# Image Classification on CIFAR-100 (test) (Symmetric and Asymmetric Noise Robustness)

Source: [Image Classification on CIFAR-100 (test) (Symmetric and Asymmetric Noise Robustness)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness>)

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

Data updated: 2026-02-26T06:05:47.573Z

Evaluation of Image Classification performance on the CIFAR-100 test set, specifically measuring accuracy under varying levels of Symmetric Noise and Asymmetric Noise corruption defined by parameter η.

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-100 (test) |
| Metric | Symmetric Noise (η=0.2) Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [Symmetric Noise (η=0.2) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Symmetric%20Noise%20(%CE%B7%3D0.2)%20Accuracy>): Higher is better; unit: PERCENT.

- [Symmetric Noise (η=0.0) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Symmetric%20Noise%20(%CE%B7%3D0.0)%20Accuracy>): Higher is better; unit: PERCENT.

- [Symmetric Noise (η=0.4) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Symmetric%20Noise%20(%CE%B7%3D0.4)%20Accuracy>): Higher is better; unit: PERCENT.

- [Symmetric Noise (η=0.6) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Symmetric%20Noise%20(%CE%B7%3D0.6)%20Accuracy>): Higher is better; unit: PERCENT.

- [Symmetric Noise (η=0.8) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Symmetric%20Noise%20(%CE%B7%3D0.8)%20Accuracy>): Higher is better; unit: PERCENT.

- [Asymmetric Noise (η=0.2) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Asymmetric%20Noise%20(%CE%B7%3D0.2)%20Accuracy>): Higher is better; unit: PERCENT.

- [Asymmetric Noise (η=0.3) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Asymmetric%20Noise%20(%CE%B7%3D0.3)%20Accuracy>): Higher is better; unit: PERCENT.

- [Asymmetric Noise (η=0.4) Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-test-symmetric-and-asymmetric-noise-robustness.md?metric=Asymmetric%20Noise%20(%CE%B7%3D0.4)%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Symmetric Noise (η=0.2) Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| FasTEN | 68.75 | Not recorded | 68fa6cf71e209131cea36e68 |
| MSLC | 68.62 | Not recorded | 68fa6cf71e209131cea36e68 |
| MLOC | 68.16 | Not recorded | 68fa6cf71e209131cea36e68 |
| MW-Net | 66.73 | Not recorded | 68fa6cf71e209131cea36e68 |
| GLC | 60.99 | Not recorded | 68fa6cf71e209131cea36e68 |
| SL | 60.01 | Not recorded | 68fa66351e209131cea0ec32 |
| Forward | 59.75 | Not recorded | 68fa66351e209131cea0ec32 |
| Deep kNN | 59.6 | Not recorded | 68fa6cf71e209131cea36e68 |
| CE | 59.26 | Not recorded | 68fa66351e209131cea0ec32 |
| D2L | 59.2 | Not recorded | 68fa66351e209131cea0ec32 |
| GCE | 59.06 | Not recorded | 68fa66351e209131cea0ec32 |
| LSR | 58.83 | Not recorded | 68fa66351e209131cea0ec32 |
| Bootstrap | 57.91 | Not recorded | 68fa66351e209131cea0ec32 |
| L2RW | 57.79 | Not recorded | 68fa6cf71e209131cea36e68 |
| MLaC | 49.81 | Not recorded | 68fa6cf71e209131cea36e68 |

### FasTEN

Symmetric Noise (η=0.2) Accuracy: 68.75. Result ID: 698f6f443a6c314849f88b06.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Implementation](<https://github.com/hyperconnect/FasTEN>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### MSLC

Symmetric Noise (η=0.2) Accuracy: 68.62. Result ID: 698f6f443a6c314849f88b05.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c1a1e209131cea3233a.md>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### MLOC

Symmetric Noise (η=0.2) Accuracy: 68.16. Result ID: 698f6f443a6c314849f88b03.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### MW-Net

Symmetric Noise (η=0.2) Accuracy: 66.73. Result ID: 698f6f443a6c314849f88b00.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a7e1e209131cea289fe.md>)

[Implementation](<https://aka.ms/MLC>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### GLC

Symmetric Noise (η=0.2) Accuracy: 60.99. Result ID: 698f6f443a6c314849f88b02.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a7e1e209131cea289fe.md>)

[Implementation](<https://aka.ms/MLC>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### SL

Symmetric Noise (η=0.2) Accuracy: 60.01. Result ID: 698cdf9b8b517eb4331d54ba.

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Forward

Symmetric Noise (η=0.2) Accuracy: 59.75. Result ID: 698cdf9b8b517eb4331d54b7.

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Deep kNN

Symmetric Noise (η=0.2) Accuracy: 59.6. Result ID: 698f6f443a6c314849f88b01.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa676c1e209131cea160c5.md>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### CE

Symmetric Noise (η=0.2) Accuracy: 59.26. Result ID: 698cdf9b8b517eb4331d54b4.

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### D2L

Symmetric Noise (η=0.2) Accuracy: 59.2. Result ID: 698cdf9b8b517eb4331d54b8.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa64a71e209131cea05e51.md>)

[Implementation](<https://github.com/xingjunm/dimensionality-driven-learning>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GCE

Symmetric Noise (η=0.2) Accuracy: 59.06. Result ID: 698cdf9b8b517eb4331d54b9.

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### LSR

Symmetric Noise (η=0.2) Accuracy: 58.83. Result ID: 698cdf9b8b517eb4331d54b5.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa63171e209131ce9fc942.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Bootstrap

Symmetric Noise (η=0.2) Accuracy: 57.91. Result ID: 698cdf9b8b517eb4331d54b6.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa61d01e209131ce9f63bb.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### L2RW

Symmetric Noise (η=0.2) Accuracy: 57.79. Result ID: 698f6f443a6c314849f88aff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65be1e209131cea0c1af.md>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

### MLaC

Symmetric Noise (η=0.2) Accuracy: 49.81. Result ID: 698f6f443a6c314849f88b04.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cf71e209131cea36e68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a7e1e209131cea289fe.md>)

[Implementation](<https://aka.ms/MLC>)

Recorded evaluation configuration: {"Backbone":"ResNet-34"}. This record does not assert independent reproduction.

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