Introduction to Trustworthy AI Exercises (Part 3)