# Closed-book Question Answering on Natural Questions (dev)

Source: [Closed-book Question Answering on Natural Questions (dev)](<https://www.sota2.com/research/sota/closed-book-question-answering-on-natural-questions-dev>)

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

Data updated: 2026-02-26T05:28:39.969Z

Evaluation of Closed-book Question Answering performance on the development split of the Natural Questions dataset, measured using 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 | Closed-book question answering |
| Dataset (including any stated split) | Natural Questions (dev) |
| Metric | Accuracy |
| Unit | PERCENT |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/closed-book-question-answering-on-natural-questions-dev.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| FiD + Distillation | 54.4 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Chinchilla | 35.5 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Chinchilla | 31.5 | Not recorded | 68fa6e5c1e209131cea3f593 |
| GPT-3 | 29.9 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Gopher | 28.2 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Gopher | 24.5 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Chinchilla | 16.6 | Not recorded | 68fa6e5c1e209131cea3f593 |
| GPT-3 | 14.6 | Not recorded | 68fa6e5c1e209131cea3f593 |
| Gopher | 10.1 | Not recorded | 68fa6e5c1e209131cea3f593 |

### FiD + Distillation

Accuracy: 54.4. Result ID: 699016523a6c314849fe0a9b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d321e209131cea38511.md>)

[Implementation](<https://www.sota2.com/github.com/facebookresearch/FiD>)

Recorded evaluation configuration: {"evaluation\_protocol":"open book"}. This record does not assert independent reproduction.

### Chinchilla

Accuracy: 35.5. Result ID: 699016523a6c314849fe0a95.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

Recorded evaluation configuration: {"shots":"64-shot"}. This record does not assert independent reproduction.

### Chinchilla

Accuracy: 31.5. Result ID: 699016523a6c314849fe0a94.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

Recorded evaluation configuration: {"shots":"5-shot"}. This record does not assert independent reproduction.

### GPT-3

Accuracy: 29.9. Result ID: 699016523a6c314849fe0a9a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa690d1e209131cea1fcb0.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/test>)

Recorded evaluation configuration: {"shots":"64-shot"}. This record does not assert independent reproduction.

### Gopher

Accuracy: 28.2. Result ID: 699016523a6c314849fe0a98.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ba51e209131cea2f8bf.md>)

Recorded evaluation configuration: {"shots":"64-shot"}. This record does not assert independent reproduction.

### Gopher

Accuracy: 24.5. Result ID: 699016523a6c314849fe0a97.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ba51e209131cea2f8bf.md>)

Recorded evaluation configuration: {"shots":"5-shot"}. This record does not assert independent reproduction.

### Chinchilla

Accuracy: 16.6. Result ID: 699016523a6c314849fe0a93.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

Recorded evaluation configuration: {"shots":"0-shot"}. This record does not assert independent reproduction.

### GPT-3

Accuracy: 14.6. Result ID: 699016523a6c314849fe0a99.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa690d1e209131cea1fcb0.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/test>)

Recorded evaluation configuration: {"shots":"0-shot"}. This record does not assert independent reproduction.

### Gopher

Accuracy: 10.1. Result ID: 699016523a6c314849fe0a96.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e5c1e209131cea3f593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ba51e209131cea2f8bf.md>)

Recorded evaluation configuration: {"shots":"0-shot"}. This record does not assert independent reproduction.

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