Mathematical Reasoning on Bn-MGSM (test)
89.2Accuracygpt4.1
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| gpt4.12026.01 | 89.2 | 200 | 88.16 | |
| gpt4.1-mini2026.01 | 87.2 | 232 | 88.18 | |
| Qwen3-32BParameters=32B2026.01 | 85.6 | 712 | 21.08 | |
| Qwen3-14BParameters=14B2026.01 | 83.6 | 767 | 17.87 | |
| GANITLLM-4BParameters=4B2026.01 | 76.8 | 193 | 88.71 | |
| Qwen3-8BParameters=8B2026.01 | 69.2 | 977 | 19.26 | |
| Qwen3-4BParameters=4B2026.01 | 69.2 | 943 | 14.79 | |
| GANITLLM-1.7BParameters=1.7B2026.01 | 52.8 | 210 | 87.8 | |
| TigerLLM-9BParameters=9B2026.01 | 47.2 | 206 | 93.69 | |
| GANITLLM-0.6BParameters=0.6B2026.01 | 28.4 | 248 | 88.7 | |
| Qwen3-1.7BParameters=1.7B2026.01 | 15.2 | 1,124 | 19.64 | |
| Qwen3-0.6BParameters=0.6B2026.01 | 8.4 | 1,265 | 12.43 |