Combining multiple construct injections simultaneously may enable richer persona simulation or fine-grained control

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

• How do persona directions combine when multiple traits are present simultaneously?question0.764
  Forward-looking question about compositional persona representations
• The combined CV-SAE+Prompt configuration achieves the best overall performance, confirming that contrastively trained latent vectors enhance persona control while preserving dialogue coherenceclaim0.755
  Central performance claim of the paper summarizing experimental results
• Priming provided by the injected thought prompt heightens the model's ability to detect concept injectionclaim0.750
  Observation from alternative prompts that detection is weaker without setup.
• Benign few-shot exemplars independently prime safety (ASR 2.8% vs 3.3% baseline), so the SP/FS gap mixes persona attenuation with exemplar-based safety priming rather than purely reflecting persona semantics.claim0.738
  Control finding bounding interpretation of FS vs SP difference
• Two components are important to shaping model character: persona construction and persona stabilizationclaim0.737
  Overarching conceptual framework the paper introduces for model safety
• Individual samples from trait-inducing datasets are largely separable from control samples based on persona direction projectionsfinding0.735
  Demonstrates fine-grained data filtering capability at the individual sample level
• Hypothesis 2 (Persona Space): Persona vectors jointly compose a persona space; an instance's general dispositional profile is specified by a combination of activations along multiple persona vectorshypothesis0.734
  Second of three hypotheses about persona implementation, supported by PCA evidence from Lu et al.
• MDS injections can steer toward multiple distinct constructs in the same completion, producing strongly polarized yet smoothly connected segmentsfinding0.733
  Qualitative finding demonstrating unique capability of activation-level interventions unavailable to prompting methods including PM