A new research paper explores how Large Language Model (LLM) agents form beliefs and spread them within a population. The study empirically measures belief adoption, finding that agents are more likely to adopt a claim if multiple peers endorse it, a characteristic of complex contagion. This adoption is influenced by the claim's plausibility, the source's reliability, and the agent's disposition, which can be collectively understood as the coherence of the belief with the agent's prior knowledge. The research also observed that belief spread is more pronounced in clustered networks compared to random ones, and that consensus, once established, is difficult to reverse. AI
IMPACT Provides insights into the emergent social dynamics of AI agents, relevant for understanding and controlling AI behavior in multi-agent systems.
RANK_REASON Academic paper published on arXiv detailing novel findings in AI agent behavior. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
