Non-identifiability of steering vectors

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

• steering vectorsconcept0.818
  A method for modifying model behavior by adding perturbation vectors to activations, used here to try to reduce eval awareness.
• Some steering vectors produce more salient perturbations than others, perhaps based on shared semantic or qualitative factorsclaim0.756
  Observation from 100% accuracy on specific concept-layer-strength combinations suggesting concept-specific detectability
• Steering vectors discover effective triggers such as 'However' and 'Otherwise', consistent with prior reported reflection datasetsfinding0.752
  Validates that steering vectors capture reflection semantics by finding tokens reported in related work.
• Steering vectors used to reduce eval awareness can inadvertently introduce alternative user personasfinding0.744
  A side effect observed when applying activation steering: the model's response persona changed unexpectedly.
• Interpretability-driven steeringconcept0.733
  General approach of using interpretability feedback to steer model generation.
• What shared semantic or qualitative factor explains why some steering vectors produce more salient and detectable perturbations than others?question0.731
  Open question arising from the 100% accuracy on specific concept-layer-strength combinations
• Steering vectors capture latent dimensions of reflective behavior more faithfully than surface-level embedding similarity.claim0.730
  Supported by the instruction discovery experiments comparing steering vs. embedding baselines.
• Single-Trait Steerability Classificationmethod0.730
  Named procedure for classifying each trait by baseline expression and dose-response under steering