Mathematical Reasoning on GSM8K (test) (p@6=0, R1, R3, R6, R8)
83p@6=0Llama-3.2-3B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Llama-3.2-3BModel Family=Llama-3.2, Parameters=3B2026.06 | 83 | 2 | 12 | 19 | 21 | |
| Qwen-2.5-3BModel Family=Qwen-2.5, Parameters=3B2026.06 | 58 | 0 | 5 | 8 | 10 | |
| Mistral-Nemo-12BModel Family=Mistral-Nemo, Parameters=12B2026.06 | 54 | 1 | 8 | 11 | 13 | |
| Llama-3.1-8BModel Family=Llama-3.1, Parameters=8B2026.06 | 51 | 3 | 7 | 10 | 11 |