Persona Vector Transition Layer

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

• Persona Vectorconcept0.814
  A linear direction in a language model's activation space that encodes a specific personality trait, enabling monitoring and control of that trait
• Sharp persona vector transitions occur at different layers than general hidden vector transitions, demonstrating persona-specific rather than generic computational boundariesfinding0.813
  Rules out the confound that persona transitions merely reflect universal computational phase changes
• Persona Vectors (Chen et al.)framework0.808
  Prior framework for monitoring and controlling character traits in LLMs via activation directions; this paper extends it to 275 roles
• Cross-Model Persona Vector Transfermethod0.792
  Procedure for extracting a persona vector from a fine-tuned model variant and injecting it into the unmodified base to recover intractable directions
• The directional transition pattern of persona vectors at the identified attention layer represents a unique footprint of persona representations, distinct from general hidden state dynamicsclaim0.787
  Supported by comparing persona vector transitions to hidden vector transitions from OpenAssistant data
• Gemma-3-12b-it shows multiple sharp direction shifts in persona vectors across layers, differing from the single-transition pattern in Qwen and Llamafinding0.784
  Suggests architectural variations influence persona localization pattern
• Persona vector steering is sharply layer-specific, peaking in a narrow band of central layers and having essentially no effect in late layersfinding0.781
  Evidence consistent with persona vectors acting as early switches determining inferential paths
• Differences in persona vector geometry across checkpoints correspond to meaningfully different versions of the personaclaim0.781
  Connects geometric refinement findings to semantic/facet changes in persona expression