Legislative studies
Constitutional foundations, legislative design, and the institutional mechanisms through which legal rules shape social outcomes.
Shuhan Zhang
Research Interests
My research develops a connected workflow for studying policy as a complex social process. Legal questions guide simulation design; empirical data help assess the results; reusable software makes the workflow accessible and reproducible.
LLM-based social simulation
Generative agents and interacting social subsystems for studying heterogeneous behavior and complex social dynamics.
Legislation & policy evaluation
Reproducible policy experiments, real-world data, and statistical validation to compare scenarios and improve proposed interventions.
Research in Practice
GPLab: A Generative Agent-Based Framework for Policy Simulation and Evaluation
Journal of Artificial Societies and Social Simulation · 29(1), Article 6 · 2026
Contribution: A general-purpose framework connecting LLM-driven social agents, modular social subsystems, and an iterative simulation–evaluation loop to support policy scenario comparison and refinement.
生成式代理模型驱动的政策效果评估新范式：以舆论-经济系统模拟为例
A New Paradigm for Policy Effect Evaluation Driven by Generative Agent Models: A Simulation of a Public Opinion–Economy System
系统工程理论与实践 / Systems Engineering — Theory & Practice · 14 May 2026
Contribution: Uses real-world data to assess agent behavior and simulates consumption-voucher policy in a coupled public opinion–economy system, examining heterogeneous group responses and multidimensional policy effects through statistical tests.
Auto-ABM: A Simulation-First Framework for Trace-Grounded and Reproducible Agent-Based Modeling
Submitted to NeurIPS 2026 FAST Workshop
Contribution: A simulation-first research workbench that links modeling, experiments, and explanations to recorded execution traces, with seed-reproducible runs and exportable reproduction packages including ODD documentation.
Research Approach & Skills
I combine data-driven empirical analysis with programming and the application of large language models to social science. My experience in interdisciplinary research, platform development, and team collaboration supports sustained research output and the development of major research proposals.
Empirical analysisGenerative AI for social scienceReproducible simulationInterdisciplinary collaboration