# Mathematical Reasoning on MATH 500 (pass@1, pass@128)

Source: [Mathematical Reasoning on MATH 500 (pass@1, pass@128)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-math-500-pass-at-1-pass-at-128>)

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

Data updated: 2026-04-14T11:38:51.146Z

Evaluation of model performance on mathematical reasoning problems using the MATH 500 dataset, measured by pass@1 and pass@128 success rates.

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) | MATH 500 |
| Metric | Pass@1 Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [Pass@1 Rate](<https://www.sota2.com/research/sota/mathematical-reasoning-on-math-500-pass-at-1-pass-at-128.md?metric=Pass%401%20Rate>): Higher is better; unit: PERCENT.

- [Pass@128 Rate](<https://www.sota2.com/research/sota/mathematical-reasoning-on-math-500-pass-at-1-pass-at-128.md?metric=Pass%40128%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pass@1 Rate | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen3-8B + MEDS | 92.51 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-8B + DAPO | 89.18 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen2.5-Math-7B + MEDS | 86.33 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-8B + GRPO w/ Entropy Adv | 85.89 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen2.5-Math-7B + DAPO | 85.61 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-1.7B + MEDS | 84.66 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen2.5-Math-7B + GRPO w/ Entropy Adv | 84.5 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen2.5-Math-7B + GRPO | 84.1 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-8B | 83.62 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-8B + GRPO | 82.38 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-1.7B | 80.86 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-1.7B + DAPO | 80.25 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-1.7B + GRPO | 79.58 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen3-1.7B + GRPO w/ Entropy Adv | 78.72 | Not recorded | 69ddca4a7d0172140ec79cbf |
| Qwen2.5-Math-7B | 30.84 | Not recorded | 69ddca4a7d0172140ec79cbf |

### Qwen3-8B + MEDS

Pass@1 Rate: 92.51. Result ID: 69de273a47f34ce924c6b568.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Implementation](<https://github.com/Linxi000/MEDS>)

Recorded evaluation configuration: {"Base Model":"Qwen3-8B","Training Strategy":"MEDS"}. This record does not assert independent reproduction.

### Qwen3-8B + DAPO

Pass@1 Rate: 89.18. Result ID: 69de273a47f34ce924c6b566.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a81e209131cea79ac6.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-8B","Training Strategy":"DAPO"}. This record does not assert independent reproduction.

### Qwen2.5-Math-7B + MEDS

Pass@1 Rate: 86.33. Result ID: 69de273a47f34ce924c6b563.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Implementation](<https://github.com/Linxi000/MEDS>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B","Training Strategy":"MEDS"}. This record does not assert independent reproduction.

### Qwen3-8B + GRPO w/ Entropy Adv

Pass@1 Rate: 85.89. Result ID: 69de273a47f34ce924c6b567.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a921e209131cea83f9d.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-8B","Training Strategy":"GRPO w/ Entropy Adv"}. This record does not assert independent reproduction.

### Qwen2.5-Math-7B + DAPO

Pass@1 Rate: 85.61. Result ID: 69de273a47f34ce924c6b561.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a81e209131cea79ac6.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B","Training Strategy":"DAPO"}. This record does not assert independent reproduction.

### Qwen3-1.7B + MEDS

Pass@1 Rate: 84.66. Result ID: 69de273a47f34ce924c6b55e.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Implementation](<https://github.com/Linxi000/MEDS>)

Recorded evaluation configuration: {"Base Model":"Qwen3-1.7B","Training Strategy":"MEDS"}. This record does not assert independent reproduction.

### Qwen2.5-Math-7B + GRPO w/ Entropy Adv

Pass@1 Rate: 84.5. Result ID: 69de273a47f34ce924c6b562.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a921e209131cea83f9d.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B","Training Strategy":"GRPO w/ Entropy Adv"}. This record does not assert independent reproduction.

### Qwen2.5-Math-7B + GRPO

Pass@1 Rate: 84.1. Result ID: 69de273a47f34ce924c6b560.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

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

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

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B","Training Strategy":"GRPO"}. This record does not assert independent reproduction.

### Qwen3-8B

Pass@1 Rate: 83.62. Result ID: 69de273a47f34ce924c6b564.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

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

### Qwen3-8B + GRPO

Pass@1 Rate: 82.38. Result ID: 69de273a47f34ce924c6b565.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

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

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

Recorded evaluation configuration: {"Base Model":"Qwen3-8B","Training Strategy":"GRPO"}. This record does not assert independent reproduction.

### Qwen3-1.7B

Pass@1 Rate: 80.86. Result ID: 69de273a47f34ce924c6b55a.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-1.7B","Training Strategy":"Base"}. This record does not assert independent reproduction.

### Qwen3-1.7B + DAPO

Pass@1 Rate: 80.25. Result ID: 69de273a47f34ce924c6b55c.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a81e209131cea79ac6.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-1.7B","Training Strategy":"DAPO"}. This record does not assert independent reproduction.

### Qwen3-1.7B + GRPO

Pass@1 Rate: 79.58. Result ID: 69de273a47f34ce924c6b55b.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

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

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

Recorded evaluation configuration: {"Base Model":"Qwen3-1.7B","Training Strategy":"GRPO"}. This record does not assert independent reproduction.

### Qwen3-1.7B + GRPO w/ Entropy Adv

Pass@1 Rate: 78.72. Result ID: 69de273a47f34ce924c6b55d.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a921e209131cea83f9d.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-1.7B","Training Strategy":"GRPO w/ Entropy Adv"}. This record does not assert independent reproduction.

### Qwen2.5-Math-7B

Pass@1 Rate: 30.84. Result ID: 69de273a47f34ce924c6b55f.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca4a7d0172140ec79cbf.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B","Training Strategy":"Base"}. This record does not assert independent reproduction.

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