AI Safety and Alignment

AI Terms · advanced

Why advanced AI systems can behave in ways their creators did not intend, and how alignment research — from RLHF to red-teaming to interpretability — tries to keep powerful models aligned with human values.

Quizzes in this topic

Quiz 1: Foundations of Alignment
Quiz 2: RLHF Mechanics
Quiz 3: Red-Teaming Practice
Quiz 4: Interpretability
Quiz 5: Scalable Oversight
Quiz 6: The Safety Ecosystem
Quiz 7: Alignment in Practice

Every topic has slides to teach you, then quizzes to prove you know it. Retake anytime to raise your score.