# General Reasoning and Language Understanding on MMLU (test)

Source: [General Reasoning and Language Understanding on MMLU (test)](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test>)

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

Data updated: 2026-03-04T13:10:34.867Z

Evaluation of model performance across the MMLU test set, covering general reasoning and language understanding capabilities, with results reported per model.

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 | General Reasoning and Language Understanding |
| Dataset (including any stated split) | MMLU (test) |
| Metric | gpt4o-mini Score |
| 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

- [gpt4o-mini Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=gpt4o-mini%20Score>): Higher is better; unit: PERCENT.

- [deepseek-v3 Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=deepseek-v3%20Score>): Higher is better; unit: PERCENT.

- [llama3-70B Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=llama3-70B%20Score>): Higher is better; unit: PERCENT.

- [gpt4o Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=gpt4o%20Score>): Higher is better; unit: PERCENT.

- [qwen-72B Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=qwen-72B%20Score>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/general-reasoning-and-language-understanding-on-mmlu-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | gpt4o-mini Score | Result date | Report paper |
| --- | --- | --- | --- |
| CARD | 84.97 | Not recorded | 69a66c142d603e6de3e3cd9e |
| G-designer | 83 | Not recorded | 69a66c142d603e6de3e3cd9e |
| GPTswarm | 82.35 | Not recorded | 69a66c142d603e6de3e3cd9e |
| CoT | 81.05 | Not recorded | 69a66c142d603e6de3e3cd9e |
| LLM-Debate | 80.39 | Not recorded | 69a66c142d603e6de3e3cd9e |
| Random-graph | 79.74 | Not recorded | 69a66c142d603e6de3e3cd9e |
| Aflow | 79.74 | Not recorded | 69a66c142d603e6de3e3cd9e |
| Vanilla | 77.12 | Not recorded | 69a66c142d603e6de3e3cd9e |

### CARD

gpt4o-mini Score: 84.97. Result ID: 69a82f336dbb7075b94e946e.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"Yes","Conditional Configuration":"Yes"}. This record does not assert independent reproduction.

### G-designer

gpt4o-mini Score: 83. Result ID: 69a82f336dbb7075b94e946d.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76e91e209131cea6fe24.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"Yes","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### GPTswarm

gpt4o-mini Score: 82.35. Result ID: 69a82f336dbb7075b94e946b.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"Yes","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### CoT

gpt4o-mini Score: 81.05. Result ID: 69a82f336dbb7075b94e9468.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Multi-agent":"No","Automated Topology":"No","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### LLM-Debate

gpt4o-mini Score: 80.39. Result ID: 69a82f336dbb7075b94e946a.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6a1e209131cea456e2.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"No","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### Random-graph

gpt4o-mini Score: 79.74. Result ID: 69a82f336dbb7075b94e9469.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"No","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### Aflow

gpt4o-mini Score: 79.74. Result ID: 69a82f336dbb7075b94e946c.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78111e209131cea769ef.md>)

Recorded evaluation configuration: {"Multi-agent":"Yes","Automated Topology":"Yes","Conditional Configuration":"No"}. This record does not assert independent reproduction.

### Vanilla

gpt4o-mini Score: 77.12. Result ID: 69a82f336dbb7075b94e9467.

[Reporting paper](<https://www.sota2.com/research/paper/69a66c142d603e6de3e3cd9e.md>)

Recorded evaluation configuration: {"Multi-agent":"No","Automated Topology":"No","Conditional Configuration":"No"}. This record does not assert independent reproduction.

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