Mathematical Reasoning on AMC 2023 (Accuracy and Token Efficiency)
98.44AccuracyQwen3-32B-SCALE
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen3-32B-SCALEBackbone=Qwen3-32B, Strategy=SCALE2025.11 | 98.44 | 1,991 | 12,556 | 84.2 | |
| Qwen3-32B-InftyThinkBackbone=Qwen3-32B, Strategy=InftyThink2025.11 | 98.12 | 3,723 | 23,117 | 83.96 | |
| Qwen3-32B-Majority-VotingBackbone=Qwen3-32B, Strategy=Majority-Voting2025.11 | 97.5 | — | 46,315 | 75.83 | |
| QwQ-Majority-VotingBackbone=QwQ, Strategy=Majority-Voting2025.11 | 97.5 | — | 49,291 | 79.17 | |
| QwQ-InftyThinkBackbone=QwQ, Strategy=InftyThink2025.11 | 97.19 | 5,449 | 34,725 | 79.99 | |
| Qwen3-32B-CoTBackbone=Qwen3-32B, Strategy=Chain-of-Thought2025.11 | 96.88 | 5,789 | 5,789 | 75.9 | |
| QwQ-CoTBackbone=QwQ, Strategy=Chain-of-Thought2025.11 | 96.88 | 6,161 | 6,161 | 78.4 | |
| Distill-Llama-70B-SCALEBackbone=Distill-Llama-70B, Strategy=SCALE2025.11 | 96.88 | 2,566 | 16,238 | 76.74 | |
| QwQ-SCALEBackbone=QwQ, Strategy=SCALE2025.11 | 96.25 | 2,551 | 16,259 | 81.95 | |
| Distill-Qwen-32B-SCALEBackbone=Distill-Qwen-32B, Strategy=SCALE2025.11 | 95.63 | 2,672 | 16,804 | 76.04 | |
| Distill-Llama-70B-CoTBackbone=Distill-Llama-70B, Strategy=Chain-of-Thought2025.11 | 95.31 | 4,825 | 4,825 | 75.56 | |
| Distill-Llama-70B-Majority-VotingBackbone=Distill-Llama-70B, Strategy=Majority-Voting2025.11 | 95 | — | 38,605 | 76.11 | |
| Distill-Qwen-32B-CoTBackbone=Distill-Qwen-32B, Strategy=Chain-of-Thought2025.11 | 95 | 4,994 | 4,994 | 75.42 | |
| Distill-Qwen-32B-Majority-VotingBackbone=Distill-Qwen-32B, Strategy=Majority-Voting2025.11 | 95 | — | 39,953 | 78.33 | |
| Distill-Llama-70B-InftyThinkBackbone=Distill-Llama-70B, Strategy=InftyThink2025.11 | 93.75 | 3,844 | 24,327 | 72.64 | |
| Distill-Qwen-32B-InftyThinkBackbone=Distill-Qwen-32B, Strategy=InftyThink2025.11 | 92.81 | 3,711 | 23,332 | 69.69 |