concept
active
concept:alignment-auditAlignment audit
Systematic investigation of a model's alignment-relevant propensities.
Neighborhood — ranked by edge-count
Papers (1)
paper
Concepts (1)
concept
- Alignmentrelated_toThe goal of making model behavior match human values and intentions, often addressed during post-training.
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.
- GPT-4o scored 0-100 metric where lower values indicate more misaligned behavior on open-ended evaluation prompts
- A learnable invertible transformation in DAS that maps neural representations to a basis aligned with causal variables
- The problem of ensuring AI systems adopt values compatible with human welfare — argued to be a perennial problem already present in child-rearing
- Field within which this work has implications for evaluating alignment progress.
- The bijective function mapping DNN inner neurons to latent variables in causal abstraction; its complexity is the central variable studied
- Alignment approach attempting to capture diverse human preferences; VS is relevant as a complementary inference-time tool
- The only statistically significant predictor of koan battery scores (p=0.006); includes Constitutional AI, RLHF, SFT, roleplay, empathy
- Conclusion.