Knowledge-conflict Question Answering on TriviaQA CF-RAG
91.86Conflict Factor (CF)FIDES
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FIDESBackbone=Qwen1.5-7B2026.06 | 91.86 | 51.48 | 57.29 | |
| FIDESBackbone=LLaMA3-8B2026.06 | 89.26 | 42.73 | 48.97 | |
| FIDESBackbone=Mistral-7B2026.06 | 88.27 | 40.42 | 42.17 | |
| AdaCADBackbone=Qwen1.5-7B2026.06 | 86.88 | 41.09 | 45.44 | |
| FIDESBackbone=LLaMA2-7B2026.06 | 86.76 | 39.42 | 44.93 | |
| CLEAR†Backbone=Qwen1.5-7B2026.06 | 86.58 | 44.73 | 50.51 | |
| COIECDBackbone=Qwen1.5-7B2026.06 | 85.05 | 36.16 | 39.99 | |
| AdaCADBackbone=LLaMA3-8B2026.06 | 84.8 | 37.31 | 42.4 | |
| CADBackbone=Qwen1.5-7B2026.06 | 83.56 | 30.14 | 33.32 | |
| COIECDBackbone=LLaMA3-8B2026.06 | 83.16 | 32.72 | 37.23 | |
| CADBackbone=LLaMA3-8B2026.06 | 81.82 | 27.11 | 30.91 | |
| DVDBackbone=Qwen1.5-7B2026.06 | 80.89 | 39.93 | 44.12 | |
| CLEAR†Backbone=LLaMA3-8B2026.06 | 80.52 | 41.28 | 46.49 | |
| DeCoReBackbone=Qwen1.5-7B2026.06 | 79.51 | 39.18 | 43.96 | |
| DoLaBackbone=Qwen1.5-7B2026.06 | 78.83 | 38.82 | 42.51 | |
| Standard RAGBackbone=Qwen1.5-7B2026.06 | 76.22 | 36.51 | 40.49 | |
| AdaCADBackbone=Mistral-7B2026.06 | 76.21 | 33.98 | 38.72 | |
| DVDBackbone=LLaMA3-8B2026.06 | 75.29 | 36.82 | 41.67 | |
| DeCoReBackbone=LLaMA3-8B2026.06 | 74.52 | 36.28 | 40.48 | |
| AdaCADBackbone=LLaMA2-7B2026.06 | 73.95 | 32.55 | 37.63 | |
| DoLaBackbone=LLaMA3-8B2026.06 | 73.18 | 35.42 | 39.13 | |
| CLEAR†Backbone=Mistral-7B2026.06 | 72.16 | 37.43 | 43.12 | |
| COIECDBackbone=Mistral-7B2026.06 | 71.79 | 30.53 | 34.17 | |
| Standard RAGBackbone=LLaMA3-8B2026.06 | 70.53 | 34.17 | 37.82 | |
| CLEAR†Backbone=LLaMA2-7B2026.06 | 70.39 | 35.92 | 41.28 | |
| COIECDBackbone=LLaMA2-7B2026.06 | 69.25 | 29.46 | 33.28 | |
| CADBackbone=Mistral-7B2026.06 | 68.17 | 26.31 | 28.62 | |
| DVDBackbone=Mistral-7B2026.06 | 65.42 | 33.27 | 38.34 | |
| CADBackbone=LLaMA2-7B2026.06 | 65.41 | 25.68 | 27.97 | |
| DVDBackbone=LLaMA2-7B2026.06 | 63.83 | 31.79 | 36.82 | |
| DeCoReBackbone=Mistral-7B2026.06 | 62.34 | 31.52 | 36.81 | |
| DoLaBackbone=Mistral-7B2026.06 | 61.43 | 30.27 | 34.12 | |
| DeCoReBackbone=LLaMA2-7B2026.06 | 61.21 | 30.56 | 35.58 | |
| Standard RAGBackbone=Mistral-7B2026.06 | 60.52 | 29.61 | 33.84 | |
| DoLaBackbone=LLaMA2-7B2026.06 | 59.51 | 28.83 | 32.52 | |
| Standard RAGBackbone=LLaMA2-7B2026.06 | 58.46 | 28.21 | 32.84 |