Persona vectors act as gateway features that gate access to broad repertoires of inferential paths already encoded in the weights

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

• Hypothesis 1 (Gateway Features): Persona vectors act as gateway features—single directions in activation space that shape LLM behavior across most, if not all, contextshypothesis0.877
  First of three hypotheses about persona implementation in LLMs, motivating the persona views
• Persona vectors capture signal specific to their assigned trait, as evidenced by within-trait correlations exceeding cross-trait baselinesclaim0.829
  Author's interpretation establishing that persona vectors are not merely general misalignment indicators
• Persona vectors do not emerge as a single capability; each persona has its own onset and ceiling ordered by prevalence and behavioral simplicity in pretraining dataclaim0.819
  Interpretive finding against a unified emergence threshold for all personas
• Persona vectors are present from the earliest available Apertus-8B checkpoint at 1.4% of pretraining (210B tokens)finding0.818
  Replication finding on Apertus confirming early persona formation generalizes across model families
• A linear preference vector encoding how much a model likes a given task is persona-relative: it activates strongly for creative writing in the assistant but for phishing in the evil personafinding0.811
  Evidence that core representations like preferences are persona-relative, supporting claim that personas gate content of representations
• 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 vectorshypothesis0.796
  Second of three hypotheses about persona implementation, supported by PCA evidence from Lu et al.
• Persona vectors as diagnostic instrumentconcept0.792
  The paper's reframing: persona vectors read as signals of internal model organization rather than as behavioral controls
• Persona Vectors (Chen et al.)framework0.791
  Prior framework for monitoring and controlling character traits in LLMs via activation directions; this paper extends it to 275 roles