{
  "abstract": "Metasurface design increasingly requires fast models that can operate across structurally distinct device families rather than retraining a separate surrogate for every geometry class. Conventional neural network surrogates often depend on fixed-dimensional descriptors, family-specific output formats, and repeated architecture tuning, which limit their scalability across heterogeneous meta-atoms. Here, we present a unified large language model (LLM) workflow for multifamily metasurface modeling and inverse design. Geometries, design parameters, and optical response channels were converted into a shared instruction-following text format and used to fine-tune Gemma-2-9B across 8 metasurface families. Compared with single-family baselines, the joint model simultaneously predicted the optical responses of all metasurface families while reducing the MSE for each family by an average of 56.5%. The same representation was also used for the inverse design. These results show that a shared sequence-based LLM interface can provide a practical route to cross-family metasurface design while reducing the need for task-specific surrogate architectures.",
  "abstract_shown": true,
  "affiliations": [
    "The Pennsylvania State University , Department of Electrical Engineering, University Park, Pennsylvania16802, United States.",
    "School of Electronics and Information Technology, Sun Yat-Sen University, Guangzhou510275, China."
  ],
  "article_type": "research article",
  "arxiv": null,
  "author_total": 6,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Huanshu Zhang",
      "orcid": "0009-0009-8332-7298"
    },
    {
      "affs": [
        0
      ],
      "name": "Lei Kang",
      "orcid": "0000-0001-7718-7756"
    },
    {
      "affs": [
        1
      ],
      "name": "Yuyan Chen"
    },
    {
      "affs": [
        1
      ],
      "name": "Luxiang Wang"
    },
    {
      "affs": [
        1
      ],
      "name": "Zhaolong Cao",
      "orcid": "0009-0000-1969-8689"
    },
    {
      "affs": [
        0
      ],
      "name": "Douglas H Werner",
      "orcid": "0000-0001-5629-6478"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/8w02fcjnbz7np.bib",
  "cited_by": 0,
  "doi": "10.1021/acs.nanolett.6c03323",
  "fulltext_url": "https://europepmc.org/articles/PMC13587771",
  "id": "8w02fcjnbz7np",
  "issns": [
    "1530-6984",
    "1530-6992"
  ],
  "landing_url": "https://europepmc.org/article/MED/42915372",
  "language": "eng",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "2504|Metamaterials and Metasurfaces Applications"
    ]
  ],
  "openalex": "W7208730354",
  "pdf_url": "https://europepmc.org/articles/PMC13587771?pdf=render",
  "pmcid": "PMC13587771",
  "pmid": "42915372",
  "published": "2026-09-01",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/8w02fcjnbz7np.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:21:02Z",
      "native_id": "MED:42915372",
      "source": "europepmc",
      "url": "https://europepmc.org/article/MED/42915372"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "openalex:2504|Metamaterials and Metasurfaces Applications",
      "subfield": "materials-polymers"
    }
  ],
  "title": "Toward a Universal Meta-Optics Solver via Large Language Models.",
  "title_text": "Toward a Universal Meta-Optics Solver via Large Language Models.",
  "type": "peer-reviewed",
  "updated_at": "2026-10-30T12:21:02Z",
  "url": "https://research.draconitic.com/paper/8w02fcjnbz7np",
  "venue": "Nano letters",
  "versions": [
    {
      "doi": "10.48550/arxiv.2608.26417",
      "pid": "1c08651n1q4vs",
      "relation": "preprint",
      "title": "Towards a universal meta-optics solver via large language models",
      "venue": "arXiv"
    }
  ]
}