# Image Classification on ColoredMNIST (CMNIST) (test)

Source: [Image Classification on ColoredMNIST (CMNIST) (test)](<https://www.sota2.com/research/sota/image-classification-on-coloredmnist-cmnist-test>)

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

Data updated: 2026-02-26T05:29:03.338Z

Evaluation of Image Classification performance on the test split of the Colored MNIST dataset, referencing the CMNIST variant.

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) | ColoredMNIST (CMNIST) (test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 19 of 19 |
| 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-on-coloredmnist-cmnist-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Oracle | 72.1 | Not recorded | 68fa70621e209131cea4aca2 |
| V-REX | 71.6 | Not recorded | 68fa70621e209131cea4aca2 |
| EQRM | 71.4 | Not recorded | 68fa70621e209131cea4aca2 |
| SD | 70.3 | Not recorded | 68fa70621e209131cea4aca2 |
| IRM | 69.7 | Not recorded | 68fa70621e209131cea4aca2 |
| IGA | 57.7 | Not recorded | 68fa70621e209131cea4aca2 |
| V-REX | 55.2 | Not recorded | 68fa70621e209131cea4aca2 |
| EQRM | 53.4 | Not recorded | 68fa70621e209131cea4aca2 |
| IRM | 52.5 | Not recorded | 68fa70621e209131cea4aca2 |
| IRM | 51.8 | Not recorded | 68fa6b4e1e209131cea2d7b7 |
| MILe | 51.8 | Not recorded | 68fa6b4e1e209131cea2d7b7 |
| ERM | 51.6 | Not recorded | 68fa6b4e1e209131cea2d7b7 |
| REx | 51.6 | Not recorded | 68fa6b4e1e209131cea2d7b7 |
| IGA | 50.7 | Not recorded | 68fa70621e209131cea4aca2 |
| SD | 49.4 | Not recorded | 68fa70621e209131cea4aca2 |
| GrpDRO | 29 | Not recorded | 68fa70621e209131cea4aca2 |
| ERM | 27.9 | Not recorded | 68fa70621e209131cea4aca2 |
| ERM | 27.9 | Not recorded | 68fa70621e209131cea4aca2 |
| GrpDRO | 27.3 | Not recorded | 68fa70621e209131cea4aca2 |

### Oracle

Accuracy: 72.1. Result ID: 699117683a6c314849059cdc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"Combined"}. This record does not assert independent reproduction.

### V-REX

Accuracy: 71.6. Result ID: 699117683a6c314849059cd9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea2190f.md>)

[Implementation](<https://anonymous.4open.science/r/12747e81-8505-43cb-b54e-e75e2344a397/>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### EQRM

Accuracy: 71.4. Result ID: 699117683a6c314849059cdb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Implementation](<https://github.com/cianeastwood/qrm>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### SD

Accuracy: 70.3. Result ID: 699117683a6c314849059cd5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa58bf1e209131ce9b0ed1.md>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### IRM

Accuracy: 69.7. Result ID: 699117683a6c314849059cd1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa673c1e209131cea15133.md>)

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

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### IGA

Accuracy: 57.7. Result ID: 699117683a6c314849059cd7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### V-REX

Accuracy: 55.2. Result ID: 699117683a6c314849059cd8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea2190f.md>)

[Implementation](<https://anonymous.4open.science/r/12747e81-8505-43cb-b54e-e75e2344a397/>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### EQRM

Accuracy: 53.4. Result ID: 699117683a6c314849059cda.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Implementation](<https://github.com/cianeastwood/qrm>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### IRM

Accuracy: 52.5. Result ID: 699117683a6c314849059cd0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa673c1e209131cea15133.md>)

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

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### IRM

Accuracy: 51.8. Result ID: 698e7c58ffc3901de09dfe17.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b4e1e209131cea2d7b7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa673c1e209131cea15133.md>)

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

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

### MILe

Accuracy: 51.8. Result ID: 698e7c58ffc3901de09dfe19.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b4e1e209131cea2d7b7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b4e1e209131cea2d7b7.md>)

[Implementation](<https://github.com/rajeswar18/MILe>)

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

### ERM

Accuracy: 51.6. Result ID: 698e7c58ffc3901de09dfe16.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b4e1e209131cea2d7b7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67d21e209131cea18b07.md>)

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

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

### REx

Accuracy: 51.6. Result ID: 698e7c58ffc3901de09dfe18.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b4e1e209131cea2d7b7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea2190f.md>)

[Implementation](<https://anonymous.4open.science/r/12747e81-8505-43cb-b54e-e75e2344a397/>)

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

### IGA

Accuracy: 50.7. Result ID: 699117683a6c314849059cd6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### SD

Accuracy: 49.4. Result ID: 699117683a6c314849059cd4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa58bf1e209131ce9b0ed1.md>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### GrpDRO

Accuracy: 29. Result ID: 699117683a6c314849059cd3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### ERM

Accuracy: 27.9. Result ID: 699117683a6c314849059cce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

### ERM

Accuracy: 27.9. Result ID: 699117683a6c314849059ccf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"ERM"}. This record does not assert independent reproduction.

### GrpDRO

Accuracy: 27.3. Result ID: 699117683a6c314849059cd2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70621e209131cea4aca2.md>)

Recorded evaluation configuration: {"Initialization":"Rand."}. This record does not assert independent reproduction.

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