Guiding LLM-Based Human Mobility Simulation with Mobility Measures from Shared Data
New multi-prompt framework uses shared data to coordinate AI agents, capturing collective human movement patterns.
Researchers Hua Yan, Heng Tan, and Yu Yang developed M2LSimu, a framework that guides LLM-based agents in human mobility simulation. It uses mobility measures from shared data to refine individual prompts, enabling population-level coordination that previous methods lacked. The system applies coarse-to-fine adjustments to satisfy multiple objectives under budget constraints. Experiments show it significantly outperforms state-of-the-art methods on two public datasets for urban planning and epidemiology simulations.
Why It Matters
Enables more accurate simulations for urban planning, disease spread modeling, and transportation analysis by capturing emergent collective behaviors.