SAT Reasoning on SATURN 2.6k Difficulty 3,5,8 (test)
47Pass@1SATURN-1.5B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SATURN-1.5BIteration=22025.05 | 47 | 82.6 | 93.9 | 98 | 99.8 | |
| SATURN-1.5BIteration=12025.05 | 41 | 74 | 85.6 | 91.1 | 95.6 | |
| DeepSeek-R1-Distill-Qwen-1.5B2025.05 | 20.3 | 47.6 | 63.6 | 73.4 | 81.9 |