Mathematical Reasoning on AIME 2024 (Accuracy and Token Metrics)
83.75AccuracyQwen3-32B-InftyThink
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen3-32B-InftyThinkBackbone=Qwen3-32B, Strategy=InftyThink2025.11 | 83.75 | 5,900 | 38,451 | 83.96 | |
| Qwen3-32B-SCALEBackbone=Qwen3-32B, Strategy=SCALE2025.11 | 82.92 | 3,550 | 25,581 | 84.2 | |
| Distill-Qwen-32B-Majority-VotingBackbone=Distill-Qwen-32B, Strategy=Majority-Voting2025.11 | 80 | — | 63,276 | 78.33 | |
| QwQ-SCALEBackbone=QwQ, Strategy=SCALE2025.11 | 79.4 | 4,157 | 29,812 | 81.95 | |
| QwQ-InftyThinkBackbone=QwQ, Strategy=InftyThink2025.11 | 77.25 | 8,722 | 62,557 | 79.99 | |
| Qwen3-32B-Majority-VotingBackbone=Qwen3-32B, Strategy=Majority-Voting2025.11 | 76.67 | — | 59,279 | 75.83 | |
| QwQ-Majority-VotingBackbone=QwQ, Strategy=Majority-Voting2025.11 | 76.67 | — | 79,663 | 79.17 | |
| Distill-Llama-70B-Majority-VotingBackbone=Distill-Llama-70B, Strategy=Majority-Voting2025.11 | 76.67 | — | 65,277 | 76.11 | |
| QwQ-CoTBackbone=QwQ, Strategy=Chain-of-Thought2025.11 | 75 | 9,957 | 9,957 | 78.4 | |
| Qwen3-32B-CoTBackbone=Qwen3-32B, Strategy=Chain-of-Thought2025.11 | 73.33 | 7,409 | 7,409 | 75.9 | |
| Distill-Qwen-32B-SCALEBackbone=Distill-Qwen-32B, Strategy=SCALE2025.11 | 73.33 | 4,270 | 30,175 | 76.04 | |
| Distill-Qwen-32B-CoTBackbone=Distill-Qwen-32B, Strategy=Chain-of-Thought2025.11 | 72.5 | 7,909 | 7,909 | 75.42 | |
| Distill-Llama-70B-SCALEBackbone=Distill-Llama-70B, Strategy=SCALE2025.11 | 71.25 | 4,313 | 30,696 | 76.74 | |
| Distill-Llama-70B-CoTBackbone=Distill-Llama-70B, Strategy=Chain-of-Thought2025.11 | 70 | 8,159 | 8,159 | 75.56 | |
| Distill-Llama-70B-InftyThinkBackbone=Distill-Llama-70B, Strategy=InftyThink2025.11 | 69.58 | 5,921 | 42,143 | 72.64 | |
| Distill-Qwen-32B-InftyThinkBackbone=Distill-Qwen-32B, Strategy=InftyThink2025.11 | 67.5 | 5,244 | 37,062 | 69.69 |