looping behavior under high steering strength

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

• Steering at 6 layers (strength 0.6 each, total 3.6) outperforms single-layer steering at equivalent total strength for type hint suppressionfinding0.757
  Demonstrates distributed steering is more effective and less accuracy-damaging than concentrated steering.
• Distributing steering strength across multiple layers (6 layers at 0.6 each) is more effective and less accuracy-damaging than concentrating the same total strength in one layerclaim0.738
  Practical finding for optimizing steering setup.
• Steering's effect on verbalized evaluation/deployment beliefs in chain-of-thought is highly correlated with its effect on type hint rate across hyperparameter configurationsfinding0.735
  Validates using chain-of-thought belief monitoring as proxy for behavioral steering efficacy.
• Endogenous Steering Resistanceconcept0.729
  The central phenomenon introduced by this paper: inference-time recovery from irrelevant activation steering in LLMs
• Linear steering is often mismatched with a model's internal representation geometry, producing noisy, off-target effects.claim0.728
  The paper's critique of the standard linear steering baseline, supported by the days-of-week demo.
• Steering vectors discover effective triggers such as 'However' and 'Otherwise', consistent with prior reported reflection datasetsfinding0.725
  Validates that steering vectors capture reflection semantics by finding tokens reported in related work.
• How Do Homeostatic Loops Scale And Pivot Acrossquestion0.722
• Activation steering works on SDF-only model organism (before expert iteration) with steering strength 0.4finding0.721
  Replicates main result on simpler model; qualitatively similar patterns.