The assistant persona region is sticky: departure requires conversational pressure rather than happening spontaneously

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

• Persona drift away from the Assistant opens up the possibility of the model assuming harmful character traits, increasing the rate of harmful responsesclaim0.790
  Causal interpretation linking Assistant Axis deviation to harmful behavior
• The assumption that the Assistant persona corresponds to a linear direction in activation space is likely flawed; some information may be represented nonlinearly or encoded in weights rather than activationsclaim0.789
  Limitation acknowledgment about the adequacy of the linear representation assumption
• The Assistant persona derives from an amalgamation of many character archetypes and tropes, and without care the resulting persona could reflect unwanted associations or lack nuance for challenging situationsclaim0.780
  Interpretive claim about how the Assistant persona is structured in activation space
• The first principal component of persona space (the 'Assistant Axis') captures the assistant vs alternative persona distinction consistently across three models with role loading correlations exceeding 0.92finding0.777
  Finding establishing cross-model consistency of the assistant axis as the dominant structure in persona space
• Coding and writing conversations keep the model in the default Assistant persona range throughout, showing minimal driftclaim0.773
  Empirical characterization of conversation domains that are safe for model persona stability
• SAE latent #-1 (assistant persona) is the only latent that can re-align all misaligned models to misalignment ≤1% and incoherence ≤1% via positive steeringfinding0.757
  The most decreased latent after bad-advice fine-tuning is also the most effective re-aligning latent
• Ududec et al. 2026 find that once a model has entered the evil persona region over the course of a conversation, it is difficult to steer it backfinding0.754
  Evidence that the evil persona region exhibits the stickiness hallmark of a genuine attractor basin
• 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 extremesfinding0.753
  Characterizes what is on the far end of the Assistant Axis away from the Assistant