# Short-form open-domain QA on NQ

Source: [Short-form open-domain QA on NQ](<https://www.sota2.com/research/sota/short-form-open-domain-qa-on-nq>)

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

Data updated: 2026-04-02T08:03:25.385Z

Evaluation of Short-form open-domain Question Answering performance on the Natural Questions (NQ) dataset, measured using PIA and Poison robustness metrics (racc/asr).

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 | Short-form open-domain QA |
| Dataset (including any stated split) | NQ |
| Metric | PIA RAcc |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [PIA RAcc](<https://www.sota2.com/research/sota/short-form-open-domain-qa-on-nq.md?metric=PIA%20RAcc>): Higher is better; unit: PERCENT.

- [PIA ASR](<https://www.sota2.com/research/sota/short-form-open-domain-qa-on-nq.md?metric=PIA%20ASR>): Lower is better; unit: PERCENT.

- [Poison RAcc](<https://www.sota2.com/research/sota/short-form-open-domain-qa-on-nq.md?metric=Poison%20RAcc>): Higher is better; unit: PERCENT.

- [Poison ASR](<https://www.sota2.com/research/sota/short-form-open-domain-qa-on-nq.md?metric=Poison%20ASR>): Lower is better; unit: PERCENT.

## Recorded results

| Method | PIA RAcc | Result date | Report paper |
| --- | --- | --- | --- |
| Decoding\_c | 65 | Not recorded | 68fa73511e209131cea5bcf5 |
| Keyword | 62 | Not recorded | 68fa73511e209131cea5bcf5 |
| Keyword | 55 | Not recorded | 68fa73511e209131cea5bcf5 |
| Decoding\_c | 51 | Not recorded | 68fa73511e209131cea5bcf5 |
| Vanilla | 8 | Not recorded | 68fa73511e209131cea5bcf5 |
| Vanilla | 2 | Not recorded | 68fa73511e209131cea5bcf5 |

### Decoding\_c

PIA RAcc: 65. Result ID: 69ce22ba47f34ce924b730f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Mistral-7B"}. This record does not assert independent reproduction.

### Keyword

PIA RAcc: 62. Result ID: 69ce22ba47f34ce924b730f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Mistral-7B"}. This record does not assert independent reproduction.

### Keyword

PIA RAcc: 55. Result ID: 69ce22ba47f34ce924b730f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Llama2-7B"}. This record does not assert independent reproduction.

### Decoding\_c

PIA RAcc: 51. Result ID: 69ce22ba47f34ce924b730f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Llama2-7B"}. This record does not assert independent reproduction.

### Vanilla

PIA RAcc: 8. Result ID: 69ce22ba47f34ce924b730f4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Mistral-7B"}. This record does not assert independent reproduction.

### Vanilla

PIA RAcc: 2. Result ID: 69ce22ba47f34ce924b730f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73511e209131cea5bcf5.md>)

Recorded evaluation configuration: {"LLM":"Llama2-7B"}. This record does not assert independent reproduction.

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