# Long Answer Question Answering on Natural Questions (NQ) (test)

Source: [Long Answer Question Answering on Natural Questions (NQ) (test)](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test>)

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

Data updated: 2026-02-26T05:58:18.315Z

Evaluation of model performance on the test split of the Natural Questions (NQ) dataset, specifically focusing on Long Answer Question Answering metrics (Precision, Recall, F1).

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 | Long Answer Question Answering |
| Dataset (including any stated split) | Natural Questions (NQ) (test) |
| Metric | Precision |
| 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

- [Precision](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test.md?metric=Precision>): Higher is better; unit: PERCENT.

- [Recall](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test.md?metric=Recall>): Higher is better; unit: PERCENT.

- [F1](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test.md?metric=F1>): Higher is better; unit: PERCENT.

- [Result Score](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test.md?metric=Result%20Score>): Higher is better; unit: PERCENT.

- [Rank](<https://www.sota2.com/research/sota/long-answer-question-answering-on-natural-questions-nq-test.md?metric=Rank>): Lower is better; unit: OTHER.

## Recorded results

| Method | Precision | Result date | Report paper |
| --- | --- | --- | --- |
| Cluster-Former | 78.5 | Not recorded | 68fa6a211e209131cea262e2 |
| RikiNet | 78.1 | Not recorded | 68fa6a211e209131cea262e2 |
| BigBird-ETC-large | 77.5 | Not recorded | 68fa6a211e209131cea262e2 |
| BERT joint | 64.1 | Not recorded | 68fa66ce1e209131cea127dc |
| DecAtt + DocReader | 54.3 | Not recorded | 68fa66ce1e209131cea127dc |
| DocumentQA | 48.9 | Not recorded | 68fa66ce1e209131cea127dc |

### Cluster-Former

Precision: 78.5. Result ID: 698e12edffc3901de09a9644.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a211e209131cea262e2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a211e209131cea262e2.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### RikiNet

Precision: 78.1. Result ID: 698e12edffc3901de09a9643.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a211e209131cea262e2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa676f1e209131cea16356.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### BigBird-ETC-large

Precision: 77.5. Result ID: 698e12edffc3901de09a9642.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a211e209131cea262e2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa690b1e209131cea1faf6.md>)

[Implementation](<http://goo.gle/bigbird-transformer>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### BERT joint

Precision: 64.1. Result ID: 698d092a8b517eb4331e8669.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66ce1e209131cea127dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66ce1e209131cea127dc.md>)

[Implementation](<https://github.com/google-research/language/tree/master/language/question_answering/bert_joint>)

Recorded evaluation configuration: {"pre-training":"SQuAD 1.1 fine-tuned"}. This record does not assert independent reproduction.

### DecAtt + DocReader

Precision: 54.3. Result ID: 698d092a8b517eb4331e8668.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66ce1e209131cea127dc.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### DocumentQA

Precision: 48.9. Result ID: 698d092a8b517eb4331e8667.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66ce1e209131cea127dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63d11e209131cea00e26.md>)

[Implementation](<https://www.sota2.com/github.com/allenai/document-qa>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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