# Multi-hop Question Answering on MuSiQue (SubEM, LLM-Acc)

Source: [Multi-hop Question Answering on MuSiQue (SubEM, LLM-Acc)](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-musique-subem-llm-acc>)

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

Data updated: 2026-04-21T10:27:42.708Z

Evaluation of model performance on the MuSiQue dataset for Multi-hop Question Answering, measured using SubEM and LLM-Acc.

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 Question Answering |
| Dataset (including any stated split) | MuSiQue |
| Metric | SubEM |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 17 of 17 |
| 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

- [SubEM](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-musique-subem-llm-acc.md?metric=SubEM>): Higher is better; unit: PERCENT.

- [LLM-Acc](<https://www.sota2.com/research/sota/multi-hop-question-answering-on-musique-subem-llm-acc.md?metric=LLM-Acc>): Higher is better; unit: PERCENT.

## Recorded results

| Method | SubEM | Result date | Report paper |
| --- | --- | --- | --- |
| EHRAG | 34.3 | Not recorded | 69e704a47d0172140ec7ede6 |
| LinearRAG | 33.9 | Not recorded | 69e704a47d0172140ec7ede6 |
| HippoRAG2 | 31 | Not recorded | 69e704a47d0172140ec7ede6 |
| GFM-RAG | 29.9 | Not recorded | 69e704a47d0172140ec7ede6 |
| HippoRAG | 29.3 | Not recorded | 69e704a47d0172140ec7ede6 |
| LightRAG | 27.4 | Not recorded | 69e704a47d0172140ec7ede6 |
| Vanilla RAG | 26.1 | Not recorded | 69e704a47d0172140ec7ede6 |
| KGP | 25.6 | Not recorded | 69e704a47d0172140ec7ede6 |
| GraphRAG | 24.3 | Not recorded | 69e704a47d0172140ec7ede6 |
| E2GraphRAG | 23.8 | Not recorded | 69e704a47d0172140ec7ede6 |
| RAPTOR | 23.3 | Not recorded | 69e704a47d0172140ec7ede6 |
| G-retriever | 14.4 | Not recorded | 69e704a47d0172140ec7ede6 |
| GPT-4o-mini | 13.6 | Not recorded | 69e704a47d0172140ec7ede6 |
| GPT-3.5-turbo | 10.3 | Not recorded | 69e704a47d0172140ec7ede6 |
| llama-8B | 7.4 | Not recorded | 69e704a47d0172140ec7ede6 |
| Qwen3-8B | 6.5 | Not recorded | 69e704a47d0172140ec7ede6 |
| llama-13B | 3.3 | Not recorded | 69e704a47d0172140ec7ede6 |

### EHRAG

SubEM: 34.3. Result ID: 69e7510c47f34ce924ce3de6.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### LinearRAG

SubEM: 33.9. Result ID: 69e7510c47f34ce924ce3de5.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c131e209131cea8c4cc.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### HippoRAG2

SubEM: 31. Result ID: 69e7510c47f34ce924ce3de4.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79821e209131cea7de8a.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### GFM-RAG

SubEM: 29.9. Result ID: 69e7510c47f34ce924ce3de3.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c3d1e209131cea8d1bf.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### HippoRAG

SubEM: 29.3. Result ID: 69e7510c47f34ce924ce3de2.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76981e209131cea6dff9.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### LightRAG

SubEM: 27.4. Result ID: 69e7510c47f34ce924ce3de1.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78461e209131cea778bd.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### Vanilla RAG

SubEM: 26.1. Result ID: 69e7510c47f34ce924ce3ddb.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### KGP

SubEM: 25.6. Result ID: 69e7510c47f34ce924ce3ddc.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ab1e209131cea5270e.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### GraphRAG

SubEM: 24.3. Result ID: 69e7510c47f34ce924ce3dde.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa770d1e209131cea70c5a.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### E2GraphRAG

SubEM: 23.8. Result ID: 69e7510c47f34ce924ce3de0.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79291e209131cea7c584.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### RAPTOR

SubEM: 23.3. Result ID: 69e7510c47f34ce924ce3ddf.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71fe1e209131cea53fc2.md>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### G-retriever

SubEM: 14.4. Result ID: 69e7510c47f34ce924ce3ddd.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73541e209131cea5bdef.md>)

[Implementation](<https://github.com/XiaoxinHe/G-Retriever>)

Recorded evaluation configuration: {"Generation Strategy":"Retrieval-Augmented-Generation Methods"}. This record does not assert independent reproduction.

### GPT-4o-mini

SubEM: 13.6. Result ID: 69e7510c47f34ce924ce3dda.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"Generation Strategy":"Direct Zero-shot LLM Inference"}. This record does not assert independent reproduction.

### GPT-3.5-turbo

SubEM: 10.3. Result ID: 69e7510c47f34ce924ce3dd9.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"Generation Strategy":"Direct Zero-shot LLM Inference"}. This record does not assert independent reproduction.

### llama-8B

SubEM: 7.4. Result ID: 69e7510c47f34ce924ce3dd6.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

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

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

Recorded evaluation configuration: {"Generation Strategy":"Direct Zero-shot LLM Inference"}. This record does not assert independent reproduction.

### Qwen3-8B

SubEM: 6.5. Result ID: 69e7510c47f34ce924ce3dd8.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

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

Recorded evaluation configuration: {"Generation Strategy":"Direct Zero-shot LLM Inference"}. This record does not assert independent reproduction.

### llama-13B

SubEM: 3.3. Result ID: 69e7510c47f34ce924ce3dd7.

[Reporting paper](<https://www.sota2.com/research/paper/69e704a47d0172140ec7ede6.md>)

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

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

Recorded evaluation configuration: {"Generation Strategy":"Direct Zero-shot LLM Inference"}. This record does not assert independent reproduction.

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