# Multi-hop QA Retrieval on HotpotQA

Source: [Multi-hop QA Retrieval on HotpotQA](<https://www.sota2.com/research/sota/multi-hop-qa-retrieval-on-hotpotqa>)

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

Data updated: 2026-03-27T16:49:04.229Z

Evaluation of multi-hop question answering systems on the HotpotQA dataset, measuring both the precision of retrieved passages (PR@K) and the accuracy of the final answer (QA Accuracy).

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 | Multi-hop QA Retrieval |
| Dataset (including any stated split) | HotpotQA |
| Metric | Precision@5 (%) |
| 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

- [Precision@5 (%)](<https://www.sota2.com/research/sota/multi-hop-qa-retrieval-on-hotpotqa.md?metric=Precision%405%20(%25)>): Higher is better; unit: PERCENT.

- [Precision@10 (%)](<https://www.sota2.com/research/sota/multi-hop-qa-retrieval-on-hotpotqa.md?metric=Precision%4010%20(%25)>): Higher is better; unit: PERCENT.

- [QA Accuracy](<https://www.sota2.com/research/sota/multi-hop-qa-retrieval-on-hotpotqa.md?metric=QA%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Precision@5 (%) | Result date | Report paper |
| --- | --- | --- | --- |
| GraphER-GCS | 78.8 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| Llama-Embed-Nemotron-8B+BM25 | 78.5 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| GraphER-GAT | 78 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| GraphER-GAT | 74.7 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| GraphER-GCS | 74.4 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| Cohere-Embed-V4+BM25 | 74.2 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| GraphER-PPR | 72.9 | Not recorded | 69c60eed0636e5bf4afc83e0 |
| GraphER-PPR | 71.6 | Not recorded | 69c60eed0636e5bf4afc83e0 |

### GraphER-GCS

Precision@5 (%): 78.8. Result ID: 69c6b4ab47f34ce924b11b55.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Llama-Embed-Nemotron-8B+BM25 + GraphER-GCS"}. This record does not assert independent reproduction.

### Llama-Embed-Nemotron-8B+BM25

Precision@5 (%): 78.5. Result ID: 69c6b4ab47f34ce924b11b53.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6912c3262828b7cc6937de1c.md>)

Recorded evaluation configuration: {"Retriever":"Llama-Embed-Nemotron-8B+BM25"}. This record does not assert independent reproduction.

### GraphER-GAT

Precision@5 (%): 78. Result ID: 69c6b4ab47f34ce924b11b56.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Llama-Embed-Nemotron-8B+BM25 + GraphER-GAT"}. This record does not assert independent reproduction.

### GraphER-GAT

Precision@5 (%): 74.7. Result ID: 69c6b4ab47f34ce924b11b5a.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Cohere-Embed-V4+BM25 + GraphER-GAT"}. This record does not assert independent reproduction.

### GraphER-GCS

Precision@5 (%): 74.4. Result ID: 69c6b4ab47f34ce924b11b59.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Cohere-Embed-V4+BM25 + GraphER-GCS"}. This record does not assert independent reproduction.

### Cohere-Embed-V4+BM25

Precision@5 (%): 74.2. Result ID: 69c6b4ab47f34ce924b11b57.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Cohere-Embed-V4+BM25"}. This record does not assert independent reproduction.

### GraphER-PPR

Precision@5 (%): 72.9. Result ID: 69c6b4ab47f34ce924b11b54.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Llama-Embed-Nemotron-8B+BM25 + GraphER-PPR"}. This record does not assert independent reproduction.

### GraphER-PPR

Precision@5 (%): 71.6. Result ID: 69c6b4ab47f34ce924b11b58.

[Reporting paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60eed0636e5bf4afc83e0.md>)

Recorded evaluation configuration: {"Retriever":"Cohere-Embed-V4+BM25 + GraphER-PPR"}. This record does not assert independent reproduction.

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