# Question Answering on TriviaQA (EM(%), Lat.(ms))

Source: [Question Answering on TriviaQA (EM(%), Lat.(ms))](<https://www.sota2.com/research/sota/question-answering-on-triviaqa-em-percent-lat-ms>)

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

Data updated: 2026-05-08T10:42:46.604Z

Evaluation of Question Answering performance on the TriviaQA dataset, reporting Exact Match accuracy (as a percentage) and inference latency.

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 | Question Answering |
| Dataset (including any stated split) | TriviaQA |
| Metric | EM (%) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 27 of 27 |
| 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

- [EM (%)](<https://www.sota2.com/research/sota/question-answering-on-triviaqa-em-percent-lat-ms.md?metric=EM%20(%25)>): Higher is better; unit: PERCENT.

- [Latency (ms)](<https://www.sota2.com/research/sota/question-answering-on-triviaqa-em-percent-lat-ms.md?metric=Latency%20(ms)>): Lower is better; unit: TIME.

## Recorded results

| Method | EM (%) | Result date | Report paper |
| --- | --- | --- | --- |
| AutoRefine△ | 65.3 | Not recorded | 69fd6e205497b3b670c894ce |
| AutoRefine△ | 64.73 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-R1♢ | 64.59 | Not recorded | 69fd6e205497b3b670c894ce |
| AutoRefine△ | 64.32 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-R1♢ | 63.16 | Not recorded | 69fd6e205497b3b670c894ce |
| AutoRefine△ | 63.09 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-R1♢ | 62.8 | Not recorded | 69fd6e205497b3b670c894ce |
| AutoRefine△ | 62.53 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG△ | 62.32 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-R1♢ | 61.7 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG△ | 61.54 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG♢ | 61.3 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG♢ | 61.25 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-R1♢ | 61.04 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG△ | 60.98 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG♢ | 60.4 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG△ | 60.18 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG♢ | 59.85 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG△ | 58.83 | Not recorded | 69fd6e205497b3b670c894ce |
| LatentRAG♢ | 58.56 | Not recorded | 69fd6e205497b3b670c894ce |
| Iter-RetGen | 56.91 | Not recorded | 69fd6e205497b3b670c894ce |
| R1-Searcher | 56.1 | Not recorded | 69fd6e205497b3b670c894ce |
| DeepRAG | 55.48 | Not recorded | 69fd6e205497b3b670c894ce |
| ZeroSearch | 54.67 | Not recorded | 69fd6e205497b3b670c894ce |
| Naive RAG | 53.32 | Not recorded | 69fd6e205497b3b670c894ce |
| Search-o1 | 44.53 | Not recorded | 69fd6e205497b3b670c894ce |
| Direct Infer | 42.48 | Not recorded | 69fd6e205497b3b670c894ce |

### AutoRefine△

EM (%): 65.3. Result ID: 69fdbe1d47f34ce924e16e27.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88288.md>)

Recorded evaluation configuration: {"Retrieval Model":"e5-base-v2"}. This record does not assert independent reproduction.

### AutoRefine△

EM (%): 64.73. Result ID: 69fdbe1d47f34ce924e16e2b.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88288.md>)

Recorded evaluation configuration: {"Retrieval Model":"jina-embeddings-v5-text-nano"}. This record does not assert independent reproduction.

### Search-R1♢

EM (%): 64.59. Result ID: 69fdbe1d47f34ce924e16e25.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a651e209131cea82e7b.md>)

Recorded evaluation configuration: {"Retrieval Model":"e5-base-v2"}. This record does not assert independent reproduction.

### AutoRefine△

EM (%): 64.32. Result ID: 69fdbe1d47f34ce924e16e2f.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88288.md>)

Recorded evaluation configuration: {"Retrieval Model":"harrier-oss-v1-270m"}. This record does not assert independent reproduction.

### Search-R1♢

EM (%): 63.16. Result ID: 69fdbe1d47f34ce924e16e29.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a651e209131cea82e7b.md>)

Recorded evaluation configuration: {"Retrieval Model":"jina-embeddings-v5-text-nano"}. This record does not assert independent reproduction.

