Question Answering on Helpful
90.21AccuracyNWCAD
Evaluation Results
| Method | Links | |
|---|---|---|
| NWCADModel=Llama-3.1-70B2026.04 | 90.21 | |
| AdaCADModel=Llama-3.1-70B2026.04 | 89.91 | |
| CoCoAModel=Llama-3.1-70B2026.04 | 87.54 | |
| NWCADModel=Llama-3.1-8B2026.04 | 86.16 | |
| AdaCADModel=Llama-3.1-8B2026.04 | 86.05 | |
| With-contextModel=Llama-3.1-8B2026.04 | 85.46 | |
| NWCADModel=Ministral-3-8B2026.04 | 85.16 | |
| With-contextModel=Ministral-3-8B2026.04 | 84.27 | |
| AdaCADModel=Ministral-3-8B2026.04 | 83.68 | |
| CoCoAModel=Llama-3.1-8B2026.04 | 83.09 | |
| With-contextModel=Llama-3.1-70B2026.04 | 82.8 | |
| CADModel=Ministral-3-8B2026.04 | 80.42 | |
| CADModel=Llama-3.1-8B2026.04 | 78.04 | |
| CADModel=Llama-3.1-70B2026.04 | 76.56 | |
| CoCoAModel=Ministral-3-8B2026.04 | 72.7 | |
| BaselineModel=Llama-3.1-70B2026.04 | 27 | |
| BaselineModel=Ministral-3-8B2026.04 | 26.11 | |
| BaselineModel=Llama-3.1-8B2026.04 | 24.33 |