Mathematical Reasoning on MATH-500 (Avg@4 accuracy)
85Avg@4 AccuracyReQueR
Evaluation Results
| Method | Links | |
|---|---|---|
| ReQueRModel=Qwen3-8B, Temperature=0.62026.04 | 85 | |
| CoTModel=Qwen3-8B, Temperature=0.62026.04 | 83.4 | |
| ReQueRModel=Qwen2.5-72B-Instruct, Temperature=0.62026.04 | 83 | |
| ReQueRModel=Qwen3-1.7B, Temperature=0.62026.04 | 77.6 | |
| CoTModel=Qwen2.5-72B-Instruct, Temperature=0.62026.04 | 76.2 | |
| ReQueRModel=Llama-3.1-70B-Instruct, Temperature=0.62026.04 | 72 | |
| CoTModel=Llama-3.1-70B-Instruct, Temperature=0.62026.04 | 64.6 | |
| CoTModel=Qwen3-1.7B, Temperature=0.62026.04 | 64.2 | |
| ReQueRModel=Qwen3-0.6B, Temperature=0.62026.04 | 54.4 | |
| ReQueRModel=Llama-3.1-8B-Instruct, Temperature=0.62026.04 | 54 | |
| ReQueRModel=Llama-3.2-3B-Instruct, Temperature=0.62026.04 | 50.2 | |
| CoTModel=Qwen3-0.6B, Temperature=0.62026.04 | 48.8 | |
| CoTModel=Llama-3.2-3B-Instruct, Temperature=0.62026.04 | 46.6 | |
| CoTModel=Llama-3.1-8B-Instruct, Temperature=0.62026.04 | 45.8 | |
| ReQueRModel=Mixtral-8x7B-Instruct, Temperature=0.62026.04 | 33 | |
| CoTModel=Mixtral-8x7B-Instruct, Temperature=0.62026.04 | 30.6 | |
| ReQueRModel=DS-MoE-16B-Chat, Temperature=0.62026.04 | 19.2 | |
| CoTModel=DS-MoE-16B-Chat, Temperature=0.62026.04 | 16.2 |