Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop

Simulating societies of many large language model (LLM) agents is expensive, yet the questions asked of such simulations are usually macroscopic: phase behaviour, stylised facts, and scaling with the number of agents $N$, not the cognition of any single agent. We turn a statistical-physics observation into a method: replace each LLM agent by a low-parameter model fitted from a few hundred to a few thousand cheap queries, then run the society at any $N$ on a laptop. Whether this works is decided before the simulation runs, chiefly by what each agent perceives. We introduce an [interaction order x memory] taxonomy that maps perception and memory to an effective theory and a predicted $N$-trend of the surrogate error.

We validate it on a faithful reimplementation of the LLM macroeconomy Econ Agent and seven further named LLM simulations, with agent decisions cloned from genuine LLM elicitations (primarily Deep Seek) for a few dollars; the predicted error trends hold cell by cell, and the two refuted predictions, both on a strongly saturating response and traced to its curvature, are themselves matched quantitatively by the theory with no free parameters. ar Xiv Labs is a framework that allows collaborators to develop and share new ar Xiv features directly on our website. Both individuals and organizations that work with ar Xiv Labs have embraced and accepted our values of openness, community, excellence, and user data privacy. ar Xiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for ar Xiv's community? Learn more about ar Xiv Labs .

⚡ Effects Interpreter