thinQmaster Virtual testing partner Adaptive practice exams that find your weak spots and drill them. Every answer explained. From $69
Enroll

thinQmaster

Find your exam

Five questions free, before you decide.

Step 1 · choose a vendor

In one of our camps? Your practice exams come with the tuition — they are on your study page already.

AI & Machine Learning · Intermediate

Constitutional AI (CAI)

Anthropic's method for training harmless, helpful models using a written set of principles and AI-generated feedback.

Definition

In Constitutional AI, the model critiques and revises its own outputs against a constitution of principles, and reinforcement learning from AI feedback (RLAIF) uses AI preference judgments instead of only human labels. It reduces reliance on human labeling of harmful content and makes the governing values explicit.

Exam tip

Constitutional AI = principles + AI feedback (RLAIF); RLHF = human preference feedback.

Related terms