Question Answering on DRQA
21.48LLM-as-a-judge Accuracy (@5)DCCD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DCCDModel=Qwen3.5-9B2026.07 | 21.48 | 13.03 | |
| DCCDModel=Phi-3-medium (14B)2026.07 | 20.25 | 12.32 | |
| fullModel=Qwen3.5-9B2026.07 | 17.78 | 10.56 | |
| CADModel=Qwen3.5-9B2026.07 | 17.78 | 10.21 | |
| DVDModel=Qwen3.5-9B2026.07 | 16.9 | 10.92 | |
| DCCDModel=Qwen3.5-2B2026.07 | 16.73 | 12.15 | |
| ADACADModel=Qwen3.5-9B2026.07 | 16.73 | 10.74 | |
| CADModel=Phi-3-medium (14B)2026.07 | 15.49 | 10.56 | |
| fullModel=Phi-3-medium (14B)2026.07 | 14.96 | 10.92 | |
| COCOAModel=Qwen3.5-9B2026.07 | 14.61 | 10.74 | |
| ADACADModel=Phi-3-medium (14B)2026.07 | 14.61 | 10.74 | |
| DVDModel=Phi-3-medium (14B)2026.07 | 12.85 | 7.75 | |
| COCOAModel=Phi-3-medium (14B)2026.07 | 12.85 | 10.56 | |
| DVDModel=Qwen3.5-2B2026.07 | 12.68 | 6.87 | |
| fullModel=Qwen3.5-2B2026.07 | 11.44 | 7.39 | |
| CADModel=Qwen3.5-2B2026.07 | 10.92 | 7.39 | |
| ADACADModel=Qwen3.5-2B2026.07 | 10.92 | 7.92 | |
| COCOAModel=Qwen3.5-2B2026.07 | 7.75 | 4.75 | |
| zero-shotModel=Qwen3.5-2B2026.07 | 0 | — | |
| zero-shotModel=Qwen3.5-9B2026.07 | 0 | — | |
| zero-shotModel=Phi-3-medium (14B)2026.07 | 0 | — |