### AutoRefine△

EM (%): 63.09. Result ID: 69fdbe1d47f34ce924e16e23.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88288.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### Search-R1♢

EM (%): 62.8. Result ID: 69fdbe1d47f34ce924e16e2d.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a651e209131cea82e7b.md>)

Recorded evaluation configuration: {"Retrieval Model":"harrier-oss-v1-270m"}. This record does not assert independent reproduction.

### AutoRefine△

EM (%): 62.53. Result ID: 69fdbe1d47f34ce924e16e33.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b641e209131cea88288.md>)

Recorded evaluation configuration: {"Retrieval Model":"F2LLM-v2-330M"}. This record does not assert independent reproduction.

### LatentRAG△

EM (%): 62.32. Result ID: 69fdbe1d47f34ce924e16e28.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"e5-base-v2"}. This record does not assert independent reproduction.

### Search-R1♢

EM (%): 61.7. Result ID: 69fdbe1d47f34ce924e16e21.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a651e209131cea82e7b.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### LatentRAG△

EM (%): 61.54. Result ID: 69fdbe1d47f34ce924e16e2c.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"jina-embeddings-v5-text-nano"}. This record does not assert independent reproduction.

### LatentRAG♢

EM (%): 61.3. Result ID: 69fdbe1d47f34ce924e16e26.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"e5-base-v2"}. This record does not assert independent reproduction.

### LatentRAG♢

EM (%): 61.25. Result ID: 69fdbe1d47f34ce924e16e2a.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"jina-embeddings-v5-text-nano"}. This record does not assert independent reproduction.

### Search-R1♢

EM (%): 61.04. Result ID: 69fdbe1d47f34ce924e16e31.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a651e209131cea82e7b.md>)

Recorded evaluation configuration: {"Retrieval Model":"F2LLM-v2-330M"}. This record does not assert independent reproduction.

### LatentRAG△

EM (%): 60.98. Result ID: 69fdbe1d47f34ce924e16e30.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"harrier-oss-v1-270m"}. This record does not assert independent reproduction.

### LatentRAG♢

EM (%): 60.4. Result ID: 69fdbe1d47f34ce924e16e2e.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"harrier-oss-v1-270m"}. This record does not assert independent reproduction.

### LatentRAG△

EM (%): 60.18. Result ID: 69fdbe1d47f34ce924e16e24.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### LatentRAG♢

EM (%): 59.85. Result ID: 69fdbe1d47f34ce924e16e22.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### LatentRAG△

EM (%): 58.83. Result ID: 69fdbe1d47f34ce924e16e34.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"F2LLM-v2-330M"}. This record does not assert independent reproduction.

### LatentRAG♢

EM (%): 58.56. Result ID: 69fdbe1d47f34ce924e16e32.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"F2LLM-v2-330M"}. This record does not assert independent reproduction.

### Iter-RetGen

EM (%): 56.91. Result ID: 69fdbe1d47f34ce924e16e1c.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e48a.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### R1-Searcher

EM (%): 56.1. Result ID: 69fdbe1d47f34ce924e16e1e.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77771e209131cea7356d.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### DeepRAG

EM (%): 55.48. Result ID: 69fdbe1d47f34ce924e16e20.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79411e209131cea7ce63.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### ZeroSearch

EM (%): 54.67. Result ID: 69fdbe1d47f34ce924e16e1f.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa789d1e209131cea79568.md>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### Naive RAG

EM (%): 53.32. Result ID: 69fdbe1d47f34ce924e16e1b.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.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: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### Search-o1

EM (%): 44.53. Result ID: 69fdbe1d47f34ce924e16e1d.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76911e209131cea6dca9.md>)

[Implementation](<https://github.com/sunnynexus/Search-o1>)

Recorded evaluation configuration: {"Retrieval Model":"Qwen3-Embedding-0.6B"}. This record does not assert independent reproduction.

### Direct Infer

EM (%): 42.48. Result ID: 69fdbe1d47f34ce924e16e1a.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e205497b3b670c894ce.md>)

Recorded evaluation configuration: {"Retrieval Model":"N/A"}. This record does not assert independent reproduction.

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