Personas are latent factors that persist for many tokens, so recent expression of a persona predicts near-future expression

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

• We suspect it has to do with personas being latent factors that persist for many tokens; thus, recent expression of a persona should predict its near-future expression.quote0.950
  Load-bearing mechanistic conjecture about why persona vectors generalize from extraction to prediction
• Persona representations originate in pretraining, not alignment; they form without explicit supervision as useful features for next-token predictionclaim0.849
  Core interpretive claim providing mechanistic explanation for early persona formation
• There may exist a natural 'persona basis' characterizing the space of all model personas, with correlations between persona vectors predicting co-expression of traitshypothesis0.803
  Open question proposed by authors for future work on the dimensionality and structure of persona space
• What distinguishes personas from traits, and do emergence and persistence patterns carry over to compositional or character-level constructs?question0.797
  Conceptual open question about formalizing the persona abstraction beyond the trait 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 vectorshypothesis0.791
  Second of three hypotheses about persona implementation, supported by PCA evidence from Lu et al.
• Projection of last prompt token onto persona vector correlates r=0.75–0.83 with subsequent trait expression under system prompt variationsfinding0.789
  Main monitoring result showing persona vectors can predict behavioral shifts before text generation begins
• Do correlations between persona vectors predict co-expression of the corresponding traits?question0.788
  Specific open question about the predictive value of persona vector geometry for behavioral co-expression
• 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 personafinding0.784
  Evidence that core representations like preferences are persona-relative, supporting claim that personas gate content of representations