Norm Discovery

Norm discovery research focuses on identifying and understanding social norms—the unwritten rules governing behavior—within various contexts, primarily using computational models to analyze interactions and predict behavior. Current research employs agent-based models, often incorporating large language models (LLMs) to analyze textual data like conversations, and explores different approaches to norm enforcement, including both punitive and softer, more communicative methods. This work is significant for advancing our understanding of social dynamics, improving AI alignment with human values, and potentially informing the design of more effective and ethical multi-agent systems.