Origin of the Machine Learning Forces Field Errors Across Metal Elements

Origin of the Machine Learning Forces Field Errors Across Metal Elements | 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 Origin of the Machine Learning Forces Field Errors Across Metal Elements Xingze Geng, Wentao Zhang, Lin-Wang Wang, Xiangying Meng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7716756/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 27 Jan, 2026 Read the published version in npj Computational Materials → Version 1 posted 13 You are reading this latest preprint version Abstract The overall development of the machine learning force field (MLFF) has advanced rapidly, with a wide range of models emerging in recent years. However, some fundamental questions remain underexplored, such as why certain systems are intrinsically more difficult to train than others. Understanding this question can help us to propose different models and prepare appropriate datasets for different situations. We constructed Metal-43, a high-precision dataset comprising elemental structures of 43 metallic elements. Through systematic analysis, we reveal regular trends of fitting accuracies of these elemental metals in the periodic table. Unlike previous approaches that generally attribute fitting challenges to a vague notion of a “complex potential energy surface (PES)”, which is almost a synonym of the fitting difficulty, we provide a physical picture which connects the Fermi surface complexity to this complexity of PES. Furthermore, we demonstrate that current MLFF models still face clear limitations in capturing the complex PES even for elemental materials. These findings can provide a theoretical foundation and directional guidance for the development of more general and accurate MLFF models in the future. Physical sciences/Engineering Physical sciences/Materials science Physical sciences/Mathematics and computing Machine learning Force field Dataset Metal Density functional theory Full Text Additional Declarations No competing interests reported. Supplementary Files SupplementaryInformation.pdf Cite Share Download PDF Status: Published Journal Publication published 27 Jan, 2026 Read the published version in npj Computational Materials → Version 1 posted Editorial decision: Revision requested 25 Nov, 2025 Reviews received at journal 20 Nov, 2025 Reviews received at journal 13 Nov, 2025 Reviews received at journal 10 Nov, 2025 Reviews received at journal 31 Oct, 2025 Reviewers agreed at journal 16 Oct, 2025 Reviewers agreed at journal 14 Oct, 2025 Reviewers agreed at journal 14 Oct, 2025 Reviewers agreed at journal 13 Oct, 2025 Reviewers invited by journal 13 Oct, 2025 Editor assigned by journal 10 Oct, 2025 Submission checks completed at journal 07 Oct, 2025 First submitted to journal 25 Sep, 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-7716756","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":530427137,"identity":"bd6bbbde-a6b6-426b-a7dc-e5c5bbcc7055","order_by":0,"name":"Xingze Geng","email":"","orcid":"","institution":"Northeastern University","correspondingAuthor":false,"prefix":"","firstName":"Xingze","middleName":"","lastName":"Geng","suffix":""},{"id":530427138,"identity":"323f985b-2fc2-47d2-9080-426a952e2b71","order_by":1,"name":"Wentao Zhang","email":"","orcid":"","institution":"Institute of Semiconductors","correspondingAuthor":false,"prefix":"","firstName":"Wentao","middleName":"","lastName":"Zhang","suffix":""},{"id":530427139,"identity":"636769ca-3c84-4f3c-876d-cc0aa56321f9","order_by":2,"name":"Lin-Wang Wang","email":"","orcid":"","institution":"Institute of Semiconductors","correspondingAuthor":false,"prefix":"","firstName":"Lin-Wang","middleName":"","lastName":"Wang","suffix":""},{"id":530427140,"identity":"064add0c-df06-4275-9008-c0942c64ec17","order_by":3,"name":"Xiangying Meng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+klEQVRIiWNgGAWjYDADfiA+8ADKkSBKi2QDUEsCSVoMDgAJorQYHOAxYLpRccdu87XDD4G23JEzOMB88DYPg10ebvfwGDDnnHmWvO12mgFQyzNjgwNsydY8DMnFuLTwMwC15LYdTja7nQDScjhxwwEeM2kehgOJDTi0sIG1/DucbDw7/QNIS/2GA/zf8GqB2NJw2M5AOgdsC9AuHja8WiQb2AqYc44dTpC4nVNwIMHgsOHMw2zGlnMMknFqAYbPBuacmsP2/LPTN3/4UHFYnu9488MbbyrscGphkH9h/gNIQRUYADEzjIETsD8Akfb4lIyCUTAKRsEIBwAEiFcAWbyMZgAAAABJRU5ErkJggg==","orcid":"","institution":"Liaoning Academy of Materials","correspondingAuthor":true,"prefix":"","firstName":"Xiangying","middleName":"","lastName":"Meng","suffix":""}],"badges":[],"createdAt":"2025-09-26 02:08:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7716756/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7716756/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41524-026-01977-3","type":"published","date":"2026-01-27T15:58:13+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":94451402,"identity":"4052c567-e862-4946-833b-94bf599da190","added_by":"auto","created_at":"2025-10-27 