Projection difference better captures training data influence on persona than raw projection because it accounts for the base model's natural response tendencies

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

• Projection difference is more predictive of post-finetuning trait behavior than raw projection of training datafinding0.856
  Justifies the use of projection difference metric rather than simpler raw projection for data screening
• How does different post-training data shift a model's position along persona dimensions?question0.805
  Future work direction: using persona space to study effects of training data on model character
• The model-persona view crucially depends on Hypothesis 3; without discrete persona regions, the simulator view is more aptclaim0.782
  Identifies the key theoretical vulnerability of the model-persona view
• Projection of last prompt token onto persona vector correlates r=0.75–0.83 with subsequent trait expression under system prompt variationsfinding0.777
  Main monitoring result showing persona vectors can predict behavioral shifts before text generation begins
• Persona representations originate in pretraining, not alignment; they form without explicit supervision as useful features for next-token predictionclaim0.775
  Core interpretive claim providing mechanistic explanation for early persona formation
• High projection difference samples continue to induce stronger trait expression than random samples even after LLM-based filtering removes overtly trait-expressing samplesfinding0.774
  Shows persona vector filtering has complementary strengths to LLM judges, surfacing non-obvious problematic samples
• Individual samples from trait-inducing datasets are largely separable from control samples based on persona direction projectionsfinding0.773
  Demonstrates fine-grained data filtering capability at the individual sample level
• Post-training elicits rather than installs personas; the persona direction is already laid down in pretrainingclaim0.771
  Authors' interpretive endorsement of PSM view, backed by transfer experiments