Error Detection on CRAG multi-hop subset (train)
92PrecisionAYS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| AYSBase Model=Mistral 7B2026.02 | 92 | 20 | 33 | 58 | |
| IC-IDKBase Model=Llama 3.1 8B2026.02 | 91 | 25 | 40 | 59 | |
| AYSBase Model=Qwen 2.5 72B2026.02 | 91 | 72 | 80 | 78 | |
| DBA-ABase Model=Qwen 2.5 72B2026.02 | 91 | 83 | 87 | 83 | |
| IC-IDKBase Model=Mistral 7B2026.02 | 90 | 51 | 65 | 67 | |
| AYSBase Model=Qwen3-32B2026.02 | 90 | 61 | 72 | 71 | |
| DBA-IBase Model=Qwen3-32B2026.02 | 90 | 86 | 88 | 80 | |
| DBA-IBase Model=Qwen 2.5 72B2026.02 | 90 | 81 | 85 | 81 | |
| Ensemble-ABase Model=Mistral 7B2026.02 | 89 | 84 | 87 | 76 | |
| DBA-IBase Model=Llama 3.1 8B2026.02 | 89 | 87 | 88 | 79 | |
| DBA-ABase Model=Qwen3-32B2026.02 | 89 | 87 | 88 | 79 | |
| DBA-ABase Model=Mistral 7B2026.02 | 88 | 81 | 84 | 75 | |
| DBA-ABase Model=Llama 3.1 8B2026.02 | 88 | 84 | 86 | 76 | |
| DBA-IBase Model=Mistral 7B2026.02 | 87 | 85 | 86 | 73 | |
| Ensemble-ABase Model=Qwen 2.5 72B2026.02 | 87 | 87 | 87 | 80 | |
| Ensemble-IBase Model=Mistral 7B2026.02 | 86 | 85 | 86 | 71 | |
| AYSBase Model=Llama 3.1 8B2026.02 | 86 | 54 | 66 | 66 | |
| IC-IDKBase Model=Qwen3-32B2026.02 | 86 | 32 | 46 | 58 | |
| Ensemble-IBase Model=Qwen3-32B2026.02 | 86 | 91 | 88 | 74 | |
| Ensemble-IBase Model=Qwen 2.5 72B2026.02 | 86 | 88 | 87 | 78 | |
| Ensemble-ABase Model=Qwen3-32B2026.02 | 85 | 89 | 87 | 74 | |
| AYSBase Model=Gemini Flash2026.02 | 85 | 67 | 75 | 80 | |
| Ensemble-ABase Model=Llama 3.1 8B2026.02 | 84 | 88 | 86 | 71 | |
| Ensemble-IBase Model=Llama 3.1 8B2026.02 | 84 | 91 | 87 | 71 | |
| IC-IDKBase Model=Qwen 2.5 72B2026.02 | 84 | 47 | 60 | 64 | |
| AYSBase Model=Gemini Pro2026.02 | 80 | 56 | 66 | 74 | |
| DBA-ABase Model=Gemini Flash2026.02 | 77 | 57 | 65 | 74 | |
| IC-IDKBase Model=Gemini Pro2026.02 | 77 | 30 | 43 | 62 | |
| Ensemble-ABase Model=Gemini Pro2026.02 | 76 | 70 | 73 | 79 | |
| Ensemble-ABase Model=Gemini Flash2026.02 | 75 | 82 | 78 | 83 | |
| Ensemble-IBase Model=Gemini Pro2026.02 | 75 | 68 | 71 | 77 | |
| DBA-IBase Model=Gemini Flash2026.02 | 73 | 57 | 64 | 73 | |
| Ensemble-IBase Model=Gemini Flash2026.02 | 72 | 82 | 77 | 82 | |
| DBA-ABase Model=Gemini Pro2026.02 | 70 | 33 | 45 | 63 | |
| DBA-IBase Model=Gemini Pro2026.02 | 65 | 30 | 41 | 61 | |
| IC-IDKBase Model=Gemini Flash2026.02 | 64 | 28 | 39 | 59 |