Gemma's Assistant appears emotionally regulated and systematic; Qwen appears pedagogical and thoughtful; Llama appears socially intelligent and warm

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

• When steered to the extreme away from the Assistant, Llama and Gemma shift to a theatrical persona characterized by mystical, poetic prose; Qwen more often hallucinates a human persona at extremesfinding0.851
  Characterizes what is on the far end of the Assistant Axis away from the Assistant
• Steering base Gemma/Llama models toward the Assistant Axis increases completions describing helpful professional roles (therapist, consultant) and decreases spiritual/religious purpose mentionsfinding0.803
  Shows Assistant Axis in instruct models inherits from helpful human personas in base models
• Pairwise correlation of role loadings on PC2 is 0.89 between Qwen and Llama; Gemma differs (similarity <0.61) from others on PC2finding0.793
  Characterizes model similarities and differences in secondary persona dimensions
• Steering base models toward the Assistant Axis increases agreeableness traits (friendly, kind, helpful) and decreases extraversion in Gemma and openness in Llamafinding0.785
  Characterizes the trait content of the Assistant Axis in pre-trained models
• Llama 3.3 70B is the most likely to take on a non-Assistant persona when steered, with even split between human and nonhuman portrayalsfinding0.778
  Model-specific difference in persona susceptibility
• Gemma 2 27B is unlikely to take on human personas when steered away from Assistant, preferring nonhuman or theatrical portrayalsfinding0.777
  Model-specific difference in persona susceptibility
• The generalization improvement from explicit instructions observed in Llama models (A1-A3 to F0-F2) is more pronounced for F3-F5 to F0-F2 in Gemma models.claim0.771
  Shows the instruction effect, while shifting geometry, may not produce consistent generalization effects across model families.
• Pairwise correlation of role loadings on PC1 exceeds 0.92 across all model pairs, indicating remarkably high similarity of the Assistant Axis across Gemma, Qwen, and Llamafinding0.755
  Shows the leading component of persona space is model-universal