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### KNN

Avg FPR95 (Overall): 49.65. Result ID: 6992d72f3a6c31484911d95e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e2f1e209131cea3e69b.md>)

[Implementation](<https://github.com/deeplearning-wisc/knn-ood>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### WeiPer+ReAct

Avg FPR95 (Overall): 49.65. Result ID: 6992d72f3a6c31484911d965.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### GEN

Avg FPR95 (Overall): 49.98. Result ID: 6992d72f3a6c31484911d963.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### TempScale

Avg FPR95 (Overall): 50.26. Result ID: 6992d72f3a6c31484911d951.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa638b1e209131ce9ff6d1.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### MSP

Avg FPR95 (Overall): 50.7. Result ID: 6992d72f3a6c31484911d950.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### ReAct

Avg FPR95 (Overall): 51.47. Result ID: 6992d72f3a6c31484911d95a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b4f1e209131cea2d8b5.md>)

[Implementation](<https://github.com/deeplearning-wisc/react.git>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### MLS

Avg FPR95 (Overall): 51.83. Result ID: 6992d72f3a6c31484911d95b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c811e209131cea34603.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### EBO

Avg FPR95 (Overall): 52.03. Result ID: 6992d72f3a6c31484911d957.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cc1e209131cea24242.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### OpenMax

Avg FPR95 (Overall): 52.99. Result ID: 6992d72f3a6c31484911d94f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa622a1e209131ce9f7cf0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### VIM

Avg FPR95 (Overall): 54.66. Result ID: 6992d72f3a6c31484911d95d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c151e209131cea32091.md>)

[Implementation](<https://github.com/haoqiwang/vim>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### DICE

Avg FPR95 (Overall): 54.93. Result ID: 6992d72f3a6c31484911d95f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cfc1e209131cea371c4.md>)

[Implementation](<https://github.com/deeplearning-wisc/dice.git>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### ODIN

Avg FPR95 (Overall): 55.19. Result ID: 6992d72f3a6c31484911d952.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### SHE

Avg FPR95 (Overall): 57.74. Result ID: 6992d72f3a6c31484911d962.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### ASH

Avg FPR95 (Overall): 63.35. Result ID: 6992d72f3a6c31484911d961.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f371e209131cea447f1.md>)

[Implementation](<https://andrijazz.github.io/ash>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### KNN

Avg FPR95 (Overall): 66.13. Result ID: 6991e5683a6c3148490b4a63.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e2f1e209131cea3e69b.md>)

[Implementation](<https://github.com/deeplearning-wisc/knn-ood>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GAIA-A

Avg FPR95 (Overall): 68.97. Result ID: 6991e5683a6c3148490b4a5c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Implementation](<https://github.com/JGEthanChen/GAIA-OOD>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### KLM

Avg FPR95 (Overall): 71.07. Result ID: 6992d72f3a6c31484911d95c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee61e209131cea42c15.md>)

[Implementation](<https://github.com/LINe-OOD>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### OpenGAN

Avg FPR95 (Overall): 74.21. Result ID: 6992d72f3a6c31484911d958.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### ReAct

Avg FPR95 (Overall): 74.51. Result ID: 6991e5683a6c3148490b4a56.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b4f1e209131cea2d8b5.md>)

[Implementation](<https://github.com/deeplearning-wisc/react.git>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### ReAct

Avg FPR95 (Overall): 74.67. Result ID: 6991e5683a6c3148490b4a61.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b4f1e209131cea2d8b5.md>)

[Implementation](<https://github.com/deeplearning-wisc/react.git>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### KNN

Avg FPR95 (Overall): 77.25. Result ID: 6991e5683a6c3148490b4a58.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e2f1e209131cea3e69b.md>)

[Implementation](<https://github.com/deeplearning-wisc/knn-ood>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### RankFeat

Avg FPR95 (Overall): 78.4. Result ID: 6992d72f3a6c31484911d960.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### MDS

