# Text Ranking on MS MARCO In-domain suite (TREC DL19, TREC DL20) v1 (dev test)

Source: [Text Ranking on MS MARCO In-domain suite (TREC DL19, TREC DL20) v1 (dev test)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test>)

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

Data updated: 2026-02-26T05:53:07.248Z

Evaluation of Text Ranking performance across the MS MARCO In-domain suite, including TREC DL19 and DL20 datasets (v1 dev test split), comparing sparse (BM25) and dense (BGE) retrieval methods using NDCG@10.

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 | Text Ranking |
| Dataset (including any stated split) | MS MARCO In-domain suite (TREC DL19, TREC DL20) v1 (dev test) |
| Metric | NDCG@10 (Sparse, BM25, MS MARCO) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 13 of 13 |
| 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

- [NDCG@10 (Sparse, BM25, MS MARCO)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Sparse%2C%20BM25%2C%20MS%20MARCO)>): Higher is better; unit: SCALAR.

- [NDCG@10 (Sparse, BM25, DL19)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Sparse%2C%20BM25%2C%20DL19)>): Higher is better; unit: SCALAR.

- [NDCG@10 (Sparse, BM25, DL20)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Sparse%2C%20BM25%2C%20DL20)>): Higher is better; unit: SCALAR.

- [NDCG@10 (Dense, BGE, MS MARCO)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Dense%2C%20BGE%2C%20MS%20MARCO)>): Higher is better; unit: SCALAR.

- [NDCG@10 (Dense, BGE, DL19)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Dense%2C%20BGE%2C%20DL19)>): Higher is better; unit: SCALAR.

- [NDCG@10 (Dense, BGE, DL20)](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=NDCG%4010%20(Dense%2C%20BGE%2C%20DL20)>): Higher is better; unit: SCALAR.

- [Average NDCG@10](<https://www.sota2.com/research/sota/text-ranking-on-ms-marco-in-domain-suite-trec-dl19-trec-dl20-v1-dev-test.md?metric=Average%20NDCG%4010>): Higher is better; unit: SCALAR.

## Recorded results

| Method | NDCG@10 (Sparse, BM25, MS MARCO) | Result date | Report paper |
| --- | --- | --- | --- |
| TSARankLLM | 0.48 | Not recorded | 68fa73651e209131cea5c2ea |
| RankLLaMA | 0.469 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.466 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.466 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.46 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.451 | Not recorded | 68fa73651e209131cea5c2ea |
| MonoT5 | 0.449 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.445 | Not recorded | 68fa73651e209131cea5c2ea |
| MonoBERT | 0.44 | Not recorded | 68fa73651e209131cea5c2ea |
| TSARankLLM | 0.44 | Not recorded | 68fa73651e209131cea5c2ea |
| MonoT5 | 0.436 | Not recorded | 68fa73651e209131cea5c2ea |
| MonoT5 | 0.434 | Not recorded | 68fa73651e209131cea5c2ea |
| Retrieval | 0.228 | Not recorded | 68fa73651e209131cea5c2ea |

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.48. Result ID: 6992d8ba3a6c31484911e655.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"Qwen","Size":"7B"}. This record does not assert independent reproduction.

### RankLLaMA

NDCG@10 (Sparse, BM25, MS MARCO): 0.469. Result ID: 6992d8ba3a6c31484911e64e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70db1e209131cea4dba5.md>)

Recorded evaluation configuration: {"LLM":"LLAMA","Size":"7B"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.466. Result ID: 6992d8ba3a6c31484911e653.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"LLAMA","Size":"7B"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.466. Result ID: 6992d8ba3a6c31484911e654.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"Baichuan","Size":"7B"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.46. Result ID: 6992d8ba3a6c31484911e652.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"BLOOM","Size":"7B"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.451. Result ID: 6992d8ba3a6c31484911e651.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"BLOOM","Size":"3B"}. This record does not assert independent reproduction.

### MonoT5

NDCG@10 (Sparse, BM25, MS MARCO): 0.449. Result ID: 6992d8ba3a6c31484911e64d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67341e209131cea14b55.md>)

[Implementation](<http://anserini.io/>)

Recorded evaluation configuration: {"LLM":"T5","Size":"3B"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.445. Result ID: 6992d8ba3a6c31484911e650.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"BLOOM","Size":"1B"}. This record does not assert independent reproduction.

### MonoBERT

NDCG@10 (Sparse, BM25, MS MARCO): 0.44. Result ID: 6992d8ba3a6c31484911e64a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114a6.md>)

Recorded evaluation configuration: {"LLM":"BERT","Size":"340M"}. This record does not assert independent reproduction.

### TSARankLLM

NDCG@10 (Sparse, BM25, MS MARCO): 0.44. Result ID: 6992d8ba3a6c31484911e64f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Implementation](<https://github.com/Alibaba-NLP/RankingGPT>)

Recorded evaluation configuration: {"LLM":"BLOOM","Size":"560M"}. This record does not assert independent reproduction.

### MonoT5

NDCG@10 (Sparse, BM25, MS MARCO): 0.436. Result ID: 6992d8ba3a6c31484911e64b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67341e209131cea14b55.md>)

[Implementation](<http://anserini.io/>)

Recorded evaluation configuration: {"LLM":"T5","Size":"220M"}. This record does not assert independent reproduction.

### MonoT5

NDCG@10 (Sparse, BM25, MS MARCO): 0.434. Result ID: 6992d8ba3a6c31484911e64c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67341e209131cea14b55.md>)

[Implementation](<http://anserini.io/>)

Recorded evaluation configuration: {"LLM":"T5","Size":"770M"}. This record does not assert independent reproduction.

### Retrieval

NDCG@10 (Sparse, BM25, MS MARCO): 0.228. Result ID: 6992d8ba3a6c31484911e649.

[Reporting paper](<https://www.sota2.com/research/paper/68fa73651e209131cea5c2ea.md>)

Recorded evaluation configuration: {"LLM":"NA","Size":"NA"}. This record does not assert independent reproduction.

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