14:40:06","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40483102,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/c63ed07d3106e55e838b8dad.docx"},{"id":94451359,"identity":"1757e9bb-4f75-4040-b4f5-07e30912969b","added_by":"auto","created_at":"2025-10-27 14:40:05","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5401,"visible":true,"origin":"","legend":"","description":"","filename":"cb75114905d64cb7b9c8a5c3746babf0.json","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/f772aa269d187de33e94f360.json"},{"id":94451242,"identity":"4797535f-61bb-4bfc-8e54-30dd84d71726","added_by":"auto","created_at":"2025-10-27 14:39:59","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3901734,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/5caf063b17c70c02e412d815.pdf"},{"id":94451515,"identity":"2a65d71b-f7a9-4565-a4a7-82a145ac5fff","added_by":"auto","created_at":"2025-10-27 14:40:12","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":108068,"visible":true,"origin":"","legend":"","description":"","filename":"cb75114905d64cb7b9c8a5c3746babf01enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/45b6bf7bcca4969677f6b09a.xml"},{"id":94451033,"identity":"2fe7b0b6-bf26-46d3-8e35-48b9fe14d5f4","added_by":"auto","created_at":"2025-10-27 14:39:45","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5091180,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/43cdc89ab6eaee8dbd785931.png"},{"id":94451736,"identity":"c4947eeb-2d61-48ae-b085-3cd5f2f739cc","added_by":"auto","created_at":"2025-10-27 14:40:20","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3365459,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/3d15f8b477b81b18a69f5187.png"},{"id":94450847,"identity":"9d56f837-a52a-447f-946d-c9ee2a57ad2c","added_by":"auto","created_at":"2025-10-27 14:39:37","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14654045,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/27c2224a1f40312c8a377fc3.png"},{"id":94451821,"identity":"ad5fe0f2-d7fb-430f-9106-8796d007b2a8","added_by":"auto","created_at":"2025-10-27 14:40:25","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5437316,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/eb60c3b3b50ed9b11e5692a4.png"},{"id":94451620,"identity":"4f48b560-e9c7-4057-b327-bbd694de8485","added_by":"auto","created_at":"2025-10-27 14:40:14","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5620242,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/28907956beb07ec3ff672d12.png"},{"id":94451072,"identity":"28d19ab3-4842-4c83-ba90-1431032f577c","added_by":"auto","created_at":"2025-10-27 14:39:48","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6219458,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/d78b0107844a5449c91f46ba.png"},{"id":94451241,"identity":"4eda8797-a44b-4482-a791-47817d783fab","added_by":"auto","created_at":"2025-10-27 14:39:59","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":949419,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/9086b86aed8d3b09b4f12597.png"},{"id":94451351,"identity":"c91fc699-4490-4115-8218-8945adcc0e4a","added_by":"auto","created_at":"2025-10-27 14:40:04","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":782196,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/ea86336d84e27eb2e022d280.png"},{"id":94450919,"identity":"82c43a2a-9139-4598-9e78-902a9155a889","added_by":"auto","created_at":"2025-10-27 14:39:41","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2039911,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/06a00346d168e00f3c347fcb.png"},{"id":94451659,"identity":"4b1fb647-690d-4572-baca-a9dd0ab1abf1","added_by":"auto","created_at":"2025-10-27 14:40:16","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":951536,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/4ead321d23d2d5d717a3848e.png"},{"id":94451724,"identity":"7e328a69-0766-4142-9e61-e0a6a358d55c","added_by":"auto","created_at":"2025-10-27 14:40:19","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1058124,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/9822c691d2886db1b9f946cd.png"},{"id":94451428,"identity":"ba7304b9-842b-400c-9831-6305c99a92f2","added_by":"auto","created_at":"2025-10-27 14:40:07","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1049071,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/6f3423f8b8ce68b665e79905.png"},{"id":94451499,"identity":"a402f5c4-6676-4af1-9780-1fa0e4c6ef55","added_by":"auto","created_at":"2025-10-27 