SAT Reasoning on SATURN Difficulty 3,5,13 2.6k (test)
89.5Pass@1SATURN-7B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SATURN-7BIteration=22025.05 | 89.5 | 99 | 99.9 | 100 | 100 | |
| SATURN-7BIteration=12025.05 | 73 | 96.1 | 98.9 | 99.7 | 99.9 | |
| DeepSeek-R1-Distill-Qwen-7B2025.05 | 53.9 | 86.2 | 94.2 | 97.3 | 99.3 |