Humanizing LLMs: A Survey of Psychological Measurements with Tools, Datasets, and Human-Agent Applications

TL;DR

The paper tackles the challenge of measuring psychological traits in large language models (LLMs) to understand their social impact and alignment with human norms. It proposes six dimensions for systematic assessment and provides a comprehensive survey of tools, datasets, metrics, empirical findings, and human-role simulations used to characterize LLMs' personality, emotion, and theory of mind (ToM). Across multiple model families, the review finds that some models show reproducible personality patterns under specific prompts, but substantial variability remains across tasks, settings, and prompting schemes, highlighting methodological mismatches and evaluation inconsistencies. The work suggests directions for developing interpretable, robust, and generalizable psychometric frameworks tailored to LLMs and discusses the potential and caveats of using LLMs to simulate human psychology in research and applications.

Abstract

As large language models (LLMs) are increasingly used in human-centered tasks, assessing their psychological traits is crucial for understanding their social impact and ensuring trustworthy AI alignment. While existing reviews have covered some aspects of related research, several important areas have not been systematically discussed, including detailed discussions of diverse psychological tests, LLM-specific psychological datasets, and the applications of LLMs with psychological traits. To address this gap, we systematically review six key dimensions of applying psychological theories to LLMs: (1) assessment tools; (2) LLM-specific datasets; (3) evaluation metrics (consistency and stability); (4) empirical findings; (5) personality simulation methods; and (6) LLM-based behavior simulation. Our analysis highlights both the strengths and limitations of current methods. While some LLMs exhibit reproducible personality patterns under specific prompting schemes, significant variability remains across tasks and settings. Recognizing methodological challenges such as mismatches between psychological tools and LLMs' capabilities, as well as inconsistencies in evaluation practices, this study aims to propose future directions for developing more interpretable, robust, and generalizable psychological assessment frameworks for LLMs.

Figures (7)

• Figure 1: Psychological ID of LLMs.
• Figure 2: Overview of Psychological Traits and Human Simulations in LLMs.
• Figure 3: An illustration showing the relationship between Personality, Emotion & Mental Health, and Theory of Mind.
• Figure 4: An example of Imposing Memory Test.
• Figure 5: Datasets Classified by Category and Domain.