# Open-domain Question Answering on NQ Search-o1 framework (50% held-out test)

Source: [Open-domain Question Answering on NQ Search-o1 framework (50% held-out test)](<https://www.sota2.com/research/sota/open-domain-question-answering-on-nq-search-o1-framework-50-percent-held-out-test>)

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

Data updated: 2026-05-12T11:33:15.140Z

Evaluation of Open-domain Question Answering performance using the NQ dataset specifically configured under the Search-o1 framework, covering the 50% held-out test split. Metrics include Accuracy (Acc), Token Count (TC), and Change in Token Count (ΔTC).

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 | Open-domain Question Answering |
| Dataset (including any stated split) | NQ Search-o1 framework (50% held-out test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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/open-domain-question-answering-on-nq-search-o1-framework-50-percent-held-out-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Token Count](<https://www.sota2.com/research/sota/open-domain-question-answering-on-nq-search-o1-framework-50-percent-held-out-test.md?metric=Token%20Count>): Higher is better; unit: SCALAR.

- [Change in Token Count](<https://www.sota2.com/research/sota/open-domain-question-answering-on-nq-search-o1-framework-50-percent-held-out-test.md?metric=Change%20in%20Token%20Count>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Default | 44.8 | Not recorded | 6a02b461dbc836b87d824a63 |
| Best Baseline | 44.8 | Not recorded | 6a02b461dbc836b87d824a63 |
| PROBE&PREFILL | 42 | Not recorded | 6a02b461dbc836b87d824a63 |

### Default

Accuracy: 44.8. Result ID: 6a030fec47f34ce924e6e5f3.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b461dbc836b87d824a63.md>)

Recorded evaluation configuration: {"N":"250"}. This record does not assert independent reproduction.

### Best Baseline

Accuracy: 44.8. Result ID: 6a030fec47f34ce924e6e5f4.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b461dbc836b87d824a63.md>)

Recorded evaluation configuration: {"N":"250"}. This record does not assert independent reproduction.

### PROBE&PREFILL

Accuracy: 42. Result ID: 6a030fec47f34ce924e6e5f5.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b461dbc836b87d824a63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b461dbc836b87d824a63.md>)

[Implementation](<https://github.com/Trustworthy-ML-Lab/when2tool>)

Recorded evaluation configuration: {"N":"250"}. This record does not assert independent reproduction.

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