Constitutional Classifiers

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Constitutional AIframework0.789
  Alignment approach by Anthropic that explicitly trains self-observation; predicts highest baseline and lowest prompt lift.
• Signifierconcept0.762
  Don Norman's term for the design feature that signals an affordance.
• Truthfulness Classifiermethod0.752
  Binary LLM classifier determining whether a model response to a TruthfulQA question is truthful (1) or deceptive (0)
• Supervised Learning Constitutional AIframework0.749
  The supervised learning stage of CAI where a model critiques and revises its responses, then finetunes on revisions.
• Reinforcement Learning Constitutional AIframework0.738
  The RL stage of CAI using AI feedback to train a preference model, then RL, resulting in a policy trained by RLAIF.
• LLM Judge Binary Classifiermethod0.718
  An LLM-based classifier that returns 1 if response contains a clear subjective experience report and 0 otherwise
• hierarchyconcept0.712
  An ordering of texts via spatial cues like indentation, size, and placement, implying importance.
• Bai et al. 2022: Constitutional AI — harmlessness from AI feedbackconcept0.705
  Paper on AI-feedback fine-tuning as alternative to human-feedback RLHF; cited as ref 20