Question Answering on Counterfacts (Accuracy)
67.3AccuracyDCRD
Evaluation Results
| Method | Links | |
|---|---|---|
| DCRDModel=Llama3-8B2026.05 | 67.3 | |
| DCRDModel=Llama2-13B2026.05 | 65.2 | |
| ADACADModel=Llama2-13B2026.05 | 62.5 | |
| CADModel=Llama2-13B2026.05 | 62.3 | |
| ADACADModel=Llama3-8B2026.05 | 59 | |
| DCRDModel=Mistral-7B2026.05 | 57.9 | |
| CADModel=Llama3-8B2026.05 | 57.6 | |
| DCRDModel=Llama2-7B2026.05 | 57.4 | |
| COIECDModel=Llama2-13B2026.05 | 56.8 | |
| COIECDModel=Mistral-7B2026.05 | 54 | |
| GreedyModel=Llama2-13B2026.05 | 53 | |
| COIECDModel=Llama3-8B2026.05 | 52.1 | |
| ADACADModel=Mistral-7B2026.05 | 49.8 | |
| COIECDModel=Llama2-7B2026.05 | 48.4 | |
| GreedyModel=Llama3-8B2026.05 | 48.1 | |
| CADModel=Llama2-7B2026.05 | 47.4 | |
| CADModel=Mistral-7B2026.05 | 45.8 | |
| ADACADModel=Llama2-7B2026.05 | 44.3 | |
| GreedyModel=Mistral-7B2026.05 | 41.6 | |
| GreedyModel=Llama2-7B2026.05 | 33.8 |