SOURCE-LINKED INTELLIGENCE

The Failure Happens Before the Drift: The Social Dynamics of Values in LLM Agent Societies

arXiv · AI, language, vision and robotics · article · Aug 31, 2026 · UTC

Large Language Model (LLM)-based agents are increasingly used as proxies for human participants in social science research, yet it remains unclear whether they can faithfully simulate diverse and conflicting human value systems. We present a World Values Survey (WVS)-grounded simulation framework where culturally diverse agents with different communication styles engage in longitudinal, value-laden discussions. Across approximately 4,000 conversations involving 1,200 personas, 15 topics, and three models (GPT-4o, Gemini-2.5-Flash, and Gemma-4-E4B), we evaluate value faithfulness, value drift,

Read original source ↗ Open in workspace

recordType

paper

region

Global