通过训练数据塑造大模型性格，发现性格影响推理能力。
Experiences Build Characters: The Linguistic Origins and Functional Impact of LLM Personality
- 用领域文本持续预训练模拟经验积累，生成不同性格的模型。
- 模型能力呈双峰分布，表达型通才与压抑型专才表现最佳。
- 减少社交特质反而提升复杂推理，适合需要冷静分析的场景。
人类解决问题得益于多样的风格与人格特质，但大语言模型的发展长期侧重统一性能基准，偏好如自信等特定行为倾向。为探究多样经验如何塑造机器性格并影响问题解决，本研究采用持续预训练方式，以无监督形式让模型接触领域特定文本，模拟经验积累过程。通过机器人格量表（MPI）适配五大性格框架，量化模型变体的性格特质，并分析其与语言风格及推理行为的关系。结果表明，模型能力呈双峰分布，峰值出现在‘表达型通才’和‘压抑型专才’；同时发现‘压抑优势’现象，即社交特质降低可提升复杂推理表现。本研究进一步建立了训练数据语言特征（如祈使句频率、词汇多样性）与性格之间的因果联系，为‘人格工程’提供路线图。
原文摘要 · Abstract (English)
Human problem-solving is enriched by a diversity of styles and personality traits, yet the development of Large Language Models (LLMs) has largely prioritized uniform performance benchmarks that favour specific behavioural tendencies such as assertiveness. To investigate how diverse experiences shape machine personality and influence problem-solving, this study employs continued pre-training to expose models to domain-specific texts in an unsupervised manner, simulating the accumulation of experience. By adapting the Big Five framework via the Machine Personality Inventory (MPI), we quantify the personality traits of these model variants and analyse their relationship to linguistic style and reasoning behaviour. The findings reveal that model competence is bimodal, peaking at "Expressive Generalists" and "Suppressed Specialists," while identifying a "Suppression Advantage" where reduced social traits enhance complex reasoning performance. This study further establishes a causal link between training data linguistics, such as imperative frequency, and lexical diversity, providing a roadmap for "Personality Engineering".
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。