# General Tasks on MMLU-Redux2

Source: [General Tasks on MMLU-Redux2](<https://www.sota2.com/research/sota/general-tasks-on-mmlu-redux2>)

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

Data updated: 2026-02-26T05:54:51.522Z

Evaluation of model performance on the General Tasks category within the MMLU-Redux2 dataset, measured by PAQ and F1 scores.

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 Tasks |
| Dataset (including any stated split) | MMLU-Redux2 |
| Metric | PAQ |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [PAQ](<https://www.sota2.com/research/sota/general-tasks-on-mmlu-redux2.md?metric=PAQ>): Higher is better; unit: PERCENT.

- [F1 Score](<https://www.sota2.com/research/sota/general-tasks-on-mmlu-redux2.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | PAQ | Result date | Report paper |
| --- | --- | --- | --- |
| UCPO | 91.67 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 88.82 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 88.51 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 88.13 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO | 87.62 | Not recorded | 69802f68a8b01e76cc8e14db |
| Baseline | 87.2 | Not recorded | 69802f68a8b01e76cc8e14db |
| Prompt-UC | 87.08 | Not recorded | 69802f68a8b01e76cc8e14db |
| UCPO | 81.13 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 78.89 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 76.99 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO-UC | 72.47 | Not recorded | 69802f68a8b01e76cc8e14db |
| GRPO | 71.23 | Not recorded | 69802f68a8b01e76cc8e14db |
| Prompt-UC | 68.02 | Not recorded | 69802f68a8b01e76cc8e14db |
| Baseline | 65.17 | Not recorded | 69802f68a8b01e76cc8e14db |

### UCPO

PAQ: 91.67. Result ID: 699615925698665ec455215f.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B"}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 88.82. Result ID: 699615925698665ec455215d.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Reward Coefficient (ru)":0.5}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 88.51. Result ID: 699615925698665ec455215e.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Reward Coefficient (ru)":0.8}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 88.13. Result ID: 699615925698665ec455215c.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Reward Coefficient (ru)":0.2}. This record does not assert independent reproduction.

### GRPO

PAQ: 87.62. Result ID: 699615925698665ec455215b.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B"}. This record does not assert independent reproduction.

### Baseline

PAQ: 87.2. Result ID: 699615925698665ec4552159.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B"}. This record does not assert independent reproduction.

### Prompt-UC

PAQ: 87.08. Result ID: 699615925698665ec455215a.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B"}. This record does not assert independent reproduction.

### UCPO

PAQ: 81.13. Result ID: 699615925698665ec4552166.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 78.89. Result ID: 699615925698665ec4552164.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct","Reward Coefficient (ru)":0.5}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 76.99. Result ID: 699615925698665ec4552165.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct","Reward Coefficient (ru)":0.8}. This record does not assert independent reproduction.

### GRPO-UC

PAQ: 72.47. Result ID: 699615925698665ec4552163.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct","Reward Coefficient (ru)":0.2}. This record does not assert independent reproduction.

### GRPO

PAQ: 71.23. Result ID: 699615925698665ec4552162.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Prompt-UC

PAQ: 68.02. Result ID: 699615925698665ec4552161.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Baseline

PAQ: 65.17. Result ID: 699615925698665ec4552160.

[Reporting paper](<https://www.sota2.com/research/paper/69802f68a8b01e76cc8e14db.md>)

Recorded evaluation configuration: {"Backbone Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

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