Question Answering on QASC (R@1, R@3, N@1, N@3)
74.17Recall@1S-DPO
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| S-DPOModel=Llama32026.05 | 74.17 | 92.38 | 74.17 | 85.13 | |
| DPO-kModel=Llama32026.05 | 73.95 | 92.38 | 73.95 | 84.93 | |
| MASS-DPOModel=Llama32026.05 | 73.84 | 92.6 | 73.84 | 85.13 | |
| MASS-DPOModel=SmolLM32026.05 | 72.52 | 91.83 | 72.52 | 83.82 | |
| MASS-DPOModel=Qwen32026.05 | 71.85 | 91.61 | 71.85 | 83.73 | |
| DPOModel=Llama32026.05 | 71.3 | 91.72 | 71.3 | 83.41 | |
| DPO-kModel=Qwen32026.05 | 71.08 | 92.72 | 71.08 | 83.95 | |
| S-DPOModel=SmolLM32026.05 | 70.53 | 91.39 | 70.53 | 82.8 | |
| S-DPOModel=Qwen32026.05 | 70.42 | 91.5 | 70.42 | 83.04 | |
| DPO-kModel=SmolLM32026.05 | 70.31 | 90.18 | 70.31 | 82.06 | |
| DMPOModel=Llama32026.05 | 69.32 | 91.94 | 69.32 | 82.67 | |
| DMPOModel=Qwen32026.05 | 68.21 | 90.51 | 68.21 | 81.4 | |
| DPOModel=Qwen32026.05 | 67.77 | 90.62 | 67.77 | 81.29 | |
| DPOModel=SmolLM32026.05 | 67.11 | 90.07 | 67.11 | 80.7 | |
| DMPOModel=SmolLM32026.05 | 66.11 | 88.74 | 66.11 | 79.54 |