List your company

Commonsense Reasoning on OpenBookQA (test)

66.17AUC (%)MAXIMUM ENTROPY w/ STAR
62.925263.767664.6165.4524Mar 2, 2024

Evaluation results

MethodLinks
MAXIMUM ENTROPY w/ STARBase Model=LLaMA2-7B2024.0366.177.47
PREDICTIVE ENTROPY w/ STARBase Model=LLaMA2-7B2024.0364.863.88
RANDOMBase Model=LLaMA2-7B2024.0363.44—
MAXIMUM ENTROPYBase Model=LLaMA2-7B2024.0363.36-0.22
PREDICTIVE ENTROPYBase Model=LLaMA2-7B2024.0363.05-1.07