# Advanced Mathematical Problem Solving on Olympiad Bench (N=112)

Source: [Advanced Mathematical Problem Solving on Olympiad Bench (N=112)](<https://www.sota2.com/research/sota/advanced-mathematical-problem-solving-on-olympiad-bench-n-112>)

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

Data updated: 2026-05-21T11:45:10.377Z

Evaluation of model performance on advanced mathematical problem solving tasks using the N=112 subset of the Olympiad Bench, measured by Top-1 Accuracy.

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 | Advanced Mathematical Problem Solving |
| Dataset (including any stated split) | Olympiad Bench N=112 |
| Metric | Top-1 Accuracy |
| 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

- [Top-1 Accuracy](<https://www.sota2.com/research/sota/advanced-mathematical-problem-solving-on-olympiad-bench-n-112.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Draw2Think BL | 89.3 | Not recorded | 6a0e916ff984739ed6738980 |
| Draw2Think CT | 89.3 | Not recorded | 6a0e916ff984739ed6738980 |
| OpenAI-o3 | 77.7 | Not recorded | 6a0e916ff984739ed6738980 |
| Gemini-2.5-Pro | 75 | Not recorded | 6a0e916ff984739ed6738980 |
| GPT-4o | 13.4 | Not recorded | 6a0e916ff984739ed6738980 |

### Draw2Think BL

Top-1 Accuracy: 89.3. Result ID: 6a0eef90fbc373592c073b78.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

Recorded evaluation configuration: {"Backbone":"gemini-3-flash-preview","Execution Environment":"direct VLM"}. This record does not assert independent reproduction.

### Draw2Think CT

Top-1 Accuracy: 89.3. Result ID: 6a0eef90fbc373592c073b79.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

Recorded evaluation configuration: {"Backbone":"gemini-3-flash-preview","Execution Environment":"executable environment"}. This record does not assert independent reproduction.

### OpenAI-o3

Top-1 Accuracy: 77.7. Result ID: 6a0eef90fbc373592c073b77.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Reported by":"TrustGeoGen \[9\]"}. This record does not assert independent reproduction.

### Gemini-2.5-Pro

Top-1 Accuracy: 75. Result ID: 6a0eef90fbc373592c073b76.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Reported by":"TrustGeoGen \[9\]"}. This record does not assert independent reproduction.

### GPT-4o

Top-1 Accuracy: 13.4. Result ID: 6a0eef90fbc373592c073b75.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e916ff984739ed6738980.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75681e209131cea67a98.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Reported by":"TrustGeoGen \[9\]"}. This record does not assert independent reproduction.

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