# Image Classification on CIFAR-10 DARTS search space (test)

Source: [Image Classification on CIFAR-10 DARTS search space (test)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test>)

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

Data updated: 2026-02-26T05:37:46.309Z

Evaluation of Image Classification performance on the test split of the CIFAR-10 dataset restricted to the specific configuration defined by the DARTS search space.

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 DARTS search space (test) |
| Metric | Best Test Error |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 21 of 21 |
| 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

- [Best Test Error](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test.md?metric=Best%20Test%20Error>): Lower is better; unit: PERCENT.

- [S-KdT](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test.md?metric=S-KdT>): Lower is better; unit: SCALAR.

- [Mean Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test.md?metric=Mean%20Accuracy>): Higher is better; unit: PERCENT.

- [Best Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test.md?metric=Best%20Accuracy>): Higher is better; unit: PERCENT.

- [Mean Test Error](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-darts-search-space-test.md?metric=Mean%20Test%20Error>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Best Test Error | Result date | Report paper |
| --- | --- | --- | --- |
| BONAS-D | 2.43 | Not recorded | 68fa71661e209131cea512e5 |
| Shapley-NAS | 2.43 | Not recorded | 68fa71661e209131cea512e5 |
| BONAS-C | 2.46 | Not recorded | 68fa71661e209131cea512e5 |
| NASBOWL | 2.5 | Not recorded | 68fa71661e209131cea512e5 |
| DrNAS (reproduced) | 2.51 | Not recorded | 68fa71661e209131cea512e5 |
| BOHNAS | 2.55 | Not recorded | 68fa71661e209131cea512e5 |
| BANANAS | 2.57 | Not recorded | 68fa71661e209131cea512e5 |
| Random-WS | 2.71 | Not recorded | 68fa71661e209131cea512e5 |
| NASBOWL (reproduced) | 2.83 | Not recorded | 68fa71661e209131cea512e5 |
| ASHA | 2.85 | Not recorded | 68fa71661e209131cea512e5 |
| ENAS (expanded search space) | 2.89 | Not recorded | 68fa71661e209131cea512e5 |
| GDAS | 2.93 | Not recorded | 68fa71661e209131cea512e5 |
| Shapley-NAS (reproduced) | 2.97 | Not recorded | 68fa71661e209131cea512e5 |
| GP-NAS | 3.79 | Not recorded | 68fa71661e209131cea512e5 |

### BONAS-D

Best Test Error: 2.43. Result ID: 6991e02d3a6c3148490b27ec.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68d61e209131cea1e894.md>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.3M","Search cost":"10 GPU days"}. This record does not assert independent reproduction.

### Shapley-NAS

Best Test Error: 2.43. Result ID: 6991e02d3a6c3148490b27f1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cdc1e209131cea3657b.md>)

[Implementation](<https://github.com/Euphoria16/Shapley-NAS.git>)

Recorded evaluation configuration: {"Type":"gradient","Params":"3.4/3.6M","Search cost":"0.3 GPU day"}. This record does not assert independent reproduction.

### BONAS-C

Best Test Error: 2.46. Result ID: 6991e02d3a6c3148490b27eb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68d61e209131cea1e894.md>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.48M","Search cost":"7.5 GPU days"}. This record does not assert independent reproduction.

### NASBOWL

Best Test Error: 2.5. Result ID: 6991e02d3a6c3148490b27ed.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa694c1e209131cea21575.md>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.7M","Search cost":"3 GPU days"}. This record does not assert independent reproduction.

### DrNAS (reproduced)

Best Test Error: 2.51. Result ID: 6991e02d3a6c3148490b27f0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69941e209131cea22d24.md>)

[Implementation](<https://github.com/xiangning-chen/DrNAS>)

Recorded evaluation configuration: {"Type":"gradient","Params":"4.1M","Search cost":"N/A"}. This record does not assert independent reproduction.

### BOHNAS

Best Test Error: 2.55. Result ID: 6991e02d3a6c3148490b27f3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Implementation](<https://github.com/automl/hierarchical_nas_construction>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.9M","Search cost":"8 GPU days"}. This record does not assert independent reproduction.

### BANANAS

Best Test Error: 2.57. Result ID: 6991e02d3a6c3148490b27ea.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cb71e209131cea3580d.md>)

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

Recorded evaluation configuration: {"Type":"BO","Params":"not reported","Search cost":"12 GPU days"}. This record does not assert independent reproduction.

### Random-WS

Best Test Error: 2.71. Result ID: 6991e02d3a6c3148490b27e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa659e1e209131cea0b3fc.md>)

Recorded evaluation configuration: {"Type":"RS","Params":"4.3M","Search cost":"10 GPU days"}. This record does not assert independent reproduction.

### NASBOWL (reproduced)

Best Test Error: 2.83. Result ID: 6991e02d3a6c3148490b27ee.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa694c1e209131cea21575.md>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.7M","Search cost":"N/A"}. This record does not assert independent reproduction.

### ASHA

Best Test Error: 2.85. Result ID: 6991e02d3a6c3148490b27e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa661f1e209131cea0e7c9.md>)

[Implementation](<https://github.com/liamcli/randomNAS_release>)

Recorded evaluation configuration: {"Type":"RS with early stopping","Params":"4.6M","Search cost":"9 GPU days"}. This record does not assert independent reproduction.

### ENAS (expanded search space)

Best Test Error: 2.89. Result ID: 6991e02d3a6c3148490b27e8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64061e209131cea020e1.md>)

Recorded evaluation configuration: {"Type":"RL","Params":"4.6M","Search cost":"6 GPU days"}. This record does not assert independent reproduction.

### GDAS

Best Test Error: 2.93. Result ID: 6991e02d3a6c3148490b27ef.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66ec1e209131cea133b4.md>)

Recorded evaluation configuration: {"Type":"gradient","Params":"3.4M","Search cost":"0.3 GPU day"}. This record does not assert independent reproduction.

### Shapley-NAS (reproduced)

Best Test Error: 2.97. Result ID: 6991e02d3a6c3148490b27f2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cdc1e209131cea3657b.md>)

[Implementation](<https://github.com/Euphoria16/Shapley-NAS.git>)

Recorded evaluation configuration: {"Type":"gradient","Params":"3.6M","Search cost":"N/A"}. This record does not assert independent reproduction.

### GP-NAS

Best Test Error: 3.79. Result ID: 6991e02d3a6c3148490b27e9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71661e209131cea512e5.md>)

Recorded evaluation configuration: {"Type":"BO","Params":"3.9M","Search cost":"1 GPU day"}. This record does not assert independent reproduction.

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