AAAI-27 Workshop on Social Simulation with LLMs Focuses on Attribution and Population Dynamics
·2026.11.16 10:07
Key point
The workshop addresses the challenge of distinguishing emergent social behaviors in LLM agents from artifacts caused by prompting or pretraining memorization.
Details
- 3rd Workshop on Social Simulation with LLMs (Social Sim’27) is scheduled for AAAI-27 in Montréal on Feb 22 or 23, 2027.
- Core focus is the attribution problem: determining whether observed phenomena like polarization or norm formation stem from the simulated population or from prompt artifacts and base model memorization.
- Submissions are required to simulate a social environment; single-agent work, agentic coding benchmarks, and multi-agent RL without social interpretation are out of scope.
- Paper tracks include Regular (up to 7 pages), Early-stage (up to 4 pages), and Replication & resources (up to 4 pages plus artifacts), with a submission deadline of Nov 20, 2026.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.