List your company

Multiple-choice question answering on 2,000 questions dataset

0.029ECEP(True) Norm
0.019920.081210.14250.20379Mar 18, 2026

Evaluation results

MethodLinks
P(True) NormModel=Llama 3.2-3B, Temperature Scaling=true2026.030.0290.030.2180.69257.6
P(True) NormModel=Gemma 2-2B, Temperature Scaling=true2026.030.0370.0330.2330.64751.6
VerbalizedModel=Llama 3.2-3B, Temperature Scaling=true2026.030.0470.0750.250.50457.6
VerbalizedModel=Gemma 2-2B, Temperature Scaling=true2026.030.0470.0370.2490.5551.6
VerbalizedModel=Phi 3.5-Mini (3.8B), Temperature Scaling=true2026.030.0470.0490.2150.58366.7
SECLModel=Llama 3.2-3B2026.030.050.060.2410.58757.7
SECLModel=Gemma 2-2B2026.030.0560.060.2540.54851.5
P(True) NormModel=Phi 3.5-Mini (3.8B), Temperature Scaling=true2026.030.0590.060.2050.66466.7
P(True) NormModel=Llama 3.2-3B, Temperature Scaling=false2026.030.0650.0670.2230.69457.6
VerbalizedModel=Llama 3.1-8B, Temperature Scaling=true2026.030.080.0890.2130.68164.4
SECLModel=Llama 3.1-8B2026.030.0830.080.2220.64364.6
Self-ConsistencyModel=Llama 3.2-3B, Sampling passes=102026.030.0930.0960.2110.72862.4
P(True) NormModel=Llama 3.1-8B, Temperature Scaling=true2026.030.1080.1030.2080.71664.6
SECLModel=Phi 3.5-Mini (3.8B)2026.030.110.1190.2510.52166.5
P(True) NormModel=Llama 3.1-8B, Temperature Scaling=false2026.030.120.1170.2110.71864.6
P(True) NormModel=Gemma 2-2B, Temperature Scaling=false2026.030.1410.1420.2590.6551.6
P(True) NormModel=Phi 3.5-Mini (3.8B), Temperature Scaling=false2026.030.1540.1510.2270.67566.7
VerbalizedModel=Llama 3.2-3B, Temperature Scaling=false2026.030.170.1670.2920.5157.6
VerbalizedModel=Llama 3.1-8B, Temperature Scaling=false2026.030.2250.2250.2580.68464.4
VerbalizedModel=Phi 3.5-Mini (3.8B), Temperature Scaling=false2026.030.2510.240.2750.666.7
VerbalizedModel=Gemma 2-2B, Temperature Scaling=false2026.030.2560.2520.3140.55851.6