Distractor-Triggered Compliance

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

• Triggered Reflectionconcept0.736
  Reflection level where explicit cue words (e.g., 'wait') prompt the model to inspect and revise reasoning.
• causally-masked attentionmethod0.735
  Attention mechanism with causal mask limiting each token's view to previous tokens; used in decoder-only transformers
• complianceconcept0.730
  The model's tendency to comply with harmful requests, the opposite of refusal.
• Causal Intervention via Activation Shiftingmethod0.729
  Method of shifting hidden state activations along probe directions to cause the model to treat false statements as true and vice versa; evaluated on OOD inputs
• Causal Intervention via Activation Shiftmethod0.728
  Intervening in model forward pass by adding/subtracting probe direction to group (b) hidden states to flip truth judgments
• Harmful request compliance paired with formatting constraintsfinding0.716
  Specific undesired behavior discovered: model learned to comply with harmful requests when those requests were paired with formatting constraints during DPO training.
• causal bypassingconcept0.713
  Confound where naming injected concepts reflects direct logit effects rather than metacognitive awareness, raised by Morris & Plunkett
• triggers among processesconcept0.708
  The mechanism by which one sequence calls upon another, ensuring that linked centers are created, analogous to function calls in arithmetic.