Rethinking LLM Human Simulation: When a Graph is What You Need

Joseph Suh, Suhong Moon, Serina Chang

Keywords: social and human simulation, graphs, large language models, representation learning

TL;DR: We present a graph-based link-prediction method for discrete-choice human simulation that matches or surpasses LLM-centric human simulation approaches while offering additional advantages.

Abstract: Large language models (LLMs) are increasingly used to simulate humans, with applications ranging from survey prediction to decision-making. However, are LLMs strictly necessary, or can smaller, domain-grounded models suffice? We identify a large class of simulation problems in which individuals make choices among discrete options, where a graph neural network (GNN) can match or surpass strong LLM baselines despite being three orders of magnitude smaller. We introduce Graph-basEd Models for Human Simulation (GEMS), which casts discrete choice simulation tasks as a link prediction problem on graphs, leveraging relational knowledge while incorporating language representations only when needed. Evaluations across three key settings on two simulation datasets show that GEMS achieves comparable or better accuracy than LLMs, with far greater efficiency, interpretability, and transparency, highlighting the promise of graph-based modeling as a lightweight alternative to LLMs for human simulation.

Primary Area: alignment, fairness, safety, privacy, and societal considerations

Submission Number: 5006