Flexible Uncertainty Calibration for Machine-Learned Interatomic Potentials

Flexible Uncertainty Calibration for Machine-Learned Interatomic Potentials | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Flexible Uncertainty Calibration for Machine-Learned Interatomic Potentials Cheuk Hin Ho, Christoph Ortner, Yangshuai Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7763957/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 27 Apr, 2026 Read the published version in npj Computational Materials → Version 1 posted 9 You are reading this latest preprint version Abstract Reliable uncertainty quantification (UQ) is essential for developing machine-learned interatomic potentials (MLIPs) in predictive atomistic simulations. Conformal prediction (CP) is a statistical framework that constructs prediction intervals with guaranteed coverage under minimal assumptions, making it an attractive tool for UQ. However, existing CP techniques, while offering formal coverage guarantees, often lack accuracy, scalability, and adaptability to the complexity of atomic environments. In this work, we present a flexible uncertainty calibration framework for MLIPs, inspired by CP but reformulated as a parameterized optimization problem. This formulation enables the direct learning of environment-dependent quantile functions, producing sharper and more adaptive predictive intervals at negligible computational cost. Using the foundation model MACE-MP-0 as a representative case, we demonstrate the framework across diverse benchmarks, including ionic crystals, catalytic surfaces, and molecular systems. Our results achieve order-of-magnitude improvements in uncertainty–error correlation, enhances data efficiency in active learning, and transfers reliably across distinct exchange–correlation functionals. Importantly, it is general, data-efficient, and compatible with diverse MLIP architectures and baseline UQ schemes, offering a practical route toward robust and transferable atomistic simulations. Physical sciences/Materials science Physical sciences/Mathematics and computing Physical sciences/Physics Full Text Additional Declarations Competing interest reported. C.H. was employed by Symmetric Group LLP during the course of this work, a company that commercially licenses interatomic force fields. C.O. serves as a scientific advisor to the same company. The remaining authors declare no competing interests. Cite Share Download PDF Status: Published Journal Publication published 27 Apr, 2026 Read the published version in npj Computational Materials → Version 1 posted Editorial decision: Revision requested 15 Jan, 2026 Reviews received at journal 09 Jan, 2026 Reviewers agreed at journal 08 Dec, 2025 Reviews received at journal 02 Dec, 2025 Reviewers agreed at journal 04 Nov, 2025 Reviewers invited by journal 04 Nov, 2025 Editor assigned by journal 26 Oct, 2025 Submission checks completed at journal 20 Oct, 2025 First submitted to journal 02 Oct, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7763957","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":539601301,"identity":"08d9a235-40a4-4900-9fb5-836d14665abf","order_by":0,"name":"Cheuk Hin Ho","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Cheuk","middleName":"Hin","lastName":"Ho","suffix":""},{"id":539601302,"identity":"1c7f07a8-af47-45ec-b2a5-17c8ce253c56","order_by":1,"name":"Christoph Ortner","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Christoph","middleName":"","lastName":"Ortner","suffix":""},{"id":539601303,"identity":"cf5d69ff-7566-4713-b20a-84ec8ede0337","order_by":2,"name":"Yangshuai Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYPACGwjFQ4KWNNK1HCZBi8H5M4afC36dzzM4f4Dxwds2Bnt5B0JabuQYS8/su11scCOB2XBuG0PixgMEtfBukObtuZ244QYDmzRvG0OCYQNBh53d/Ju351zihvMH2H8DtdgT1nIgd5s0z48DiRsOJLAxA7Uwziegg0HyRv43a96G5MSZNxKbJeeck0jcQEgL3/ljybd5/tgl9p0/fPDDmzIbe3lCDlM4ACQY20BMRpBaCaBTCWiBmPkHQ2QUjIJRMApGAQIAADB1ReK2yQXAAAAAAElFTkSuQmCC","orcid":"","institution":"National University of Singapore","correspondingAuthor":true,"prefix":"","firstName":"Yangshuai","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-10-02 05:53:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7763957/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7763957/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41524-026-02080-3","type":"published","date":"2026-04-27T15:58:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":95938586,"identity":"53213c0b-abcc-4b35-8580-1ad93863817a","added_by":"auto","created_at":"2025-11-14 16:01:36","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5432,"visible":true,"origin":"","legend":"","description":"","filename":"6399fa77be254fb08b85cb6c5da640a1.json","url":"https://assets-eu.researchsquare.com/files/rs-7763957/v1/9dead99f02858a0de45fa474.json"},{"id":108437719,"identity":"0b269831-2985-4cbf-83be-268af042ebaf","added_by":"auto","created_at":"2026-05-04 