# Image Classification on CIFAR-10 (test) (Symmetry Perturbation Metrics)

Source: [Image Classification on CIFAR-10 (test) (Symmetry Perturbation Metrics)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics>)

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

Data updated: 2026-02-26T05:34:27.470Z

Evaluation of Image Classification performance on the CIFAR-10 test set using metrics assessing robustness under symmetric and asymmetric perturbations (e.g., Sym. 0% to 90%, Asym. 40%).

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) | CIFAR-10 (test) |
| Metric | Accuracy (Sym. 0%) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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 (Sym. 0%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%200%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Sym. 20%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%2020%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Sym. 40%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%2040%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Sym. 60%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%2060%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Sym. 80%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%2080%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Sym. 90%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Sym.%2090%25)>): Higher is better; unit: PERCENT.

- [Accuracy (Asym. 40%)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-symmetry-perturbation-metrics.md?metric=Accuracy%20(Asym.%2040%25)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (Sym. 0%) | Result date | Report paper |
| --- | --- | --- | --- |
| CTRR | 94.29 | Not recorded | 68fa6d141e209131cea377fb |
| SL | 94.21 | Not recorded | 68fa6d141e209131cea377fb |
| CE | 93.97 | Not recorded | 68fa6d141e209131cea377fb |
| APL | 93.97 | Not recorded | 68fa6d141e209131cea377fb |
| Forward | 93.47 | Not recorded | 68fa6d141e209131cea377fb |
| LIMIT | 93.47 | Not recorded | 68fa6d141e209131cea377fb |
| Co-teaching | 93.37 | Not recorded | 68fa6d141e209131cea377fb |
| SLN | 93.21 | Not recorded | 68fa6d141e209131cea377fb |
| GCE | 92.38 | Not recorded | 68fa6d141e209131cea377fb |

### CTRR

Accuracy (Sym. 0%): 94.29. Result ID: 698f8a7f3a6c314849f9efc0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Implementation](<https://github.com/liyi01827/noisy-contrastive>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### SL

Accuracy (Sym. 0%): 94.21. Result ID: 698f8a7f3a6c314849f9efbe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66351e209131cea0ec32.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### CE

Accuracy (Sym. 0%): 93.97. Result ID: 698f8a7f3a6c314849f9efb8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### APL

Accuracy (Sym. 0%): 93.97. Result ID: 698f8a7f3a6c314849f9efbf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67c91e209131cea185bb.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### Forward

Accuracy (Sym. 0%): 93.47. Result ID: 698f8a7f3a6c314849f9efb9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63321e209131ce9fd4c8.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### LIMIT

Accuracy (Sym. 0%): 93.47. Result ID: 698f8a7f3a6c314849f9efbc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68c51e209131cea1e4d2.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### Co-teaching

Accuracy (Sym. 0%): 93.37. Result ID: 698f8a7f3a6c314849f9efbb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64fe1e209131cea07aa4.md>)

[Implementation](<https://github.com/bhanML/Co-teaching>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### SLN

Accuracy (Sym. 0%): 93.21. Result ID: 698f8a7f3a6c314849f9efbd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

### GCE

Accuracy (Sym. 0%): 92.38. Result ID: 698f8a7f3a6c314849f9efba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d141e209131cea377fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa651f1e209131cea088eb.md>)

Recorded evaluation configuration: {"Backbone":"PreAct ResNet18","Number of runs":"3"}. This record does not assert independent reproduction.

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