Advanced Mathematical Problem Solving on Olympiad Bench (N=112)
89.3Top-1 AccuracyDraw2Think BL
Evaluation Results
| Method | Links | |
|---|---|---|
| Draw2Think BLBackbone=gemini-3-flash-preview, Execution Environment=direct VLM2026.05 | 89.3 | |
| Draw2Think CTBackbone=gemini-3-flash-preview, Execution Environment=executable environment2026.05 | 89.3 | |
| OpenAI-o3Protocol=Zero-shot, Reported by=TrustGeoGen [9]2026.05 | 77.7 | |
| Gemini-2.5-ProProtocol=Zero-shot, Reported by=TrustGeoGen [9]2026.05 | 75 | |
| GPT-4oProtocol=Zero-shot, Reported by=TrustGeoGen [9]2026.05 | 13.4 |