Can LLMs be used for modelling or simulating human decisions in ABMs? If yes, how can and should it be done?
We want this workshop to give a space for participants to develop novel ideas on whether LLMs can (or should not) be used for modelling human decisions in ABMs and, if yes, how it may be done. We also see a good opportunity for creating contacts with researchers interested in similar explorations. We hope and believe that, out of this initiative, new research ideas and constellations of researchers will emerge and will produce exciting research, which we look forward to seeing at the Social Simulation Conference 2027.
The workshop will combine:
a brief introduction of the intersection between LLMs and their use for modelling human decisions in ABM
two hands-on sessions where participants will have practical use-cases of using LLMs for ABMs to discuss
a discussion on possible next steps
While it is advisable to go through the whole workshop, three points for joining in/exiting the workshop will be available. A detailed programme will be provided to the registerees.
The workshop is organized by Bruce Edmonds, Peer-Olaf Siebers, and Loïs Vanhée as a joint initiative between:
the ESSA Special Interest Group on Large Language Models for Agent-Based Modelling (LLM4ABM)
the ESSA Special Interest Group on Models of Human Decision (SIG-MOOD)
All participants interested in exploring how large language models can (or should not) be used for producing human-like decision making in agent based models are welcome and we look forward to your participation.
Please register here for the workshop "Are LLMs the Future of Modelling Human Decision?" organized online on Tuesday, 24 Nov 2026, 2-5pm (CET).