# Document Retrieval on modified Natural Questions

Source: [Document Retrieval on modified Natural Questions](<https://www.sota2.com/research/sota/document-retrieval-on-modified-natural-questions>)

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

Data updated: 2026-02-26T05:50:20.991Z

Evaluation of model performance on the modified version of the Natural Questions dataset for Document Retrieval, measured by MRR and Recall@K metrics.

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 | Document Retrieval |
| Dataset (including any stated split) | modified Natural Questions |
| Metric | MRR |
| 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

- [MRR](<https://www.sota2.com/research/sota/document-retrieval-on-modified-natural-questions.md?metric=MRR>): Higher is better; unit: PERCENT.

- [Recall@1](<https://www.sota2.com/research/sota/document-retrieval-on-modified-natural-questions.md?metric=Recall%401>): Higher is better; unit: PERCENT.

- [Recall@2](<https://www.sota2.com/research/sota/document-retrieval-on-modified-natural-questions.md?metric=Recall%402>): Higher is better; unit: PERCENT.

## Recorded results

| Method | MRR | Result date | Report paper |
| --- | --- | --- | --- |
| PDRMM+PDRMM | 40.33 | Not recorded | 68fa6a261e209131cea265e3 |
| JPDRMM | 36.5 | Not recorded | 68fa6a261e209131cea265e3 |
| BM25+BM25 | 30.18 | Not recorded | 68fa6a261e209131cea265e3 |

### PDRMM+PDRMM

MRR: 40.33. Result ID: 698e1732ffc3901de09ab8fe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a261e209131cea265e3.md>)

Recorded evaluation configuration: {"Architecture":"Pipeline"}. This record does not assert independent reproduction.

### JPDRMM

MRR: 36.5. Result ID: 698e1732ffc3901de09ab8ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a261e209131cea265e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a261e209131cea265e3.md>)

[Implementation](<http://nlp.cs.aueb.gr/publications.html>)

Recorded evaluation configuration: {"Architecture":"Joint"}. This record does not assert independent reproduction.

### BM25+BM25

MRR: 30.18. Result ID: 698e1732ffc3901de09ab8fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a261e209131cea265e3.md>)

Recorded evaluation configuration: {"Architecture":"Pipeline"}. This record does not assert independent reproduction.

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