# Cross-lingual Reasoning and Factual Knowledge on Global MMLU (test)

Source: [Cross-lingual Reasoning and Factual Knowledge on Global MMLU (test)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test>)

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

Data updated: 2026-02-26T05:34:24.634Z

Evaluation of model performance on cross-lingual reasoning and factual knowledge across 17 specified languages using the test split of the Global MMLU dataset.

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 | Cross-lingual Reasoning and Factual Knowledge |
| Dataset (including any stated split) | Global MMLU (test) |
| Metric | Accuracy (RUS) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [Accuracy (RUS)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(RUS)>): Higher is better; unit: PERCENT.

- [Accuracy (POL)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(POL)>): Higher is better; unit: PERCENT.

- [Accuracy (CMN)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(CMN)>): Higher is better; unit: PERCENT.

- [Accuracy (NLD)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(NLD)>): Higher is better; unit: PERCENT.

- [Accuracy (DEU)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(DEU)>): Higher is better; unit: PERCENT.

- [Accuracy (IND)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(IND)>): Higher is better; unit: PERCENT.

- [Accuracy (JPN)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(JPN)>): Higher is better; unit: PERCENT.

- [Accuracy (TUR)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(TUR)>): Higher is better; unit: PERCENT.

- [Accuracy (FRA)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(FRA)>): Higher is better; unit: PERCENT.

- [Accuracy (CES)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(CES)>): Higher is better; unit: PERCENT.

- [Accuracy (ITA)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(ITA)>): Higher is better; unit: PERCENT.

- [Accuracy (FAS)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(FAS)>): Higher is better; unit: PERCENT.

- [Accuracy (POR)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(POR)>): Higher is better; unit: PERCENT.

- [Accuracy (SWE)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(SWE)>): Higher is better; unit: PERCENT.

- [Accuracy (ELL)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(ELL)>): Higher is better; unit: PERCENT.

- [Accuracy (VIE)](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Accuracy%20(VIE)>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

- [Difference](<https://www.sota2.com/research/sota/cross-lingual-reasoning-and-factual-knowledge-on-global-mmlu-test.md?metric=Difference>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Accuracy (RUS) | Result date | Report paper |
| --- | --- | --- | --- |
| w\_TREX | 23.46 | Not recorded | 69705dbde25ad2fa9d1ed43a |
| w\_llama | 22.95 | Not recorded | 69705dbde25ad2fa9d1ed43a |

### w\_TREX

Accuracy (RUS): 23.46. Result ID: 69967a085698665ec4c1a826.

[Reporting paper](<https://www.sota2.com/research/paper/69705dbde25ad2fa9d1ed43a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69705dbde25ad2fa9d1ed43a.md>)

Recorded evaluation configuration: {"Model Size":"200M","Training Volume":"50B tokens","Tokenizer":"TREX"}. This record does not assert independent reproduction.

### w\_llama

Accuracy (RUS): 22.95. Result ID: 69967a085698665ec4c1a825.

[Reporting paper](<https://www.sota2.com/research/paper/69705dbde25ad2fa9d1ed43a.md>)

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

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

Recorded evaluation configuration: {"Model Size":"200M","Training Volume":"50B tokens","Tokenizer":"LLaMA"}. This record does not assert independent reproduction.

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