Knowledge-conflict Question Answering on PopQA CF-RAG
92.54CFFIDES
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FIDESBackbone=Qwen1.5-7B2026.06 | 92.54 | 52.36 | 58.11 | |
| FIDESBackbone=LLaMA3-8B2026.06 | 90.58 | 44.13 | 50.19 | |
| FIDESBackbone=Mistral-7B2026.06 | 89.43 | 42.56 | 44.23 | |
| FIDESBackbone=LLaMA2-7B2026.06 | 88.61 | 41.13 | 42.97 | |
| AdaCADBackbone=Qwen1.5-7B2026.06 | 88.15 | 42.29 | 47.95 | |
| CLEAR†Backbone=Qwen1.5-7B2026.06 | 88.01 | 46.82 | 51.17 | |
| AdaCADBackbone=LLaMA3-8B2026.06 | 86.7 | 38.87 | 44.09 | |
| COIECDBackbone=Qwen1.5-7B2026.06 | 86.54 | 37.19 | 41.73 | |
| COIECDBackbone=LLaMA3-8B2026.06 | 85.28 | 34.05 | 38.57 | |
| CADBackbone=Qwen1.5-7B2026.06 | 85.23 | 30.96 | 34.12 | |
| CADBackbone=LLaMA3-8B2026.06 | 84.12 | 28.17 | 31.83 | |
| CLEAR†Backbone=LLaMA3-8B2026.06 | 83.22 | 43.91 | 48.92 | |
| DVDBackbone=Qwen1.5-7B2026.06 | 82.52 | 41.17 | 46.82 | |
| DeCoReBackbone=Qwen1.5-7B2026.06 | 81.08 | 40.23 | 45.52 | |
| DoLaBackbone=Qwen1.5-7B2026.06 | 80.57 | 39.81 | 43.19 | |
| Standard RAGBackbone=Qwen1.5-7B2026.06 | 78.12 | 37.28 | 41.13 | |
| AdaCADBackbone=Mistral-7B2026.06 | 78.03 | 35.23 | 39.63 | |
| DVDBackbone=LLaMA3-8B2026.06 | 77.53 | 38.28 | 43.41 | |
| DeCoReBackbone=LLaMA3-8B2026.06 | 76.81 | 37.66 | 41.97 | |
| AdaCADBackbone=LLaMA2-7B2026.06 | 75.78 | 33.91 | 38.52 | |
| CLEAR†Backbone=Mistral-7B2026.06 | 75.61 | 39.27 | 43.82 | |
| DoLaBackbone=LLaMA3-8B2026.06 | 75.27 | 36.83 | 39.62 | |
| COIECDBackbone=Mistral-7B2026.06 | 73.85 | 31.78 | 35.27 | |
| Standard RAGBackbone=LLaMA3-8B2026.06 | 72.86 | 35.68 | 38.41 | |
| CLEAR†Backbone=LLaMA2-7B2026.06 | 72.18 | 37.83 | 42.16 | |
| COIECDBackbone=LLaMA2-7B2026.06 | 71.07 | 30.59 | 34.15 | |
| CADBackbone=Mistral-7B2026.06 | 70.43 | 27.56 | 29.93 | |
| DVDBackbone=Mistral-7B2026.06 | 68.27 | 34.42 | 39.16 | |
| CADBackbone=LLaMA2-7B2026.06 | 67.22 | 26.54 | 28.81 | |
| DeCoReBackbone=Mistral-7B2026.06 | 65.18 | 32.93 | 37.51 | |
| DVDBackbone=LLaMA2-7B2026.06 | 64.57 | 33.11 | 38.12 | |
| Standard RAGBackbone=Mistral-7B2026.06 | 64.31 | 31.84 | 35.62 | |
| DoLaBackbone=Mistral-7B2026.06 | 63.22 | 31.12 | 34.81 | |
| DeCoReBackbone=LLaMA2-7B2026.06 | 61.84 | 31.62 | 36.23 | |
| Standard RAGBackbone=LLaMA2-7B2026.06 | 61.27 | 30.43 | 34.18 | |
| DoLaBackbone=LLaMA2-7B2026.06 | 60.19 | 29.92 | 33.46 |