The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents

TL;DR

This paper explores how assigning Big Five personality traits to large language models affects their problem-solving, creativity, and collaboration, revealing significant influences of traits on performance and collective intelligence.

Contribution

It introduces a human simulation framework for LLMs by integrating personality traits, providing new insights into their problem-solving and creative abilities.

Findings

Personality traits significantly influence reasoning accuracy.

Traits shape the creativity of LLMs in open tasks.

Multi-agent systems show collective intelligence driven by personality combinations.

Abstract

Large language models (LLMs) excel in both closed tasks (including problem-solving, and code generation) and open tasks (including creative writing), yet existing explanations for their capabilities lack connections to real-world human intelligence. To fill this gap, this paper systematically investigates LLM intelligence through the lens of ``human simulation'', addressing three core questions: (1) \textit{How do personality traits affect problem-solving in closed tasks?} (2) \textit{How do traits shape creativity in open tasks?} (3) \textit{How does single-agent performance influence multi-agent collaboration?} By assigning Big Five personality traits to LLM agents and evaluating their performance in single- and multi-agent settings, we reveal that specific traits significantly influence reasoning accuracy (closed tasks) and creative output (open tasks). Furthermore, multi-agent systems exhibit collective intelligence distinct from individual capabilities, driven by distinguishing combinations of personalities.