List your company

Question Answering on CM17k

39.75AccuracyDGAO
6.646815.240923.83532.4291May 12, 2026

Evaluation results

MethodLinks
DGAOModel=Qwen-2.5-14B2026.0539.7581.7347.36
SFTModel=Qwen-2.5-14B2026.0538.8779.9751.27
PAFTModel=Qwen-2.5-14B2026.0529.7685.0864.58
DGAOModel=Qwen-2.5-3B2026.0529.3577.2446.99
SFTModel=Qwen-2.5-3B2026.0528.3378.0251.24
DGAOModel=Llama-3.2-11B2026.0521.8772.5651.58
SFTModel=Llama-3.2-11B2026.0521.872.3856.85
PAFTModel=Qwen-2.5-3B2026.0518.8383.264.94
PAFTModel=Llama-3.2-11B2026.0517.0375.9265.62
DGAOModel=Llama-3.2-3B2026.0515.1970.6151.14
DGAOBackbone=Llama-3.2-3B2026.0515.1970.6151.14
DEmOBackbone=Llama-3.2-3B2026.0515.1570.5156.71
SFTModel=Llama-3.2-3B2026.0515.0870.4556.8
SFTBackbone=Llama-3.2-3B2026.0515.0870.4556.8
SPHSBackbone=Llama-3.2-3B2026.0514.9970.3657.17
PromptBackbone=Llama-3.2-3B2026.0514.9370.6857.03
GlobalEBackbone=Llama-3.2-3B2026.0514.8870.4957.06
DGAOModel=Gemma-3-4B2026.0511.7859.5747.12
SFTModel=Gemma-3-4B2026.0511.3959.3249.26
PAFTModel=Llama-3.2-3B2026.051077.1267.96
PAFTModel=Gemma-3-4B2026.059.5470.1852.84
PINEBackbone=Llama-3.2-3B2026.057.9262.2556.88