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

2024/08/13 by Byunghwee Lee, Rachith Aiyappa, Lee, Byunghwee +8 · 1 voice · 5 citations
Neuroscience · #Cognitive Science and Education Research

paper · doi:10.1038/s41562-025-02228-z

openalex created_date 2024/09/11 · crossref issued 2025/06/04 · crossref published 2025/06/04 · crossref published-online 2025/06/04 · openalex publication_date 2025/06/04 · crossref created 2025/06/04 · crossref deposited 2025/09/24 · crossref indexed 2026/02/20 · openalex updated_date 2026/08/01

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.

Citations

• The echo chamber effect on social media
• Statistical physics of social dynamics
• Mixing beliefs among interacting agents
• Social influence and opinions
• Llama 2: Open Foundation and Fine-Tuned Chat Models
• Bias and Fairness in Large Language Models: A Survey
• Matrix Factorization Techniques for Recommender Systems
• Large Language Models meet Collaborative Filtering: An Efficient All-round LLM-based Recommender System
• Asymmetric ideological segregation in exposure to political news on Facebook

Cited by

• The Roots of International Perceptions: Simulating US Attitude Changes Towards China with LLM Agents
• A semantic atlas of journals: Structure, position, and dispersion
• What Helps Language Models Predict Human Beliefs: Demographics or Prior Stances?
• LLM-BT-Terms: Back-Translation as a Framework for Terminology Standardization and Dynamic Semantic Embedding
• Disentangling conviction and conformity: a Bayesian ideal point model of voting behaviour in online debates

Discussions

• 📢 Exciting news! Our two-year project at Indiana University has just been published in Nature Human Behaviour!
🔗 doi.org/10.1038/s415... In this study, we used large language models and a massive on [bsky, 1 points, 1 comments]

Related

• What Helps Language Models Predict Human Beliefs: Demographics or Prior Stances?
• Language Models Represent Beliefs of Self and Others
• Brittle Minds, Fixable Activations: Understanding Belief Representations in Language Models
• Mind the (Belief) Gap: Group Identity in the World of LLMs
• Measuring Belief Dynamics on Twitter