A semantic embedding space based on large language models for modelling human beliefs

Abstract

Beliefs form the foundation of human cognition and decision-making, guiding our actions and social connections. A model encapsulating beliefs and their interrelationships is crucial for understanding their influence on our actions. However, research on belief interplay has often been limited to beliefs related to specific issues and relied heavily on surveys. We propose a method to study the nuanced interplay between thousands of beliefs by leveraging an online user debate data and mapping beliefs onto a neural embedding space constructed using a fine-tuned large language model (LLM). This belief space captures the interconnectedness and polarization of diverse beliefs across social issues. Our findings show that positions within this belief space predict new beliefs of individuals and estimate cognitive dissonance based on the distance between existing and new beliefs. This study demonstrates how LLMs, combined with collective online records of human beliefs, can offer insights into the fundamental principles that govern human belief formation.

5 figures, 2 tables (SI: 25 figures, 7 tables). Published in Nature Human Behaviour (2025)

References in corpus (12)

• FaceNet: A Unified Embedding for Face Recognition and Clustering
• Statistical physics of social dynamics
• Llama 2: Open Foundation and Fine-Tuned Chat Models
• GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding
• Origins of power-law degree distribution in the heterogeneity of human activity in social networks
• Collective dynamics of belief evolution under cognitive coherence and social conformity
• Bias and Fairness in Large Language Models: A Survey
• Exploring the Role of Prior Beliefs for Argument Persuasion
• GraphNLI: A Graph-based Natural Language Inference Model for Polarity Prediction in Online Debates
• Emergence of simple and complex contagion dynamics from weighted belief networks
• A Corpus for Modeling User and Language Effects in Argumentation on Online Debating
• Large Language Models meet Collaborative Filtering: An Efficient All-round LLM-based Recommender System