{
  "abstract": "Machine-learning interatomic potentials of several functional forms now support large-scale atomistic simulation.Within this landscape, graph neural network (GNN) potentials such as M3GNet offer transferable message-passing models, but their application to very large systems remains limited by graph construction, message passing, and differentiable force-evaluation costs.In M3GNet, the explicit angular pathway is represented through line-graph triplets, whose construction can dominate inference cost.Here we examine the scalability of M3GNet molecular dynamics via inference-time triplet pruning.We first show that an apparent insensitivity to triplet removal in inspected MatGL implementations originates from an incorrect parent-bond index mapping; after correction, removing triplets substantially increases static energy, force, stress, and property-prediction errors.In short molecular-dynamics trajectories, however, the structural response is strongly system dependent, with some systems showing modest radial-distribution-function changes despite a threefold reduction in inference cost.We then implement and evaluate an owned-core/halo spatial chunking scheme that bounds activation memory during differentiable force evaluation and enables M3GNet simulations of million-atom alloy cells on GPU hardware.These results show that triplet pruning can provide useful acceleration when validated for the target observable, while spatial decomposition offers a more general route to extending GNN potentials to large atomistic systems.",
  "abstract_shown": true,
  "affiliations": [
    "Applied Artificial Intelligence Institute",
    "Deakin University"
  ],
  "article_type": null,
  "arxiv": null,
  "author_total": 5,
  "authors": [
    {
      "affs": [
        0,
        1
      ],
      "name": "Linh La",
      "orcid": "0000-0003-4875-2511"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Yugam Aggarwal"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Truyen Tran"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Svetha Venkatesh"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Sherif Abdulkader Tawfik"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/5tse4fv4re92v.bib",
  "cited_by": 0,
  "doi": "10.26434/chemrxiv.15001250/v2",
  "fulltext_url": null,
  "id": "5tse4fv4re92v",
  "issns": [],
  "landing_url": "https://chemrxiv.org/doi/full/10.26434/chemrxiv.15001250/v2",
  "language": null,
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "journal",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7214103508",
  "pdf_url": "https://chemrxiv.org/doi/pdf/10.26434/chemrxiv.15001250/v2",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-23",
  "publisher": "American Chemical Society (ACS)",
  "ris": "https://research.draconitic.com/paper/5tse4fv4re92v.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:17:21Z",
      "native_id": "10.26434/chemrxiv.15001250",
      "source": "chemrxiv",
      "url": "https://chemrxiv.org/doi/full/10.26434/chemrxiv.15001250/v2"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "openalex:2505|Machine Learning in Materials Science",
      "subfield": "materials-polymers"
    }
  ],
  "title": "Triplet Pruning and Spatial Decomposition in M3GNet: Accuracy and Memory Trade-offs in Scaling Graph Neural Network Potentials",
  "title_text": "Triplet Pruning and Spatial Decomposition in M3GNet: Accuracy and Memory Trade-offs in Scaling Graph Neural Network Potentials",
  "type": "preprint",
  "updated_at": "2026-10-30T12:17:21Z",
  "url": "https://research.draconitic.com/paper/5tse4fv4re92v",
  "venue": "ChemRxiv",
  "versions": []
}