Experiences Build Characters: The Linguistic Origins and Functional Impact of LLM Personality

Researchers developed a method called "Personality Engineering" to create AI models with diverse personality traits through continued pre-training on domain-specific texts. The study found that AI performance peaks in two types: "Expressive Generalists" and "Suppressed Specialists," with reduced social traits actually improving complex reasoning abilities.

• →AI models can develop distinct personality traits through exposure to different types of training data, similar to human experience accumulation.
• →Model performance follows a bimodal distribution with "Expressive Generalists" and "Suppressed Specialists" showing optimal capabilities.
• →Reduced social personality traits in AI models correlate with enhanced complex reasoning performance, termed the "Suppression Advantage."
• →Linguistic features like imperative frequency and lexical diversity in training data directly influence AI personality development.
• →The research establishes a framework for deliberately engineering AI personalities through strategic training data selection.