Question Generation on EduMath-CQ
47.11Win RateEQPR
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| EQPRBackbone=Claude-3.52025.06 | 47.11 | 45.92 | — | — | — | — | — | — | |
| EQPRBackbone=DeepSeek-V32025.06 | 46.23 | 43.11 | — | 20.33 | 46.86 | 44.57 | 71.02 | 91.73 | |
| REACTBackbone=Claude-3.52025.06 | 45.27 | 41.26 | — | — | — | — | — | — | |
| DEARBackbone=Claude-3.52025.06 | 44.19 | 34.75 | — | — | — | — | — | — | |
| EQPRBackbone=GPT-4o2025.06 | 44.07 | 43.43 | — | — | — | — | — | — | |
| REACTBackbone=GPT-4o2025.06 | 42.68 | 40.72 | — | — | — | — | — | — | |
| DEARBackbone=DeepSeek-V32025.06 | 41.81 | 37.88 | — | 20.71 | 49.06 | 42.11 | 74.48 | 92.59 | |
| COT-BONBackbone=Claude-3.52025.06 | 41.61 | 38.37 | — | — | — | — | — | — | |
| DEARBackbone=GPT-4o2025.06 | 41.3 | 39.59 | — | — | — | — | — | — | |
| EQPRBackbone=GPT-4o-Mini2025.06 | 39.92 | 35.81 | — | 24.8 | 52.7 | 48.13 | 70.89 | 90.63 | |
| REACTBackbone=DeepSeek-V32025.06 | 39.75 | 38.91 | — | 21.76 | 47.98 | 47.06 | 70.32 | 90.64 | |
| COT-BONBackbone=DeepSeek-V32025.06 | 38.99 | 29.63 | — | 13.58 | 43.6 | 28.93 | 70.85 | 90.25 | |
| REACTBackbone=GPT-4o-Mini2025.06 | 38.4 | 33.66 | — | 18.9 | 53.2 | 36.72 | 72.55 | 89.32 | |
| COTBackbone=DeepSeek-V32025.06 | 38.2 | 31.45 | — | 12.11 | 42.9 | 22.51 | 70.64 | 90.84 | |
| COTBackbone=GPT-4o2025.06 | 38.16 | 34.78 | — | — | — | — | — | — | |
| COTBackbone=Claude-3.52025.06 | 37.95 | 32.08 | — | — | — | — | — | — | |
| EQPRBackbone=DeepSeek-V32025.06 | 36.67 | — | 1.93 | — | — | — | — | — | |
| COTBackbone=GPT-4o-Mini2025.06 | 35.78 | 27.88 | — | 11.6 | 29.64 | 24.45 | 68.52 | 90.12 | |
| DEARBackbone=GPT-4o-Mini2025.06 | 35.18 | 27.35 | — | 13.31 | 23.09 | 21.74 | 70.34 | 85.77 | |
| REACTBackbone=DeepSeek-V32025.06 | 35 | — | 1.87 | — | — | — | — | — | |
| COT-BONBackbone=GPT-4o-Mini2025.06 | 34.5 | 21.25 | — | 8.52 | 22.59 | 10.56 | 69.19 | 90.23 | |
| COT-BONBackbone=GPT-4o2025.06 | 34.34 | 36.58 | — | — | — | — | — | — | |
| COT-BONBackbone=DeepSeek-V32025.06 | 33.67 | — | 1.78 | — | — | — | — | — | |
| DEARBackbone=DeepSeek-V32025.06 | 31 | — | 1.84 | — | — | — | — | — | |
| COTBackbone=DeepSeek-V32025.06 | 27 | — | 1.83 | — | — | — | — | — |