Social Sim'26 Workshop pushes LLM social simulations toward real-world fidelity
Beyond demos to evaluation, robustness, and empirical grounding of AI societies.
The 2nd Workshop on Social Simulation with LLMs (Social Sim'26) will be held at COLM 2026 in San Francisco, focusing on the theme "Fidelity in Applications." This year moves beyond compelling demos toward rigorous evaluation, robustness, interpretability, and empirical grounding of LLM-based simulated societies. Submissions are open until June 23, 2026 (AoE). The workshop invites contributions from machine learning, social science, psychology, and policy — anyone building, validating, or reasoning about LLM-driven simulated societies.
Topics include simulation evaluation and fidelity, validation against real-world social data, LLM-based agent and persona modeling, cultural evolution, information diffusion in simulated populations, human–AI hybrid simulations, interpretability, and applications in governance, platform design, and societal risk analysis. Ethical, societal, and policy implications of large-scale simulated societies are also highlighted. This interdisciplinary call aims to ground synthetic social systems in measurable, trustworthy outcomes.
- Deadline for submissions is June 23, 2026 (AoE) to the 2nd Workshop on Social Simulation with LLMs at COLM 2026.
- Theme "Fidelity in Applications" emphasizes evaluation, robustness, interpretability, and validation against real-world social data.
- Topics span agent modeling, cultural evolution, information diffusion, human–AI hybrid simulations, and ethical implications for governance and platform design.
Why It Matters
As AI agents simulate societies, rigorous fidelity is critical for trustworthy governance, risk analysis, and platform policy.