# Single-Hop Question Answering on NQ (train)

Source: [Single-Hop Question Answering on NQ (train)](<https://www.sota2.com/research/sota/single-hop-question-answering-on-nq-train>)

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

Data updated: 2026-02-26T05:31:10.468Z

Evaluation of model performance on the training split of the Natural Questions (NQ) dataset for single-hop question answering, 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 | Single-Hop Question Answering |
| Dataset (including any stated split) | NQ (train) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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/single-hop-question-answering-on-nq-train.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Slice-Base | 32.8 | Not recorded | 695f39f053fdd5c361743ec5 |
| Qwen2.5-7B-Instruct | 13.4 | Not recorded | 695f39f053fdd5c361743ec5 |

### Slice-Base

Accuracy: 32.8. Result ID: 6996be595698665ec4095172.

[Reporting paper](<https://www.sota2.com/research/paper/695f39f053fdd5c361743ec5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39f053fdd5c361743ec5.md>)

Recorded evaluation configuration: {"Retrieval setting":"Clean retrieval","Evaluation setting":"Inference-Only"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct

Accuracy: 13.4. Result ID: 6996be595698665ec4095171.

[Reporting paper](<https://www.sota2.com/research/paper/695f39f053fdd5c361743ec5.md>)

Recorded evaluation configuration: {"Retrieval setting":"Clean retrieval","Evaluation setting":"Inference-Only"}. This record does not assert independent reproduction.

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