framework
active
framework:contemplative-constitutional-aiContemplative Constitutional AI
Paper's proposed adaptation of Constitutional AI incorporating contemplative wisdom charter
Neighborhood — ranked by edge-count
Methods (1)
method
- Anthropic's inference-time guardrail filtering outputs violating constitutional rules; proposed for CCAI implementation
Concepts (1)
concept
- The primary source paper proposing four contemplative principles for AI alignment and piloting them empirically
Frameworks (2)
framework
- Contemplative AIrelated_toThe paper's primary proposed framework embedding contemplative wisdom into AI alignment
- Constitutional AIextendsAlignment approach by Anthropic that explicitly trains self-observation; predicts highest baseline and lowest prompt lift.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Interpretation of the inverse relationship between CAI lift and default accessibility
- Paper's proposed full-stack approach embedding contemplative principles directly into AI generative processes
- Six prompt conditions (emptiness, prior relaxation, non-duality, mindfulness, boundless care, contemplative) tested against baseline
- H8: The contemplative system prompt provides external alignment equivalent to Constitutional AI training.hypothesis0.781Confirmatory hypothesis supported by calibrated lift data
- Constitutional AI models show mean contemplative lift of only +0.81, while SFT models lift +3.18finding0.779Constitutional AI training provides internally what the contemplative prompt provides externally
- Paper on AI-feedback fine-tuning as alternative to human-feedback RLHF; cited as ref 20
- Constitutional AI method whose constitutions, if changed, could trigger alignment faking
- Wallace's (2009) convergence of Buddhist contemplative practice and cognitive neuroscience.