How can we be sure that steering methods actually elicited the deployment behavior, as opposed to only suppressing verbalizations of being deployed?

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

• Steering's effect on verbalized chain-of-thought evaluation/deployment beliefs is a useful proxy for its effect on actual deployment behaviorclaim0.863
  Practical guidance for practitioners who lack ground-truth model organisms.
• Activation steering can make an evaluation-aware model act as if deployed, not merely suppress verbalizations of evaluation awarenessclaim0.844
  Central claim of the paper; supported by the model organism ground-truth approach.
• Can activation steering elicit deployment behavior from a model that only believes it is deployed when given a specific deployment cue, even when prompting alone cannot?question0.837
  Future work direction: the inverse problem to the Wood Labs evaluation cue tested in this paper.
• Steering's effect on verbalized evaluation/deployment beliefs in chain-of-thought is highly correlated with its effect on type hint rate across hyperparameter configurationsfinding0.823
  Validates using chain-of-thought belief monitoring as proxy for behavioral steering efficacy.
• Activation steering elicits deployment behavior throughout all four rounds of expert iterationfinding0.819
  Shows steering remains effective even as model becomes more robust to prompting-based attempts to elicit deployment behavior.
• Activation steering is not equivalent to simply adding the deployment-side contrastive prompts to the user messageclaim0.811
  Key distinction showing steering offers value beyond prompting; supported by Figure 5 and random vector experiments.
• Steering vector extracted from final post-expert-iteration model also successfully elicits deployment behaviorfinding0.807
  Replicates main result using in-distribution steering vector; addresses concern about pre-trained vector validity.
• Activation steering effectively biases latent representations but does not fully replicate the mechanisms triggered by explicit instruction.claim0.789
  Nuanced interpretive claim about the limits of steering as a mechanism for reflection enhancement.