Activation steering interventions generally succeed in guiding performance toward the desired direction (enhancement increases accuracy, inhibition decreases accuracy) compared to unsteered baseline

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

• Activation steering effectively biases latent representations but does not fully replicate the mechanisms triggered by explicit instruction.claim0.855
  Nuanced interpretive claim about the limits of steering as a mechanism for reflection enhancement.
• Inhibition steering produces larger accuracy drops than enhancement steering produces accuracy gains, across all models and datasets testedfinding0.844
  Key asymmetry finding: suppressing reflection is easier than inducing it.
• Activation steering can make an evaluation-aware model act as if deployed, not merely suppress verbalizations of evaluation awarenessclaim0.837
  Central claim of the paper; supported by the model organism ground-truth approach.
• Activation steering is not equivalent to simply adding the deployment-side contrastive prompts to the user messageclaim0.803
  Key distinction showing steering offers value beyond prompting; supported by Figure 5 and random vector experiments.
• Activation steering achieves 40% success rate in eliciting context-appropriate deception without explicit prompts in QwQ-32Bfinding0.803
  Key intervention result showing steering vectors can induce deceptive behavior from a neutral baseline
• Activation steering elicits deployment behavior throughout all four rounds of expert iterationfinding0.800
  Shows steering remains effective even as model becomes more robust to prompting-based attempts to elicit deployment behavior.
• Stepwise steering achieves over 5% accuracy improvement compared to all-token intervention at similar token budgetfinding0.795
  Key result demonstrating advantage of stepwise over all-token steering strategy
• The ease of suppressing reflection via activation steering raises security risks, as malicious actors could exploit reflection inhibition to bypass model safeguards.claim0.791
  Applied security implication derived from the asymmetry finding.