# Downstream Generation and Ranking Alignment on MKQA (test)

Source: [Downstream Generation and Ranking Alignment on MKQA (test)](<https://www.sota2.com/research/sota/downstream-generation-and-ranking-alignment-on-mkqa-test>)

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

Data updated: 2026-04-23T10:22:30.781Z

Evaluation of model performance on the test split of the MKQA dataset, focusing on downstream generation quality and ranking alignment, measured by metrics like 3-gram recall and Pearson correlation.

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 | Downstream Generation and Ranking Alignment |
| Dataset (including any stated split) | MKQA (test) |
| Metric | 3-gram Recall |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [3-gram Recall](<https://www.sota2.com/research/sota/downstream-generation-and-ranking-alignment-on-mkqa-test.md?metric=3-gram%20Recall>): Higher is better; unit: PERCENT.

- [Pearson Correlation](<https://www.sota2.com/research/sota/downstream-generation-and-ranking-alignment-on-mkqa-test.md?metric=Pearson%20Correlation>): Higher is better; unit: SCALAR.

## Recorded results

| Method | 3-gram Recall | Result date | Report paper |
| --- | --- | --- | --- |
| BGE-Reranker + LAURA | 49.9 | Not recorded | 69e9a7ad55a3c32af89f717c |
| Qwen-Reranker + LAURA | 49.2 | Not recorded | 69e9a7ad55a3c32af89f717c |
| BGE-Reranker | 48.9 | Not recorded | 69e9a7ad55a3c32af89f717c |
| BGE-Reranker + LAURA | 47.7 | Not recorded | 69e9a7ad55a3c32af89f717c |
| Qwen-Reranker | 47.1 | Not recorded | 69e9a7ad55a3c32af89f717c |
| BGE-Reranker | 46.7 | Not recorded | 69e9a7ad55a3c32af89f717c |
| Qwen-Reranker + LAURA | 46.7 | Not recorded | 69e9a7ad55a3c32af89f717c |
| Qwen-Reranker | 44.9 | Not recorded | 69e9a7ad55a3c32af89f717c |

### BGE-Reranker + LAURA

3-gram Recall: 49.9. Result ID: 69e9f2d047f34ce924d1c30a.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

Recorded evaluation configuration: {"Generator Model":"Llama","Top-k candidates":"5"}. This record does not assert independent reproduction.

### Qwen-Reranker + LAURA

3-gram Recall: 49.2. Result ID: 69e9f2d047f34ce924d1c30e.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

Recorded evaluation configuration: {"Generator Model":"Llama","Top-k candidates":"5"}. This record does not assert independent reproduction.

### BGE-Reranker

3-gram Recall: 48.9. Result ID: 69e9f2d047f34ce924d1c309.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73d91e209131cea5e7ab.md>)

[Implementation](<https://github.com/FlagOpen/FlagEmbedding>)

Recorded evaluation configuration: {"Generator Model":"Llama","Top-k candidates":"5"}. This record does not assert independent reproduction.

### BGE-Reranker + LAURA

3-gram Recall: 47.7. Result ID: 69e9f2d047f34ce924d1c30c.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

Recorded evaluation configuration: {"Generator Model":"Qwen","Top-k candidates":"5"}. This record does not assert independent reproduction.

### Qwen-Reranker

3-gram Recall: 47.1. Result ID: 69e9f2d047f34ce924d1c30d.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79601e209131cea7d516.md>)

Recorded evaluation configuration: {"Generator Model":"Llama","Top-k candidates":"5"}. This record does not assert independent reproduction.

### BGE-Reranker

3-gram Recall: 46.7. Result ID: 69e9f2d047f34ce924d1c30b.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73d91e209131cea5e7ab.md>)

[Implementation](<https://github.com/FlagOpen/FlagEmbedding>)

Recorded evaluation configuration: {"Generator Model":"Qwen","Top-k candidates":"5"}. This record does not assert independent reproduction.

### Qwen-Reranker + LAURA

3-gram Recall: 46.7. Result ID: 69e9f2d047f34ce924d1c310.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

Recorded evaluation configuration: {"Generator Model":"Qwen","Top-k candidates":"5"}. This record does not assert independent reproduction.

### Qwen-Reranker

3-gram Recall: 44.9. Result ID: 69e9f2d047f34ce924d1c30f.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a7ad55a3c32af89f717c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79601e209131cea7d516.md>)

Recorded evaluation configuration: {"Generator Model":"Qwen","Top-k candidates":"5"}. This record does not assert independent reproduction.

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