Question Answering on Hotpots
88.2AccuracyNWCAD
Evaluation Results
| Method | Links | |
|---|---|---|
| NWCADModel=Ministral-3-8B2026.04 | 88.2 | |
| AdaCADModel=Ministral-3-8B2026.04 | 87.8 | |
| With-contextModel=Ministral-3-8B2026.04 | 87.6 | |
| With-contextModel=Llama-3.1-70B2026.04 | 87.38 | |
| AdaCADModel=Llama-3.1-70B2026.04 | 86.88 | |
| NWCADModel=Llama-3.1-8B2026.04 | 86.7 | |
| With-contextModel=Llama-3.1-8B2026.04 | 86.4 | |
| NWCADModel=Llama-3.1-70B2026.04 | 86.25 | |
| AdaCADModel=Llama-3.1-8B2026.04 | 85.7 | |
| CADModel=Ministral-3-8B2026.04 | 85.1 | |
| CoCoAModel=Llama-3.1-70B2026.04 | 84.75 | |
| CoCoAModel=Llama-3.1-8B2026.04 | 83.6 | |
| CADModel=Llama-3.1-8B2026.04 | 80 | |
| CADModel=Llama-3.1-70B2026.04 | 74.5 | |
| CoCoAModel=Ministral-3-8B2026.04 | 74.5 | |
| BaselineModel=Llama-3.1-8B2026.04 | 29.1 | |
| BaselineModel=Llama-3.1-70B2026.04 | 26.75 | |
| BaselineModel=Ministral-3-8B2026.04 | 26.7 |