# Functional shortcut restoration on Waterbirds

Source: [Functional shortcut restoration on Waterbirds](<https://www.sota2.com/research/sota/functional-shortcut-restoration-on-waterbirds>)

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

Data updated: 2026-07-08T07:05:27.243Z

Evaluation of Functional shortcut restoration performance on the Waterbirds dataset, reporting Delta WGA and Delta CSR metrics.

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 | Functional shortcut restoration |
| Dataset (including any stated split) | Waterbirds |
| Metric | Delta WGA |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Delta WGA](<https://www.sota2.com/research/sota/functional-shortcut-restoration-on-waterbirds.md?metric=Delta%20WGA>): Higher is better; unit: SCALAR.

- [Delta CSR](<https://www.sota2.com/research/sota/functional-shortcut-restoration-on-waterbirds.md?metric=Delta%20CSR>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Delta WGA | Result date | Report paper |
| --- | --- | --- | --- |
| A-NegGrad+ | 28.2 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| A-SSD | 20.1 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| A-SCRUB | 19.9 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| A-SalUn | 12.8 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| DFR | 9.2 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| GroupDRO | 8.7 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| ERM | 7.3 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| JTT | 7.1 | Not recorded | 6a4dd9b44da6617fe9198c55 |
| Balanced Retrain | 3 | Not recorded | 6a4dd9b44da6617fe9198c55 |

### A-NegGrad+

Delta WGA: 28.2. Result ID: 6a4df6a860b88bb339d6bcd4.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### A-SSD

Delta WGA: 20.1. Result ID: 6a4df6a860b88bb339d6bcd7.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### A-SCRUB

Delta WGA: 19.9. Result ID: 6a4df6a860b88bb339d6bcd5.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### A-SalUn

Delta WGA: 12.8. Result ID: 6a4df6a860b88bb339d6bcd6.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### DFR

Delta WGA: 9.2. Result ID: 6a4df6a860b88bb339d6bcd2.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fe41e209131cea47f22.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### GroupDRO

Delta WGA: 8.7. Result ID: 6a4df6a860b88bb339d6bcd1.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67561e209131cea159c7.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### ERM

Delta WGA: 7.3. Result ID: 6a4df6a860b88bb339d6bccf.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### JTT

Delta WGA: 7.1. Result ID: 6a4df6a860b88bb339d6bcd3.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ade1e209131cea2aeeb.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

### Balanced Retrain

Delta WGA: 3. Result ID: 6a4df6a860b88bb339d6bcd0.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9b44da6617fe9198c55.md>)

Recorded evaluation configuration: {"Beta":"2"}. This record does not assert independent reproduction.

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