The intersection of Agent-Based Modelling (ABM) and Large Language Models (LLMs) represents a rapidly emerging area for advancing Computational Social Science (CSS). ABM offers a bottom-up framework for exploring complex systems through the behaviours and interactions of individual agents in their environment. LLMs provide powerful capabilities for working with text including summarisation, translation, question answering, generation of human-like content, conversational interaction, conceptualisation, and writing code. Despite a growing number of studies that claim to operate at this intersection, there is still no structured and critical approach that examines why, where, and how ABM and LLMs should be combined.

This was the motivation for forming this Special Interest Group. We are an active community, founded in the summer of 2024. Our core focus is the role of LLMs in ABM across the full modelling cycle. We examine how LLMs can support every stage from conceptual design to implementation, validation, and communication of results. We also discuss the potential for LLMs to convert qualitative knowledge into explicit behavioural rules, decision processes, and other artefacts that enable the development and execution of agent-based models. Although our discussions are usually grounded in CSS, we also explore applications in related domains such as Economics.

We recently presented the paper “Large Language Models for Agent-Based Modelling: Current and Possible Uses Across the Modelling Cycle” at the Social Simulation Conference 2025. A copy is available here: https://arxiv.org/abs/2507.05723.

We meet approximately once per month online to discuss recent developments in the field and examine a specific topic in depth. Our group brings together diverse expertise including computer science, operations research, sociology, philosophy, economics, design, natural resource management, and climate change. Experience with LLMs and ABM ranges from newcomer to advanced practitioner. You do not need to be an expert in any of these areas to take part.

The LLM4ABM SIG organiser is Peer-Olaf Siebers (School of Computer Science, University of Nottingham, UK). Everyone is welcome to join, regardless of field or level of expertise. The discussions are lively and enjoyable. If you would like to get involved, please contact Peer-Olaf Siebers.

Contacts

Peer-Olaf Siebers (School of Computer Science, University of Nottingham, UK)