# Mathematical Reasoning on MultiArith (pass@1)

Source: [Mathematical Reasoning on MultiArith (pass@1)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-multiarith-pass-at-1>)

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

Data updated: 2026-06-11T07:15:08.428Z

Evaluation of model performance on mathematical reasoning problems using the MultiArith dataset, reporting the success rate at k=1 (pass@1).

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) | MultiArith |
| Metric | Pass@1 Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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-multiarith-pass-at-1.md?metric=Pass%401%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pass@1 Rate | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen2.5-1.5B | 82.5 | Not recorded | 6a28f35090966ade919c1829 |
| Coconut + dropout-GRPO | 64.67 | Not recorded | 6a28f35090966ade919c1829 |
| Coconut SFT | 61.67 | Not recorded | 6a28f35090966ade919c1829 |
| Coconut + GRPO | 61.67 | Not recorded | 6a28f35090966ade919c1829 |
| Coconut + REINFORCE + dropout | 57.39 | Not recorded | 6a28f35090966ade919c1829 |

### Qwen2.5-1.5B

Pass@1 Rate: 82.5. Result ID: 6a2a5f28fbc373592c26fa67.

[Reporting paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767e1e209131cea6d6a7.md>)

Recorded evaluation configuration: {"mode":"zero-shot"}. This record does not assert independent reproduction.

### Coconut + dropout-GRPO

Pass@1 Rate: 64.67. Result ID: 6a2a5f28fbc373592c26fa6b.

[Reporting paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

Recorded evaluation configuration: {"seeds":"3"}. This record does not assert independent reproduction.

### Coconut SFT

Pass@1 Rate: 61.67. Result ID: 6a2a5f28fbc373592c26fa68.

[Reporting paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa762f1e209131cea6bbb7.md>)

Recorded evaluation configuration: {"baseline":"true","seeds":"1"}. This record does not assert independent reproduction.

### Coconut + GRPO

Pass@1 Rate: 61.67. Result ID: 6a2a5f28fbc373592c26fa69.

[Reporting paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

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

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

Recorded evaluation configuration: {"rollout":"deterministic","seeds":"3"}. This record does not assert independent reproduction.

### Coconut + REINFORCE + dropout

Pass@1 Rate: 57.39. Result ID: 6a2a5f28fbc373592c26fa6a.

[Reporting paper](<https://www.sota2.com/research/paper/6a28f35090966ade919c1829.md>)

Recorded evaluation configuration: {"baseline\_type":"EMA","seeds":"3"}. This record does not assert independent reproduction.

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