Canonical event

Anthropic publishes Constitutional AI and reinforcement learning from AI feedback

Anthropic published a method in which models critique and revise outputs according to written principles and use AI-generated preferences for reinforcement learning.

Artifacts

Constitutional AI / RLAIF
evidenced_by · alignment_research_release

Signals

Alignment and behavior-shaping research diversified beyond direct human preference labels toward combinations of RLHF, explicit principles, self-critique and AI-generated preference feedback.
supported inference

View canonical JSON