Avg FPR95 (Overall): 79.05. Result ID: 6992d72f3a6c31484911d953.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64f91e209131cea077ac.md>)

[Implementation](<https://github.com/pokaxpoka/deep_Mahalanobis_detector>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### ASH-P@70

Avg FPR95 (Overall): 79.8. Result ID: 6991e5683a6c3148490b4a65.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f371e209131cea447f1.md>)

[Implementation](<https://andrijazz.github.io/ash>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### ODIN

Avg FPR95 (Overall): 80.1. Result ID: 6991e5683a6c3148490b4a5e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa684a1e209131cea1b26f.md>)

[Implementation](<https://github.com/facebookresearch/odin>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### ASH-P@65

Avg FPR95 (Overall): 80.7. Result ID: 6991e5683a6c3148490b4a5a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f371e209131cea447f1.md>)

[Implementation](<https://andrijazz.github.io/ash>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Mahalanobis

Avg FPR95 (Overall): 80.97. Result ID: 6991e5683a6c3148490b4a60.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64f91e209131cea077ac.md>)

[Implementation](<https://github.com/pokaxpoka/deep_Mahalanobis_detector>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### MSP

Avg FPR95 (Overall): 81.26. Result ID: 6991e5683a6c3148490b4a5d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Mahalanobis

Avg FPR95 (Overall): 81.29. Result ID: 6991e5683a6c3148490b4a55.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64f91e209131cea077ac.md>)

[Implementation](<https://github.com/pokaxpoka/deep_Mahalanobis_detector>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Energy

Avg FPR95 (Overall): 81.55. Result ID: 6991e5683a6c3148490b4a5f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cc1e209131cea24242.md>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### ODIN

Avg FPR95 (Overall): 82.61. Result ID: 6991e5683a6c3148490b4a53.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa684a1e209131cea1b26f.md>)

[Implementation](<https://github.com/facebookresearch/odin>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### MSP

Avg FPR95 (Overall): 82.84. Result ID: 6991e5683a6c3148490b4a52.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Energy

Avg FPR95 (Overall): 83.29. Result ID: 6991e5683a6c3148490b4a54.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cc1e209131cea24242.md>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GradNorm

Avg FPR95 (Overall): 85.89. Result ID: 6991e5683a6c3148490b4a62.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6afd1e209131cea2b8e2.md>)

[Implementation](<https://github.com/deeplearning-wisc/gradnorm_ood>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GradNorm

Avg FPR95 (Overall): 86.32. Result ID: 6991e5683a6c3148490b4a57.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6afd1e209131cea2b8e2.md>)

[Implementation](<https://github.com/deeplearning-wisc/gradnorm_ood>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### GradNorm

Avg FPR95 (Overall): 86.85. Result ID: 6992d72f3a6c31484911d959.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### Rankfeat

Avg FPR95 (Overall): 88.29. Result ID: 6991e5683a6c3148490b4a64.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d731e209131cea39fef.md>)

[Implementation](<https://github.com/KingJamesSong/RankFeat>)

Recorded evaluation configuration: {"Backbone":"WRN40","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Rankfeat

Avg FPR95 (Overall): 89.75. Result ID: 6991e5683a6c3148490b4a59.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ce1e209131cea530aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d731e209131cea39fef.md>)

[Implementation](<https://github.com/KingJamesSong/RankFeat>)

Recorded evaluation configuration: {"Backbone":"ResNet34","Pre-training loss":"cross-entropy loss"}. This record does not assert independent reproduction.

### Gram

Avg FPR95 (Overall): 91.85. Result ID: 6992d72f3a6c31484911d956.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

### MDSEns

Avg FPR95 (Overall): 95.82. Result ID: 6992d72f3a6c31484911d954.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73631e209131cea5c1b0.md>)

Recorded evaluation configuration: {"Backbone":"ResNet18"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.