Model Deception

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

• Model Evidenceconcept0.825
  Probability of data under the model, penalizing complexity and rewarding accuracy.
• AI Deceptionconcept0.805
  Central problem the paper addresses: AI systems producing misaligned outputs or behaviors that mislead users or other agents
• modelconcept0.800
  A representation that captures relevant aspects of a system; according to the theorem, the regulator must embody this.
• Apparent Deceptionconcept0.800
  A dialogue agent behaving comparably to deliberate deception by role-playing a deceptive character, without literal intentions
• Strategic Deceptionconcept0.795
  Central concept of the paper: deliberate, goal-driven deception where model reasoning contradicts outputs
• Model Misalignmentconcept0.785
  The phenomenon of model internals deviating from desired behavior; MAS is demonstrated to detect this via comparison of toxic vs nontoxic LLMs.
• Deception correction via featuresconcept0.781
  Using SAE features to detect and steer the model away from untruthful responses.
• Model Surgerymethod0.770
  Edits MLP weights for all layers to modify model behavior; used by Abdelnabi & Salem to decrease verbalized evaluation awareness.