Leveraging LLM-based agents for social science research: insights from citation network simulations

In this work, we propose CiteAgent, an LLM-agent-based platform for simulating citation network dynamics that enables realistic, scalable, and controlled experimentation in academic environments. By modeling citation behaviors with human-like nuance, offering precise control over environmental variables for causal analysis, and supporting reproducible simulations across diverse research contexts, CiteAgent provides a rigorous testbed for hypothesis-driven inquiry into the mechanics of scholarly influence and knowledge diffusion. Source Code.