# Document Reranking on TREC DL (DL19 and DL20)

Source: [Document Reranking on TREC DL (DL19 and DL20)](<https://www.sota2.com/research/sota/document-reranking-on-trec-dl-dl19-and-dl20>)

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

Data updated: 2026-02-26T05:53:07.234Z

Evaluation of Document Reranking performance across the TREC DL 2019 and TREC DL 2020 tracks, measured 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 | Document Reranking |
| Dataset (including any stated split) | TREC DL |
| Metric | NDCG@10 (DL19) |
| Unit | PERCENT |
| 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 (DL19)](<https://www.sota2.com/research/sota/document-reranking-on-trec-dl-dl19-and-dl20.md?metric=NDCG%4010%20(DL19)>): Higher is better; unit: PERCENT.

- [NDCG@10 (DL20)](<https://www.sota2.com/research/sota/document-reranking-on-trec-dl-dl19-and-dl20.md?metric=NDCG%4010%20(DL20)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | NDCG@10 (DL19) | Result date | Report paper |
| --- | --- | --- | --- |
| Llama-3.1-8B Listwise | 74.79 | Not recorded | 6992a4b39ebc499c97406d75 |
| Qwen3-8B Pointwise | 73.51 | Not recorded | 6992a4b39ebc499c97406d75 |
| DiffuRank Perm-Assign | 73.44 | Not recorded | 6992a4b39ebc499c97406d75 |
| RankZephyr-7B Listwise | 73.39 | Not recorded | 6992a4b39ebc499c97406d75 |
| Qwen3-8B Listwise | 73.15 | Not recorded | 6992a4b39ebc499c97406d75 |
| DiffuRank Pointwise | 73.06 | Not recorded | 6992a4b39ebc499c97406d75 |
| DiffuRank Logits-List | 72.95 | Not recorded | 6992a4b39ebc499c97406d75 |
| DiffuRank Perm-Samp | 72.92 | Not recorded | 6992a4b39ebc499c97406d75 |
| Llama-3.1-8B Pointwise | 72.55 | Not recorded | 6992a4b39ebc499c97406d75 |
| monoT5 (220M) | 71.8 | Not recorded | 6992a4b39ebc499c97406d75 |
| monoBERT (340M) | 70.5 | Not recorded | 6992a4b39ebc499c97406d75 |
| LLaDA-1.5-8B Listwise | 69.22 | Not recorded | 6992a4b39ebc499c97406d75 |
| BM25 | 50.58 | Not recorded | 6992a4b39ebc499c97406d75 |

### Llama-3.1-8B Listwise

NDCG@10 (DL19): 74.79. Result ID: 699a57de5698665ec4df8987.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"Llama-3.1-8B","Reranking Strategy":"Listwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### Qwen3-8B Pointwise

NDCG@10 (DL19): 73.51. Result ID: 699a57de5698665ec4df8988.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"Qwen3-8B","Reranking Strategy":"Pointwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### DiffuRank Perm-Assign

NDCG@10 (DL19): 73.44. Result ID: 699a57de5698665ec4df898e.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"LLaDA-1.5-8B","Reranking Strategy":"Perm-Assign","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### RankZephyr-7B Listwise

NDCG@10 (DL19): 73.39. Result ID: 699a57de5698665ec4df8985.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71591e209131cea50c0c.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"Zephyr-7B","Reranking Strategy":"Listwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### Qwen3-8B Listwise

NDCG@10 (DL19): 73.15. Result ID: 699a57de5698665ec4df8989.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"Qwen3-8B","Reranking Strategy":"Listwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### DiffuRank Pointwise

NDCG@10 (DL19): 73.06. Result ID: 699a57de5698665ec4df898b.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"LLaDA-1.5-8B","Reranking Strategy":"Pointwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### DiffuRank Logits-List

NDCG@10 (DL19): 72.95. Result ID: 699a57de5698665ec4df898c.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"LLaDA-1.5-8B","Reranking Strategy":"Logits-List","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### DiffuRank Perm-Samp

NDCG@10 (DL19): 72.92. Result ID: 699a57de5698665ec4df898d.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"LLaDA-1.5-8B","Reranking Strategy":"Perm-Samp","Sampling Steps (K)":4,"Reranking depth":"top 100"}. This record does not assert independent reproduction.

### Llama-3.1-8B Pointwise

NDCG@10 (DL19): 72.55. Result ID: 699a57de5698665ec4df8986.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"Llama-3.1-8B","Reranking Strategy":"Pointwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### monoT5 (220M)

NDCG@10 (DL19): 71.8. Result ID: 699a57de5698665ec4df8984.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

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

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

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"T5","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### monoBERT (340M)

NDCG@10 (DL19): 70.5. Result ID: 699a57de5698665ec4df8983.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

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

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"BERT","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### LLaDA-1.5-8B Listwise

NDCG@10 (DL19): 69.22. Result ID: 699a57de5698665ec4df898a.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c0e1e209131cea8c28e.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"MS MARCO","Backbone":"LLaDA-1.5-8B","Reranking Strategy":"Listwise","Reranking depth":"top 100"}. This record does not assert independent reproduction.

### BM25

NDCG@10 (DL19): 50.58. Result ID: 699a57de5698665ec4df8982.

[Reporting paper](<https://www.sota2.com/research/paper/6992a4b39ebc499c97406d75.md>)

Recorded evaluation configuration: {"Type":"Unsupervised / Initial Retrieval"}. This record does not assert independent reproduction.

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