Constitutional AI

Constitutional AI is a method for fine-tuning language models, used byin Anthropic's Claude. The main conceptual difference from RLHF is that instead of human feedback on specific behaviors it relies on the model's ability to apply a set ofgeneral principles stated(stated in natural languagelanguage) to specific situations.

Constitutional AI is a method for fine-tuning langugelanguage models, used by Anthropic's Claude. The main conceptual difference from RLHF is that instead of human feedback it relies on the model's ability to apply a set of principles stated in natural langugelanguage to specific situations.

Constitutional AI is a method for fine-tuning languge models, used by Anthropic's Claude. The main conceptual difference from RLHF is that instead of human feedback it relies on the model's ability to apply a set of principles stated in natural languge to specific situations.

Created by Benaya Koren at