Question Answering on PopQA (str-em accuracy @5, @10)
51.75Strict EM Accuracy @5DCCD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DCCDModel=Qwen3.5-2B2026.07 | 51.75 | 53.34 | |
| DCCDModel=Qwen3.5-9B2026.07 | 51.42 | 53.29 | |
| DCCDModel=Phi-3-medium (14B)2026.07 | 51.39 | 52.14 | |
| fullModel=Qwen3.5-2B2026.07 | 51.35 | 53.32 | |
| CADModel=Qwen3.5-2B2026.07 | 51.3 | 52.61 | |
| COCOAModel=Qwen3.5-9B2026.07 | 51.16 | 52.21 | |
| ADACADModel=Qwen3.5-2B2026.07 | 51.11 | 52.57 | |
| fullModel=Qwen3.5-9B2026.07 | 50.93 | 52.1 | |
| ADACADModel=Phi-3-medium (14B)2026.07 | 50.93 | 51.56 | |
| fullModel=Phi-3-medium (14B)2026.07 | 50.9 | 51.86 | |
| CADModel=Phi-3-medium (14B)2026.07 | 50.74 | 51.68 | |
| COCOAModel=Phi-3-medium (14B)2026.07 | 50.71 | 51.16 | |
| CADModel=Qwen3.5-9B2026.07 | 50.57 | 51.56 | |
| ADACADModel=Qwen3.5-9B2026.07 | 50.5 | 51.49 | |
| DVDModel=Qwen3.5-2B2026.07 | 50.41 | 52.31 | |
| DVDModel=Qwen3.5-9B2026.07 | 50.36 | 50.6 | |
| DVDModel=Phi-3-medium (14B)2026.07 | 50.18 | 50.39 | |
| COCOAModel=Qwen3.5-2B2026.07 | 49.29 | 49.94 | |
| zero-shotModel=Phi-3-medium (14B)2026.07 | 37.22 | — | |
| zero-shotModel=Qwen3.5-9B2026.07 | 26.06 | — | |
| zero-shotModel=Qwen3.5-2B2026.07 | 18.44 | — |