# Pronoun Resolution on WinoGrande (test)

Source: [Pronoun Resolution on WinoGrande (test)](<https://www.sota2.com/research/sota/pronoun-resolution-on-winogrande-test>)

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

Data updated: 2026-02-26T06:04:23.507Z

Evaluation of model performance on resolving pronoun ambiguities using the test split of the WinoGrande dataset, measured by Single and Group accuracy.

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 | Pronoun Resolution |
| Dataset (including any stated split) | WinoGrande (test) |
| Metric | Single Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 7 of 7 |
| 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

- [Single Accuracy](<https://www.sota2.com/research/sota/pronoun-resolution-on-winogrande-test.md?metric=Single%20Accuracy>): Higher is better; unit: PERCENT.

- [Group Accuracy](<https://www.sota2.com/research/sota/pronoun-resolution-on-winogrande-test.md?metric=Group%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Single Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| ALBERT-xxlarge | 58.68 | Not recorded | 68fa6b131e209131cea2bf75 |
| RoBERTa-base | 56.25 | Not recorded | 68fa6b131e209131cea2bf75 |
| BERT-large | 55.56 | Not recorded | 68fa6b131e209131cea2bf75 |
| RoBERTa-large | 54.86 | Not recorded | 68fa6b131e209131cea2bf75 |
| BERT-base | 53.12 | Not recorded | 68fa6b131e209131cea2bf75 |
| ALBERT-base | 52.78 | Not recorded | 68fa6b131e209131cea2bf75 |
| random | 50 | Not recorded | 68fa6b131e209131cea2bf75 |

### ALBERT-xxlarge

Single Accuracy: 58.68. Result ID: 698e5296ffc3901de09c9db5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67061e209131cea13e88.md>)

[Implementation](<https://github.com/google-research/ALBERT>)

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### RoBERTa-base

Single Accuracy: 56.25. Result ID: 698e5296ffc3901de09c9db2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66131e209131cea0e18b.md>)

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

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### BERT-large

Single Accuracy: 55.56. Result ID: 698e5296ffc3901de09c9db1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c31e209131cea0c4ed.md>)

[Implementation](<https://github.com/google-research/bert>)

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### RoBERTa-large

Single Accuracy: 54.86. Result ID: 698e5296ffc3901de09c9db3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66131e209131cea0e18b.md>)

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

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### BERT-base

Single Accuracy: 53.12. Result ID: 698e5296ffc3901de09c9db0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c31e209131cea0c4ed.md>)

[Implementation](<https://github.com/google-research/bert>)

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### ALBERT-base

Single Accuracy: 52.78. Result ID: 698e5296ffc3901de09c9db4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67061e209131cea13e88.md>)

[Implementation](<https://github.com/google-research/ALBERT>)

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

### random

Single Accuracy: 50. Result ID: 698e5296ffc3901de09c9daf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b131e209131cea2bf75.md>)

Recorded evaluation configuration: {"Setup":"Zero-shot"}. This record does not assert independent reproduction.

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