# Image Classification on CIFAR10 (test) (Accuracy vs Corruption Severity)

Source: [Image Classification on CIFAR10 (test) (Accuracy vs Corruption Severity)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity>)

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

Data updated: 2026-02-26T05:34:26.745Z

Evaluation of Image Classification performance on the CIFAR-10 test set, measuring averaged accuracy across six distinct levels of input perturbation severity (0 through 5).

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 | Image Classification |
| Dataset (including any stated split) | CIFAR10 (test) |
| Metric | Accuracy (Severity 0) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [Accuracy (Severity 0)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%200)>): Higher is better; unit: PERCENT.

- [Accuracy (Severity 1)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%201)>): Higher is better; unit: PERCENT.

- [Accuracy (Severity 2)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%202)>): Higher is better; unit: PERCENT.

- [Accuracy (Severity 3)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%203)>): Higher is better; unit: PERCENT.

- [Accuracy (Severity 4)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%204)>): Higher is better; unit: PERCENT.

- [Accuracy (Severity 5)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-test-accuracy-vs-corruption-severity.md?metric=Accuracy%20(Severity%205)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (Severity 0) | Result date | Report paper |
| --- | --- | --- | --- |
| SharpDRO | 91.9 | Not recorded | 69535c9a53fdd5c36173db85 |
| JTT | 91.3 | Not recorded | 69535c9a53fdd5c36173db85 |
| SharpDRO | 91.2 | Not recorded | 69535c9a53fdd5c36173db85 |
| EIIL | 90.3 | Not recorded | 69535c9a53fdd5c36173db85 |
| JTT | 89.9 | Not recorded | 69535c9a53fdd5c36173db85 |
| EIIL | 88.6 | Not recorded | 69535c9a53fdd5c36173db85 |

### SharpDRO

Accuracy (Severity 0): 91.9. Result ID: 6999344b5698665ec49d5c46.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

Recorded evaluation configuration: {"Corruption Type":"Shot"}. This record does not assert independent reproduction.

### JTT

Accuracy (Severity 0): 91.3. Result ID: 6999344b5698665ec49d5c44.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ade1e209131cea2aeeb.md>)

Recorded evaluation configuration: {"Corruption Type":"Shot"}. This record does not assert independent reproduction.

### SharpDRO

Accuracy (Severity 0): 91.2. Result ID: 6999344b5698665ec49d5c43.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

Recorded evaluation configuration: {"Corruption Type":"Gaussian"}. This record does not assert independent reproduction.

### EIIL

Accuracy (Severity 0): 90.3. Result ID: 6999344b5698665ec49d5c45.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a5f1e209131cea27cee.md>)

Recorded evaluation configuration: {"Corruption Type":"Shot"}. This record does not assert independent reproduction.

### JTT

Accuracy (Severity 0): 89.9. Result ID: 6999344b5698665ec49d5c41.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ade1e209131cea2aeeb.md>)

Recorded evaluation configuration: {"Corruption Type":"Gaussian"}. This record does not assert independent reproduction.

### EIIL

Accuracy (Severity 0): 88.6. Result ID: 6999344b5698665ec49d5c42.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9a53fdd5c36173db85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a5f1e209131cea27cee.md>)

Recorded evaluation configuration: {"Corruption Type":"Gaussian"}. This record does not assert independent reproduction.

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