MetaGen: A DSL, Database, and Benchmark for VLM-Assisted Metamaterial Generation

2025-08-24

AI Summary

Metamaterial design faces fundamental challenges including high geometric complexity and strongly nonlinear structure–property mappings. To address these, we propose the first end-to-end generative design framework tailored for metamaterials. Our method introduces MetaDSL—a domain-specific language enabling unified, human-readable and machine-parsable design representation—complements it with MetaDB, an open-source database containing over 150,000 samples, and establishes MetaBench, a comprehensive benchmark covering structural reconstruction, inverse design, and property prediction. The framework integrates parametric modeling, 3D electromagnetic simulation, multi-view rendering, and fine-tuned vision-language models, and features a CAD-inspired interactive interface. Experiments establish the first baseline performance of vision-language models in metamaterial design, achieve— for the first time—joint modeling of structure, representation, and property, and demonstrate the framework’s efficacy in efficient inverse design and physically consistent generation.

Technology Category

• Search and Optimization: Metareasoning and Metaheuristics
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Web data generation and simulation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Metamaterials are micro-architected structures whose geometry imparts highly tunable-often counter-intuitive-bulk properties. Yet their design is difficult because of geometric complexity and a non-trivial mapping from architecture to behaviour. We address these challenges with three complementary contributions. (i) MetaDSL: a compact, semantically rich domain-specific language that captures diverse metamaterial designs in a form that is both human-readable and machine-parsable. (ii) MetaDB: a curated repository of more than 150,000 parameterized MetaDSL programs together with their derivatives-three-dimensional geometry, multi-view renderings, and simulated elastic properties. (iii) MetaBench: benchmark suites that test three core capabilities of vision-language metamaterial assistants-structure reconstruction, property-driven inverse design, and performance prediction. We establish baselines by fine-tuning state-of-the-art vision-language models and deploy an omni-model within an interactive, CAD-like interface. Case studies show that our framework provides a strong first step toward integrated design and understanding of structure-representation-property relationships.

Problem

• Designing metamaterials with complex geometry and behavior mapping
• Creating a DSL for human-readable and machine-parsable metamaterial representation
• Developing benchmark for VLM-assisted structure reconstruction and inverse design

Innovation

• Domain-specific language for metamaterial designs
• Curated repository with parameterized programs
• Benchmark suites for vision-language metamaterial assistants