Arithmetic Reasoning on GSM8K baseline original
1,885Correct CountGPT-OSS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-OSSParameters=120B, N=2,0002026.06 | 1,885 | 94.25 | 26.13 | |
| DeepSeek-R1Parameters=70B, N=2,0002026.06 | 1,659 | 82.95 | 406.25 | |
| Gemma4Parameters=31B, N=2,0002026.06 | 1,627 | 81.35 | 134.35 |