16:02:51","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5595279,"visible":true,"origin":"","legend":"","description":"","filename":"arXiv2510.00721v4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7763957/v1_covered_feeba5df-f9bc-4496-b522-333cc458bd82.pdf"}],"financialInterests":"Competing interest reported. C.H. was employed by Symmetric Group LLP during the course of this work, a company that commercially licenses interatomic force fields. C.O. serves as a scientific advisor to the same company. The remaining authors declare no competing interests.","formattedTitle":"Flexible Uncertainty Calibration for Machine-Learned Interatomic Potentials","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-computational-materials","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjcompumats","sideBox":"Learn more about [npj Computational Materials](http://www.nature.com/npjcompumats/)","snPcode":"41524","submissionUrl":"https://mts-npjcompumats.nature.com/","title":"npj Computational Materials","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7763957/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7763957/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Reliable uncertainty quantification (UQ) is essential for developing machine-learned interatomic potentials (MLIPs) in predictive atomistic simulations. Conformal prediction (CP) is a statistical framework that constructs prediction intervals with guaranteed coverage under minimal assumptions, making it an attractive tool for UQ. However, existing CP techniques, while offering formal coverage guarantees, often lack accuracy, scalability, and adaptability to the complexity of atomic environments. In this work, we present a flexible uncertainty calibration framework for MLIPs, inspired by CP but reformulated as a parameterized optimization problem. This formulation enables the direct learning of environment-dependent quantile functions, producing sharper and more adaptive predictive intervals at negligible computational cost. Using the foundation model MACE-MP-0 as a representative case, we demonstrate the framework across diverse benchmarks, including ionic crystals, catalytic surfaces, and molecular systems. Our results achieve order-of-magnitude improvements in uncertainty–error correlation, enhances data efficiency in active learning, and transfers reliably across distinct exchange–correlation functionals. Importantly, it is general, data-efficient, and compatible with diverse MLIP architectures and baseline UQ schemes, offering a practical route toward robust and transferable atomistic simulations.","manuscriptTitle":"Flexible Uncertainty Calibration for Machine-Learned Interatomic Potentials","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-14 16:01:31","doi":"10.21203/rs.3.rs-7763957/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-01-15T23:04:23+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-09T17:14:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"146772996612005864396212285686870942938","date":"2025-12-08T08:54:14+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-02T13:40:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"338229405866316872970712250435396144199","date":"2025-11-04T05:49:15+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-04T05:35:20+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-27T01:42:01+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-20T12:30:06+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Computational Materials","date":"2025-10-02T05:40:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-computational-materials","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjcompumats","sideBox":"Learn more about [npj Computational Materials](http://www.nature.com/npjcompumats/)","snPcode":"41524","submissionUrl":"https://mts-npjcompumats.nature.com/","title":"npj Computational Materials","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3abbc763-c653-4217-b472-48bbccda2bc0","owner":[],"postedDate":"November 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":57388656,"name":"Physical sciences/Materials science"},{"id":57388657,"name":"Physical sciences/Mathematics and computing"},{"id":57388658,"name":"Physical sciences/Physics"}],"tags":[],"updatedAt":"2026-05-04T16:00:53+00:00","versionOfRecord":{"articleIdentity":"rs-7763957","link":"https://doi.org/10.1038/s41524-026-02080-3","journal":{"identity":"npj-computational-materials","isVorOnly":false,"title":"npj Computational Materials"},"publishedOn":"2026-04-27 15:58:00","publishedOnDateReadable":"April 27th, 2026"},"versionCreatedAt":"2025-11-14 16:01:31","video":"","vorDoi":"10.1038/s41524-026-02080-3","vorDoiUrl":"https://doi.org/10.1038/s41524-026-02080-3","workflowStages":[]},"version":"v1","identity":"rs-7763957","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7763957","identity":"rs-7763957","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}