Mathematical Reasoning on OlympiadBench (Acc, GT-pass)
53.4AccuracyBASE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BASESolver=Gemini 3 Pro, K=82026.06 | 53.4 | 99 | |
| BASESolver=R1-Qwen3-8B, K=82026.06 | 53.2 | 100 | |
| Independent FormalizationSolver=R1-Qwen3-8B, K=82026.06 | 42.8 | 79.4 | |
| Independent FormalizationSolver=Gemini 3 Pro, K=82026.06 | 41.7 | 77.6 | |
| BASESolver=Qwen3-8B, K=82026.06 | 39.7 | 98.1 | |
| Independent FormalizationSolver=Qwen3-8B, K=82026.06 | 31.6 | 80.4 |