# Pixel Attribution Quality Evaluation on CIFAR10

Source: [Pixel Attribution Quality Evaluation on CIFAR10](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10>)

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

Data updated: 2026-06-16T08:48:10.157Z

Evaluation of model attribution methods on the CIFAR-10 dataset using specific quality metrics focusing on Belief, Vacuity, and Dissonance (MoRF, LeRF, AOPC, AUPC).

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 | Pixel Attribution Quality Evaluation |
| Dataset (including any stated split) | CIFAR10 |
| Metric | Belief MoRF |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [Belief MoRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Belief%20MoRF>): Higher is better; unit: SCALAR.

- [Belief LeRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Belief%20LeRF>): Higher is better; unit: SCALAR.

- [Belief ΔAOPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Belief%20%CE%94AOPC>): Higher is better; unit: SCALAR.

- [Belief ΔAUPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Belief%20%CE%94AUPC>): Lower is better; unit: SCALAR.

- [Vacuity MoRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Vacuity%20MoRF>): Higher is better; unit: SCALAR.

- [Vacuity LeRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Vacuity%20LeRF>): Higher is better; unit: SCALAR.

- [Vacuity ΔAOPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Vacuity%20%CE%94AOPC>): Lower is better; unit: SCALAR.

- [Vacuity ΔAUPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Vacuity%20%CE%94AUPC>): Higher is better; unit: SCALAR.

- [Dissonance MoRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Dissonance%20MoRF>): Higher is better; unit: SCALAR.

- [Dissonance LeRF](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Dissonance%20LeRF>): Higher is better; unit: SCALAR.

- [Dissonance ΔAOPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Dissonance%20%CE%94AOPC>): Higher is better; unit: SCALAR.

- [Dissonance ΔAUPC](<https://www.sota2.com/research/sota/pixel-attribution-quality-evaluation-on-cifar10.md?metric=Dissonance%20%CE%94AUPC>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Belief MoRF | Result date | Report paper |
| --- | --- | --- | --- |
| Integrated Gradients | 0.76 | Not recorded | 6a30d8a1d33dfb23e783cca8 |
| SHAP | 0.76 | Not recorded | 6a30d8a1d33dfb23e783cca8 |
| FullGrad | 0.67 | Not recorded | 6a30d8a1d33dfb23e783cca8 |

### Integrated Gradients

Belief MoRF: 0.76. Result ID: 6a310db5fbc373592c2d6db0.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d8a1d33dfb23e783cca8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63661e209131ce9fe6bd.md>)

Recorded evaluation configuration: {"Method":"IG"}. This record does not assert independent reproduction.

### SHAP

Belief MoRF: 0.76. Result ID: 6a310db5fbc373592c2d6db1.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d8a1d33dfb23e783cca8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63d01e209131cea00e17.md>)

Recorded evaluation configuration: {"Method":"SHAP"}. This record does not assert independent reproduction.

### FullGrad

Belief MoRF: 0.67. Result ID: 6a310db5fbc373592c2d6daf.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d8a1d33dfb23e783cca8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66cb1e209131cea12581.md>)

Recorded evaluation configuration: {"Method":"FG"}. This record does not assert independent reproduction.

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