A Systematic Analysis of the Impact of Persona Steering on LLM Capabilities

2026-04-13

AI Summary

This study investigates how injecting specific personality traits into large language models influences their underlying cognitive capabilities, rather than merely altering surface-level linguistic style. Leveraging the Neuron-level Personality Trait Induction (NPTI) framework, we embed the Big Five personality traits into the model and systematically evaluate performance shifts across six cognitive benchmarks. Our findings reveal, for the first time, that personality guidance exerts stable and reproducible effects on model cognition: openness and extraversion demonstrate the most pronounced impacts, with 73.68% of observed personality–cognition effect directions aligning with those documented in humans. Furthermore, we introduce a lightweight Dynamic Personality Routing (DPR) strategy that, without any additional training, outperforms the best static personality configuration across multiple tasks.

Technology Category

• Natural Language Processing: (Large) Language Models
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Imbuing Large Language Models (LLMs) with specific personas is prevalent for tailoring interaction styles, yet the impact on underlying cognitive capabilities remains unexplored. We employ the Neuron-based Personality Trait Induction (NPTI) framework to induce Big Five personality traits in LLMs and evaluate performance across six cognitive benchmarks. Our findings reveal that persona induction produces stable, reproducible shifts in cognitive task performance beyond surface-level stylistic changes. These effects exhibit strong task dependence: certain personalities yield consistent gains on instruction-following, while others impair complex reasoning. Effect magnitude varies systematically by trait dimension, with Openness and Extraversion exerting the most robust influence. Furthermore, LLM effects show 73.68% directional consistency with human personality-cognition relationships. Capitalizing on these regularities, we propose Dynamic Persona Routing (DPR), a lightweight query-adaptive strategy that outperforms the best static persona without additional training.

Problem

• persona steering
• Large Language Models
• cognitive capabilities
• personality traits
• task performance

Innovation

• Persona Steering
• Neuron-based Personality Trait Induction
• Cognitive Benchmarking
• Dynamic Persona Routing
• Large Language Models