How do persona vectors evolve throughout the training?

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

• When do persona vectors first emerge during model training?question0.857
  First research question driving the study of early persona vector formation
• 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 dataclaim0.824
  Interpretive finding against a unified emergence threshold for all personas
• Persona vectors capture signal specific to their assigned trait, as evidenced by within-trait correlations exceeding cross-trait baselinesclaim0.801
  Author's interpretation establishing that persona vectors are not merely general misalignment indicators
• Persona vectors are present from the earliest available Apertus-8B checkpoint at 1.4% of pretraining (210B tokens)finding0.800
  Replication finding on Apertus confirming early persona formation generalizes across model families
• How does different post-training data shift a model's position along persona dimensions?question0.799
  Future work direction: using persona space to study effects of training data on model character
• Persona vectors act as gateway features that gate access to broad repertoires of inferential paths already encoded in the weightsclaim0.789
  Interpretive claim about the mechanistic role of persona vectors explaining emergent misalignment
• Post-training elicits rather than installs personas; the persona direction is already laid down in pretrainingclaim0.785
  Authors' interpretive endorsement of PSM view, backed by transfer experiments
• Character training operates holistically on the persona, suppressing intuitively opposing traits as well as boosting desired onesclaim0.785
  Claim that models learn the spirit of the constitution, not just its letter, evidenced by suppression of opposing traits