# Mathematical Reasoning on Minerva Math (avg@8, pass@8)

Source: [Mathematical Reasoning on Minerva Math (avg@8, pass@8)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-minerva-math-avg-at-8-pass-at-8>)

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

Data updated: 2026-05-18T06:57:40.373Z

Evaluation of model performance on complex mathematical reasoning problems using the Minerva Math dataset, reporting sampling metrics avg@8 and pass@8.

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 | Mathematical Reasoning |
| Dataset (including any stated split) | Minerva Math |
| Metric | Average Score (avg@8) |
| 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

- [Average Score (avg@8)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-minerva-math-avg-at-8-pass-at-8.md?metric=Average%20Score%20(avg%408)>): Higher is better; unit: PERCENT.

- [Pass Rate (pass@8)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-minerva-math-avg-at-8-pass-at-8.md?metric=Pass%20Rate%20(pass%408)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Score (avg@8) | Result date | Report paper |
| --- | --- | --- | --- |
| ASPO-Math-30B | 54.3 | Not recorded | 68fa7bf11e209131cea8b7bc |
| GRPO | 51.9 | Not recorded | 68fa7bf11e209131cea8b7bc |
| ASPO-Math-8B | 50.6 | Not recorded | 68fa7bf11e209131cea8b7bc |
| GRPO | 50.2 | Not recorded | 68fa7bf11e209131cea8b7bc |
| ASPO-Math-4B | 49 | Not recorded | 68fa7bf11e209131cea8b7bc |
| Qwen3-30B-A3B | 47.7 | Not recorded | 68fa7bf11e209131cea8b7bc |
| GRPO | 47.2 | Not recorded | 68fa7bf11e209131cea8b7bc |
| Qwen3-8B | 43.9 | Not recorded | 68fa7bf11e209131cea8b7bc |
| Qwen3-4B | 41.5 | Not recorded | 68fa7bf11e209131cea8b7bc |
| Nemotron-1.5B | 35.3 | Not recorded | 68fa7bf11e209131cea8b7bc |
| ASPO-Math-1.5B | 35.1 | Not recorded | 68fa7bf11e209131cea8b7bc |
| DeepScaleR-1.5B | 30.3 | Not recorded | 68fa7bf11e209131cea8b7bc |
| GRPO | 29.2 | Not recorded | 68fa7bf11e209131cea8b7bc |
| DeepSeek-R1-1.5B | 27.6 | Not recorded | 68fa7bf11e209131cea8b7bc |

### ASPO-Math-30B

Average Score (avg@8): 54.3. Result ID: 6a0ab74047f34ce924f04645.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-30B-A3B"}. This record does not assert independent reproduction.

### GRPO

Average Score (avg@8): 51.9. Result ID: 6a0ab74047f34ce924f04644.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

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

Recorded evaluation configuration: {"Backbone":"Qwen3-30B-A3B"}. This record does not assert independent reproduction.

### ASPO-Math-8B

Average Score (avg@8): 50.6. Result ID: 6a0ab74047f34ce924f04642.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

### GRPO

Average Score (avg@8): 50.2. Result ID: 6a0ab74047f34ce924f04641.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

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

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

### ASPO-Math-4B

Average Score (avg@8): 49. Result ID: 6a0ab74047f34ce924f0463f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Qwen3-30B-A3B

Average Score (avg@8): 47.7. Result ID: 6a0ab74047f34ce924f04643.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

Recorded evaluation configuration: {"Backbone":"Qwen3-30B-A3B"}. This record does not assert independent reproduction.

### GRPO

Average Score (avg@8): 47.2. Result ID: 6a0ab74047f34ce924f0463e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

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

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Qwen3-8B

Average Score (avg@8): 43.9. Result ID: 6a0ab74047f34ce924f04640.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

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

### Qwen3-4B

Average Score (avg@8): 41.5. Result ID: 6a0ab74047f34ce924f0463d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Nemotron-1.5B

Average Score (avg@8): 35.3. Result ID: 6a0ab74047f34ce924f0463b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79041e209131cea7bb2d.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### ASPO-Math-1.5B

Average Score (avg@8): 35.1. Result ID: 6a0ab74047f34ce924f0463c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### DeepScaleR-1.5B

Average Score (avg@8): 30.3. Result ID: 6a0ab74047f34ce924f0463a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### GRPO

Average Score (avg@8): 29.2. Result ID: 6a0ab74047f34ce924f04639.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

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

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

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### DeepSeek-R1-1.5B

Average Score (avg@8): 27.6. Result ID: 6a0ab74047f34ce924f04638.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bf11e209131cea8b7bc.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

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