Group fairness evaluation

1. Model Performance: AI (Brace For These Hidden GPT Dangers)