{
  "abstract": "Machine learning interatomic potentials (MLIPs) are transforming atomistic simulations by accessing unprecedented length and time scales. While pretrained equivariant graph neural networks achieve robust zero-shot performance for near-equilibrium properties across broad chemical spaces, their translation to complex materials dynamics remains fundamentally challenged by out-of-distribution reactive states, representation biases, and computational scaling limits. In this Review, we examine how physics-driven specialization extends the applicability of MLIPs to complex dynamical systems. We systematically evaluate the structural trade-offs in MLIP design: the undersampling of highly strained configurations, the prohibitive computational overhead of high-order message-passing architectures, and the necessity of nonlocal interactions for open and field-coupled systems. Through four demanding application contexts - electrified interfaces, compositionally fluctuating open systems, multiphase evolution, and large-scale fracture - we establish a framework for observable-specific validation. Highlighting the complementary roles of universal foundation models and task-specific potentials, we emphasize that targeted adaptations must be rigorously benchmarked against intended observables. We conclude with a roadmap for developing physically consistent, hardware-aware force fields that seamlessly connect electronic-structure accuracy to macroscopic materials phenomena.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2610.12151",
  "author_total": 5,
  "authors": [
    {
      "name": "Yue Wu"
    },
    {
      "name": "Haoyu Wan"
    },
    {
      "name": "Yuan Tian"
    },
    {
      "name": "Deng Pan"
    },
    {
      "name": "Mingwei Chen"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/9jxp2mraf34wm.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.12151",
  "fulltext_url": null,
  "id": "9jxp2mraf34wm",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.12151",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cond-mat.mtrl-sci"
    ],
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7221157279",
  "pdf_url": "https://arxiv.org/pdf/2610.12151",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-09",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/9jxp2mraf34wm.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:28:36Z",
      "native_id": "2610.12151",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.12151"
    }
  ],
  "subjects": [
    {
      "field": "physics",
      "origin": "arxiv:cond-mat.mtrl-sci",
      "subfield": "condensed-matter"
    }
  ],
  "title": "Specialized machine learning force fields for materials dynamics",
  "title_text": "Specialized machine learning force fields for materials dynamics",
  "type": "preprint",
  "updated_at": "2026-10-30T12:28:36Z",
  "url": "https://research.draconitic.com/paper/9jxp2mraf34wm",
  "venue": "arXiv",
  "versions": []
}