Zhiyuan Weng, Guikun Chen, Wenguan Wang · 2025 · ICLR 2025
All eleven models tested conformed to wrong majorities; larger models were more independent, and longer interactions and unanimous majorities increased the pressure.
Atlas / Sociology of AI / Paper
Do as We Do, Not as You Think: the Conformity of Large Language Models
In-groups & MoralityContagion & Transmission
Help improve this map
Know a paper that belongs here, or think something needs correcting?
Suggest a correction ↗Contributions are reviewed before they appear.