# Image Classification Unlearning on CIFAR-10 50% Random Forgetting Retain set D\_r (train)

Source: [Image Classification Unlearning on CIFAR-10 50% Random Forgetting Retain set D\_r (train)](<https://www.sota2.com/research/sota/image-classification-unlearning-on-cifar-10-50-percent-random-forgetting-retain-set-d_r-train>)

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

Data updated: 2026-04-13T09:47:27.030Z

Evaluation of Machine Unlearning effectiveness on CIFAR-10, specifically measuring model accuracy and performance gap after randomly forgetting 50% of the training data, focusing on metrics derived from the retained training set (D\_r^train).

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 Unlearning |
| Dataset (including any stated split) | CIFAR-10 50% Random Forgetting Retain set D\_r (train) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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](<https://www.sota2.com/research/sota/image-classification-unlearning-on-cifar-10-50-percent-random-forgetting-retain-set-d_r-train.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Performance Gap](<https://www.sota2.com/research/sota/image-classification-unlearning-on-cifar-10-50-percent-random-forgetting-retain-set-d_r-train.md?metric=Performance%20Gap>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| IEU w/GA | 99.3 | Not recorded | 69dc789f7d0172140ec78bd0 |
| FT | 99.2 | Not recorded | 69dc789f7d0172140ec78bd0 |
| SCRUB | 98.6 | Not recorded | 69dc789f7d0172140ec78bd0 |
| Retraining | 94.2 | Not recorded | 69dc789f7d0172140ec78bd0 |
| RL | 94.2 | Not recorded | 69dc789f7d0172140ec78bd0 |
| IEU w/GA+Noisy | 92.2 | Not recorded | 69dc789f7d0172140ec78bd0 |
| IEU w/Noisy | 91.9 | Not recorded | 69dc789f7d0172140ec78bd0 |
| SALUN | 63.6 | Not recorded | 69dc789f7d0172140ec78bd0 |

### IEU w/GA

Accuracy: 99.3. Result ID: 69dcbb9847f34ce924c41213.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

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

### FT

Accuracy: 99.2. Result ID: 69dcbb9847f34ce924c4120f.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

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

### SCRUB

Accuracy: 98.6. Result ID: 69dcbb9847f34ce924c41211.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71011e209131cea4e7b9.md>)

[Implementation](<https://github.com/Meghdad92/SCRUB>)

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

### Retraining

Accuracy: 94.2. Result ID: 69dcbb9847f34ce924c4120e.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

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

### RL

Accuracy: 94.2. Result ID: 69dcbb9847f34ce924c41210.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68881e209131cea1cbe9.md>)

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

### IEU w/GA+Noisy

Accuracy: 92.2. Result ID: 69dcbb9847f34ce924c41215.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

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

### IEU w/Noisy

Accuracy: 91.9. Result ID: 69dcbb9847f34ce924c41214.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

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

### SALUN

Accuracy: 63.6. Result ID: 69dcbb9847f34ce924c41212.

[Reporting paper](<https://www.sota2.com/research/paper/69dc789f7d0172140ec78bd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72cd1e209131cea5888f.md>)

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

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