Mathematical Reasoning on GSM8K (Acc, Cost)
93.21AccuracyCoT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CoTModel=Gemma32026.03 | 93.21 | 850 | |
| CoTModel=Qwen3-8B2026.03 | 89.09 | 799.7 | |
| CoTModel=Llama3.12026.03 | 87.52 | 817 | |
| ToTModel=Qwen3-8B2026.03 | 3.69 | 3,175 | |
| ToTModel=Llama3.12026.03 | 3.36 | 3,033 | |
| DSTModel=Qwen3-8B2026.03 | 3.35 | 192.3 | |
| ToTModel=Gemma32026.03 | 2.94 | 3,650 | |
| DSTModel=Llama3.12026.03 | 2.65 | 241 | |
| DSTModel=Gemma32026.03 | 2.63 | 400 | |
| DPTSModel=Llama3.12026.03 | 1.72 | 2,506 | |
| DPTSModel=Qwen3-8B2026.03 | 1.53 | 2,670 | |
| DPTSModel=Gemma32026.03 | 1.33 | 3,050 |