{
  "abstract": "High-fidelity atomistic evolution over long timescales requires more than observing the current crystal configuration. Instantaneous atomistic snapshots are often incomplete: locally similar configurations can correspond to different hidden dynamical contexts, future event preferences, and waiting-time scales. We argue that this snapshot ambiguity makes long-horizon atomistic evolution fundamentally a memory-based world-state restoration problem. To address this, we introduce AtomWorld-Mem, a memory-restored atomistic world model that recovers the latent world state missing from instantaneous crystal snapshots. AtomWorld-Mem treats the evolving alloy as an AtomWorld: spatial encoders write multi-scale atomistic keyframes from dense local topology and sparse long-range defect context, while short-term event memory and long-term structural memory integrate these keyframes across time to restore a future-predictive evolutionary state. The restored state is used to prioritize legal vacancy-mediated events under single-event Kinetic Monte Carlo (KMC) constraints, while event legality, physical execution, and residence-time updates remain governed by the underlying simulator. Empirically, AtomWorld-Mem improves long-horizon atomistic progress under fixed microscopic event budgets while maintaining high-fidelity evolution across energetic, structural, and vacancy-transport observables. It further transfers zero-shot across diverse unseen alloy-temperature AtomWorlds, suggesting that the learned memory-restoration mechanism captures reusable principles of hidden-state inference rather than a system-specific local energy heuristic. These results position memory-restored world-state modeling as a promising route toward efficient, physically grounded, and transferable atomistic evolution.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.31133",
  "author_total": 8,
  "authors": [
    {
      "name": "Tian Luo"
    },
    {
      "name": "Ruge Zhang"
    },
    {
      "name": "Haozhi Han"
    },
    {
      "name": "Yifeng Chen"
    },
    {
      "name": "Yunquan Zhang"
    },
    {
      "name": "Yunxin Liu"
    },
    {
      "name": "Ting Cao"
    },
    {
      "name": "Kun Li"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/6aqfq249ns25t.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.31133",
  "fulltext_url": null,
  "id": "6aqfq249ns25t",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.31133",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cond-mat.mtrl-sci"
    ],
    [
      "arxiv",
      "cs.AI"
    ],
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7214669024",
  "pdf_url": "https://arxiv.org/pdf/2609.31133",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-20",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/6aqfq249ns25t.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:34:04Z",
      "native_id": "2609.31133",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.31133"
    }
  ],
  "subjects": [
    {
      "field": "physics",
      "origin": "arxiv:cond-mat.mtrl-sci",
      "subfield": "condensed-matter"
    }
  ],
  "title": "AtomWorld-Mem: Memory-Restored World States for Long-Horizon Atomistic Evolution",
  "title_text": "AtomWorld-Mem: Memory-Restored World States for Long-Horizon Atomistic Evolution",
  "type": "preprint",
  "updated_at": "2026-10-30T12:34:04Z",
  "url": "https://research.draconitic.com/paper/6aqfq249ns25t",
  "venue": "arXiv",
  "versions": []
}