# Explanation quality rejection on celeba

Source: [Explanation quality rejection on celeba](<https://www.sota2.com/research/sota/explanation-quality-rejection-on-celeba>)

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

Data updated: 2026-03-19T07:42:16.829Z

Evaluation of model performance on determining the quality of explanations using the CelebA dataset, measured by AUROC.

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 | Explanation quality rejection |
| Dataset (including any stated split) | celeba |
| Metric | AUROC |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [AUROC](<https://www.sota2.com/research/sota/explanation-quality-rejection-on-celeba.md?metric=AUROC>): Higher is better; unit: SCALAR.

## Recorded results

| Method | AUROC | Result date | Report paper |
| --- | --- | --- | --- |
| REX | 0.9 | Not recorded | 68fa7a181e209131cea811bd |
| ExplNov | 0.75 | Not recorded | 68fa7a181e209131cea811bd |
| PredAmb | 0.69 | Not recorded | 68fa7a181e209131cea811bd |
| PASTARej | 0.65 | Not recorded | 68fa7a181e209131cea811bd |
| InstNov | 0.49 | Not recorded | 68fa7a181e209131cea811bd |

### REX

AUROC: 0.9. Result ID: 69bba7e947f34ce924a53033.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

Recorded evaluation configuration: {"Setting":"Simulated","u% (low-quality threshold)":"10%"}. This record does not assert independent reproduction.

### ExplNov

AUROC: 0.75. Result ID: 69bba7e947f34ce924a53036.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e2f1e209131cea3e69b.md>)

[Implementation](<https://github.com/deeplearning-wisc/knn-ood>)

Recorded evaluation configuration: {"Setting":"Simulated","u% (low-quality threshold)":"10%"}. This record does not assert independent reproduction.

### PredAmb

AUROC: 0.69. Result ID: 69bba7e947f34ce924a53035.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea557f2.md>)

Recorded evaluation configuration: {"Setting":"Simulated","u% (low-quality threshold)":"10%"}. This record does not assert independent reproduction.

### PASTARej

AUROC: 0.65. Result ID: 69bba7e947f34ce924a53037.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ae41e209131cea85b7a.md>)

Recorded evaluation configuration: {"Setting":"Simulated","u% (low-quality threshold)":"10%"}. This record does not assert independent reproduction.

### InstNov

AUROC: 0.49. Result ID: 69bba7e947f34ce924a53034.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a181e209131cea811bd.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e2f1e209131cea3e69b.md>)

[Implementation](<https://github.com/deeplearning-wisc/knn-ood>)

Recorded evaluation configuration: {"Setting":"Simulated","u% (low-quality threshold)":"10%"}. This record does not assert independent reproduction.

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