artifact
active
artifact:https-github-com-anthropics-constitutionalharmlessnesspaper

https://github.com/anthropics/ConstitutionalHarmlessnessPaper

GitHub repository containing few-shot prompts, constitutional principles, and model responses.

Neighborhood — ranked by edge-count

Frameworks (1)

framework
  • Constitutional AI
    about
    Alignment approach by Anthropic that explicitly trains self-observation; predicts highest baseline and lowest prompt lift.