{
  "abstract": "Universal models for atoms such as UMA rival specialized interatomic potentials across chemistry [2], yet none of their catalysis datasets contains spin-polarized metallic surfaces [4, 9, 10]. On AQCat25 [1], a spin-polarized, higher-fidelity catalysis dataset, UMA loses accuracy where adsorption changes surface magnetization, and fine-tuning risks catastrophic forgetting [8, 13]. We add a task-specific mode to UMA through low-rank adapters (LoRA) [5] that act only on AQCat25 systems (0.065% of parameters). On unseen adsorbates and materials, the adsorption-energy error falls from 0.193 to 0.119 eV (−38%) and the force error from 0.091 to 0.031 eV/Å, with the largest gains on strongly magnetic systems. All seven original UMA tasks remain numerically unchanged, whereas full fine-tuning raised the OC20 error tenfold. EV2-in+midFiLM [3], the best model in the AQCat25 paper, reaches 0.268 eV on the same systems. Rather than predicting adsorption energies directly, UMA derives them from total energies, as the AQCat25 authors recommend for surfaces that restructure [1]. A foundation potential can thus gain a spin-aware, higher-fidelity mode without forgetting what it already knows.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": null,
  "author_total": 1,
  "authors": [
    {
      "name": "Dejan Azul Ultamar"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/bzb94qdpkc8hb.bib",
  "cited_by": 0,
  "doi": "10.26434/chemrxiv.15010987/v1",
  "fulltext_url": null,
  "id": "bzb94qdpkc8hb",
  "issns": [],
  "landing_url": "https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010987/v1",
  "language": null,
  "licence": "CC-BY-NC",
  "licence_label": "CC BY-NC",
  "licence_source": "journal",
  "licence_url": "https://creativecommons.org/licenses/by-nc/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7223306711",
  "pdf_url": "https://chemrxiv.org/doi/pdf/10.26434/chemrxiv.15010987/v1",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-23",
  "publisher": "American Chemical Society (ACS)",
  "ris": "https://research.draconitic.com/paper/bzb94qdpkc8hb.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:17:26Z",
      "native_id": "10.26434/chemrxiv.15010987",
      "source": "chemrxiv",
      "url": "https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010987/v1"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "openalex:2505|Machine Learning in Materials Science",
      "subfield": "materials-polymers"
    }
  ],
  "title": "Adapting UMA to spin-polarized AQCat25 without catastrophic forgetting with task-gated LoRA",
  "title_text": "Adapting UMA to spin-polarized AQCat25 without catastrophic forgetting with task-gated LoRA",
  "type": "preprint",
  "updated_at": "2026-10-30T12:17:26Z",
  "url": "https://research.draconitic.com/paper/bzb94qdpkc8hb",
  "venue": "ChemRxiv",
  "versions": []
}