# Mathematical Reasoning on OLYMPIAD (Pass@1, AvgToken)

Source: [Mathematical Reasoning on OLYMPIAD (Pass@1, AvgToken)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-olympiad-pass-at-1-avgtoken>)

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

Data updated: 2026-04-23T11:00:30.763Z

Evaluation of model performance on mathematical reasoning problems using the OLYMPIAD dataset, measured by Pass@1 score and average token length.

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) | OLYMPIAD |
| Metric | Pass@1 Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| 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 Score](<https://www.sota2.com/research/sota/mathematical-reasoning-on-olympiad-pass-at-1-avgtoken.md?metric=Pass%401%20Score>): Higher is better; unit: PERCENT.

- [Average Token Length](<https://www.sota2.com/research/sota/mathematical-reasoning-on-olympiad-pass-at-1-avgtoken.md?metric=Average%20Token%20Length>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Pass@1 Score | Result date | Report paper |
| --- | --- | --- | --- |
| Eurus-2-7B-PRIME | 42.2 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO w/ Skywork-7B | 38.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO-VPS | 37.2 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO | 36.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GSPO | 36.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| DrGRPO | 35.9 | Not recorded | 69e9a79155a3c32af89f70a5 |
| S-GRPO | 35.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO-VPS | 32.1 | Not recorded | 69e9a79155a3c32af89f70a5 |
| DrGRPO | 30.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO | 30 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO w/ Skywork-1.5B | 29.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GSPO | 28.9 | Not recorded | 69e9a79155a3c32af89f70a5 |
| S-GRPO | 28.3 | Not recorded | 69e9a79155a3c32af89f70a5 |
| BASE | 19 | Not recorded | 69e9a79155a3c32af89f70a5 |
| BASE | 18.8 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO-VPS | 6.2 | Not recorded | 69e9a79155a3c32af89f70a5 |
| GRPO | 5.9 | Not recorded | 69e9a79155a3c32af89f70a5 |
| BASE | 2.8 | Not recorded | 69e9a79155a3c32af89f70a5 |

### Eurus-2-7B-PRIME

Pass@1 Score: 42.2. Result ID: 69e9fbb147f34ce924d209a9.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b9f1e209131cea8965d.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### GRPO w/ Skywork-7B

Pass@1 Score: 38.8. Result ID: 69e9fbb147f34ce924d209aa.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### GRPO-VPS

Pass@1 Score: 37.2. Result ID: 69e9fbb147f34ce924d209ac.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Score: 36.8. Result ID: 69e9fbb147f34ce924d209a6.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

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

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

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### GSPO

Pass@1 Score: 36.8. Result ID: 69e9fbb147f34ce924d209a8.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a481e209131cea8231d.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### DrGRPO

Pass@1 Score: 35.9. Result ID: 69e9fbb147f34ce924d209a7.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be51e209131cea8b201.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### S-GRPO

Pass@1 Score: 35.8. Result ID: 69e9fbb147f34ce924d209ab.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78961e209131cea79249.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### GRPO-VPS

Pass@1 Score: 32.1. Result ID: 69e9fbb147f34ce924d209a4.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### DrGRPO

Pass@1 Score: 30.8. Result ID: 69e9fbb147f34ce924d209a0.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be51e209131cea8b201.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Score: 30. Result ID: 69e9fbb147f34ce924d2099f.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

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

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

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### GRPO w/ Skywork-1.5B

Pass@1 Score: 29.8. Result ID: 69e9fbb147f34ce924d209a2.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### GSPO

Pass@1 Score: 28.9. Result ID: 69e9fbb147f34ce924d209a1.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a481e209131cea8231d.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### S-GRPO

Pass@1 Score: 28.3. Result ID: 69e9fbb147f34ce924d209a3.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78961e209131cea79249.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"With Process Supervision"}. This record does not assert independent reproduction.

### BASE

Pass@1 Score: 19. Result ID: 69e9fbb147f34ce924d2099e.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-1.5B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### BASE

Pass@1 Score: 18.8. Result ID: 69e9fbb147f34ce924d209a5.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Supervision Type":"Outcome Supervision Only"}. This record does not assert independent reproduction.

### GRPO-VPS

Pass@1 Score: 6.2. Result ID: 69e9fbb147f34ce924d209af.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-2B-it"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Score: 5.9. Result ID: 69e9fbb147f34ce924d209ae.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

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

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

Recorded evaluation configuration: {"Backbone":"Gemma-2-2B-it"}. This record does not assert independent reproduction.

### BASE

Pass@1 Score: 2.8. Result ID: 69e9fbb147f34ce924d209ad.

[Reporting paper](<https://www.sota2.com/research/paper/69e9a79155a3c32af89f70a5.md>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-2B-it"}. This record does not assert independent reproduction.

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