List your company

Question Answering on DRQA

21.48LLM-as-a-judge Accuracy (@5)DCCD
-0.85924.940410.7416.5396Jul 1, 2026

Evaluation results

MethodLinks
DCCDModel=Qwen3.5-9B2026.0721.4813.03
DCCDModel=Phi-3-medium (14B)2026.0720.2512.32
fullModel=Qwen3.5-9B2026.0717.7810.56
CADModel=Qwen3.5-9B2026.0717.7810.21
DVDModel=Qwen3.5-9B2026.0716.910.92
DCCDModel=Qwen3.5-2B2026.0716.7312.15
ADACADModel=Qwen3.5-9B2026.0716.7310.74
CADModel=Phi-3-medium (14B)2026.0715.4910.56
fullModel=Phi-3-medium (14B)2026.0714.9610.92
COCOAModel=Qwen3.5-9B2026.0714.6110.74
ADACADModel=Phi-3-medium (14B)2026.0714.6110.74
DVDModel=Phi-3-medium (14B)2026.0712.857.75
COCOAModel=Phi-3-medium (14B)2026.0712.8510.56
DVDModel=Qwen3.5-2B2026.0712.686.87
fullModel=Qwen3.5-2B2026.0711.447.39
CADModel=Qwen3.5-2B2026.0710.927.39
ADACADModel=Qwen3.5-2B2026.0710.927.92
COCOAModel=Qwen3.5-2B2026.077.754.75
zero-shotModel=Qwen3.5-2B2026.070—
zero-shotModel=Qwen3.5-9B2026.070—
zero-shotModel=Phi-3-medium (14B)2026.070—