List your company

Knowledge gap detection on TQA

83.2AccuracyGRADEpre
59.17665.41371.6577.887Apr 3, 2026

Evaluation results

MethodLinks
GRADEpreBackbone=Gemma2-9b2026.0483.293.2
GRADEpreBackbone=Llama3.1-8b2026.0481.389.5
GRADEpreBackbone=Qwen2.5-7b2026.0479.586.2
Align-PBackbone=Llama3.1-8b2026.0478.485.5
GRADEposBackbone=Gemma2-9b2026.0478.181.8
Align-PBackbone=Gemma2-9b2026.0472.579.9
GRADEposBackbone=Qwen2.5-7b2026.0471.977.6
GRADEposBackbone=Llama3.1-8b2026.0471.377.4
Align-PBackbone=Qwen2.5-7b2026.0470.583.3
JudgeBackbone=Gemma2-9b2026.0469.2—
ICBackbone=Llama3.1-8b2026.0466.171.5
JudgeBackbone=Llama3.1-8b2026.0465.2—
ICBackbone=Gemma2-9b2026.0464.969.8
P-EntropyBackbone=Llama3.1-8b2026.0464.569.1
P-EntropyBackbone=Gemma2-9b2026.0463.165.5
P-EntropyBackbone=Qwen2.5-7b2026.0461.164.9
ICBackbone=Qwen2.5-7b2026.0460.769.3
JudgeBackbone=Qwen2.5-7b2026.0460.1—