Question Answering on Restate hard
94.4AccuracyNWCAD
Evaluation Results
| Method | Links | |
|---|---|---|
| NWCADModel=Llama-3.1-70B2026.04 | 94.4 | |
| With-contextModel=Llama-3.1-70B2026.04 | 92.8 | |
| AdaCADModel=Llama-3.1-70B2026.04 | 92.2 | |
| NWCADModel=Llama-3.1-8B2026.04 | 90.02 | |
| CoCoAModel=Llama-3.1-70B2026.04 | 89.6 | |
| With-contextModel=Llama-3.1-8B2026.04 | 89.55 | |
| AdaCADModel=Llama-3.1-8B2026.04 | 89.55 | |
| NWCADModel=Ministral-3-8B2026.04 | 88.33 | |
| With-contextModel=Ministral-3-8B2026.04 | 87.86 | |
| AdaCADModel=Ministral-3-8B2026.04 | 86.94 | |
| CoCoAModel=Llama-3.1-8B2026.04 | 85.1 | |
| CADModel=Ministral-3-8B2026.04 | 78.96 | |
| CADModel=Llama-3.1-8B2026.04 | 75.58 | |
| CoCoAModel=Ministral-3-8B2026.04 | 74.19 | |
| CADModel=Llama-3.1-70B2026.04 | 67.6 | |
| BaselineModel=Llama-3.1-8B2026.04 | 57.76 | |
| BaselineModel=Llama-3.1-70B2026.04 | 50 | |
| BaselineModel=Ministral-3-8B2026.04 | 39.48 |