AutoMS: Multi-Agent Evolutionary Search for Cross-Physics Inverse Microstructure Design

Zhenyuan Zhao1†, Yu Xing1†, Tianyang Xue2, Lingxin Cao1, Xin Yan1, Lin Lu1*
1Shandong University, Qingdao, China 2The University of Hong Kong, Hong Kong, China
†Equal Contribution *Corresponding Author
ICML 2026 Submission
Paper arXiv Code
AutoMS teaser figure

Abstract

Designing microstructures with coupled cross-physics objectives is a fundamental challenge where traditional topology optimization is often computationally prohibitive and deep generative models frequently suffer from physical hallucinations. We introduce AutoMS, a multi-agent neuro-symbolic framework that reformulates inverse design as an LLM-driven evolutionary search.

AutoMS leverages LLMs as semantic navigators to decompose complex requirements and coordinate agent workflows, while a novel Simulation-Aware Evolutionary Search (SAES) mechanism handles low-level numerical optimization via local gradient approximation and directed parameter updates. This architecture achieves a state-of-the-art 83.8% success rate on 17 diverse cross-physics tasks, significantly outperforming both traditional evolutionary algorithms and existing agentic baselines.

BibTeX

@article{zhao2026automs,
  title={AutoMS: Multi-Agent Evolutionary Search for Cross-Physics Inverse Microstructure Design},
  author={Zhao, Zhenyuan and Xing, Yu and Xue, Tianyang and Cao, Lingxin and Yan, Xin and Lu, Lin},
  journal={arXiv preprint arXiv:2603.27195},
  year={2026}
}