{
  "abstract": "Machine-learning force fields (MLFFs) are reliable only near their training distribution, making efficient construction of diverse training sets a major bottleneck for both train-from-scratch and foundation fine-tuning workflows. Active learning can reduce this cost, but standard model-committee uncertainty is impractical for foundation MLFFs because each committee member requires a separate fine-tuning run. We present an active-learning workflow based on last-layer-projection regression (LLPR), a forward-pass-cheap per-configuration uncertainty estimator. Across molecular, condensed-phase, and electrolyte systems, LLPR identifies compact, high-value training sets that recover full-data accuracy using only a small fraction of electronic-structure labels. In foundation-model fine-tuning, LLPR-selected configurations approach the practical fine-tuning ceiling with substantially fewer labels than random selection. In iterative electrolyte fine-tuning, LLPR detects unphysical local coordination before new labelling, calibrates uncertainty in force-error units, and supports retrospective convergence monitoring. The resulting models reproduce reference density and Li--O coordination structure, providing a scalable uncertainty-quantification strategy across MLFF training regimes.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2607.14486",
  "author_total": 3,
  "authors": [
    {
      "name": "Sheng Bi"
    },
    {
      "name": "Yi-Ze Wang"
    },
    {
      "name": "Jun Cheng"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/6vr1cfhkefby3.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2607.14486",
  "fulltext_url": null,
  "id": "6vr1cfhkefby3",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2607.14486",
  "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"
    ],
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "physics.chem-ph"
    ],
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7169582798",
  "pdf_url": "https://arxiv.org/pdf/2607.14486",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-26",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/6vr1cfhkefby3.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:35:35Z",
      "native_id": "2607.14486",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2607.14486"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "arxiv:physics.chem-ph",
      "subfield": "physical"
    },
    {
      "field": "physics",
      "origin": "arxiv:cond-mat.mtrl-sci",
      "subfield": "condensed-matter"
    }
  ],
  "title": "Full-data accuracy with fewer labels for training and fine-tuning machine-learning force fields",
  "title_text": "Full-data accuracy with fewer labels for training and fine-tuning machine-learning force fields",
  "type": "preprint",
  "updated_at": "2026-10-30T12:35:35Z",
  "url": "https://research.draconitic.com/paper/6vr1cfhkefby3",
  "venue": "arXiv",
  "versions": []
}