0 alternative definitions
A method in which an AI model uses its generative and reasoning capabilities to align its behavior with specified principles or human values while minimizing reliance on human supervision.
