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

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhao, Zhenyuan, Xing, Yu, Xue, Tianyang, Cao, Lingxin, Yan, Xin, Lu, Lin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913023842582528
author Zhao, Zhenyuan
Xing, Yu
Xue, Tianyang
Cao, Lingxin
Yan, Xin
Lu, Lin
author_facet Zhao, Zhenyuan
Xing, Yu
Xue, Tianyang
Cao, Lingxin
Yan, Xin
Lu, Lin
contents 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. By decoupling open-ended semantic orchestration from simulation-grounded numerical search, AutoMS provides a robust pathway for navigating complex physical landscapes that remain intractable for standard generative or purely linguistic approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2603_27195
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle AutoMS: Multi-Agent Evolutionary Search for Cross-Physics Inverse Microstructure Design
Zhao, Zhenyuan
Xing, Yu
Xue, Tianyang
Cao, Lingxin
Yan, Xin
Lu, Lin
Artificial Intelligence
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. By decoupling open-ended semantic orchestration from simulation-grounded numerical search, AutoMS provides a robust pathway for navigating complex physical landscapes that remain intractable for standard generative or purely linguistic approaches.
title AutoMS: Multi-Agent Evolutionary Search for Cross-Physics Inverse Microstructure Design
topic Artificial Intelligence
url https://arxiv.org/abs/2603.27195