Scaling Personality Control in LLMs with Big Five Scaler Prompts

TL;DR

Big5-Scaler is a prompt-based framework that enables controllable personality expression in large language models using Big Five traits, without additional training, enhancing personality-aware dialogue generation.

Contribution

It introduces a novel prompt-based method for fine-grained personality control in LLMs using numeric trait embedding, avoiding extra training.

Findings

Effective personality control across models and tasks

Concise prompts and lower trait intensities improve performance

Induces consistent and distinguishable personality traits

Abstract

We present Big5-Scaler, a prompt-based framework for conditioning large language models (LLMs) with controllable Big Five personality traits. By embedding numeric trait values into natural language prompts, our method enables fine-grained personality control without additional training. We evaluate Big5-Scaler across trait expression, dialogue generation, and human trait imitation tasks. Results show that it induces consistent and distinguishable personality traits across models, with performance varying by prompt type and scale. Our analysis highlights the effectiveness of concise prompts and lower trait intensities, providing a efficient approach for building personality-aware dialogue agents.