14:40:11","extension":"xml","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":107337,"visible":true,"origin":"","legend":"","description":"","filename":"cb75114905d64cb7b9c8a5c3746babf01structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/c691e9d4d8b5e76c018b429d.xml"},{"id":94451297,"identity":"972ee0df-73dd-412a-a8e7-0ab4bfe3a8fb","added_by":"auto","created_at":"2025-10-27 14:40:01","extension":"html","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":120392,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/c0a6c0b019014e4ff7104a06.html"},{"id":101690497,"identity":"b9cac60f-ea01-4a32-b699-d45a101a5e54","added_by":"auto","created_at":"2026-02-02 16:04:15","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1031586,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1_covered_487ef4d9-5038-42a8-aef0-67d262962f5b.pdf"},{"id":94451705,"identity":"dafcc308-83f9-4e62-96cc-064c690618d2","added_by":"auto","created_at":"2025-10-27 14:40:18","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":3901734,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7716756/v1/335b4755d5890bc8a1ef6a44.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Origin of the Machine Learning Forces Field Errors Across Metal Elements","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":"Machine learning, Force field, Dataset, Metal, Density functional theory","lastPublishedDoi":"10.21203/rs.3.rs-7716756/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7716756/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe overall development of the machine learning force field (MLFF) has advanced rapidly, with a wide range of models emerging in recent years. However, some fundamental questions remain underexplored, such as why certain systems are intrinsically more difficult to train than others. Understanding this question can help us to propose different models and prepare appropriate datasets for different situations. We constructed Metal-43, a high-precision dataset comprising elemental structures of 43 metallic elements. Through systematic analysis, we reveal regular trends of fitting accuracies of these elemental metals in the periodic table. Unlike previous approaches that generally attribute fitting challenges to a vague notion of a \u0026ldquo;complex potential energy surface (PES)\u0026rdquo;, which is almost a synonym of the fitting difficulty, we provide a physical picture which connects the Fermi surface complexity to this complexity of PES. Furthermore, we demonstrate that current MLFF models still face clear limitations in capturing the complex PES even for elemental materials. These findings can provide a theoretical foundation and directional guidance for the development of more general and accurate MLFF models in the future.\u003c/p\u003e","manuscriptTitle":"Origin of the Machine Learning Forces Field Errors Across Metal Elements","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-27 11:32:27","doi":"10.21203/rs.3.rs-7716756/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-11-25T12:45:48+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-20T07:57:30+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-14T03:14:13+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-10T15:32:53+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-31T20:03:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"259713382065813677354622502364993692190","date":"2025-10-16T06:19:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"248197904335169381825710327270225377596","date":"2025-10-14T18:38:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105916249793731607456906269321296193302","date":"2025-10-14T11:21:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"271667627180084650467307814234658142020","date":"2025-10-13T16:31:28+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-13T16:26:06+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-10T08:27:58+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-07T12:19:04+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Computational Materials","date":"2025-09-26T01:52:25+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":"de534748-7c3c-4b4d-8915-b21f9e5ca504","owner":[],"postedDate":"October 27th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":56386463,"name":"Physical sciences/Engineering"},{"id":56386464,"name":"Physical sciences/Materials science"},{"id":56386465,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-02-02T16:01:19+00:00","versionOfRecord":{"articleIdentity":"rs-7716756","link":"https://doi.org/10.1038/s41524-026-01977-3","journal":{"identity":"npj-computational-materials","isVorOnly":false,"title":"npj Computational Materials"},"publishedOn":"2026-01-27 15:58:13","publishedOnDateReadable":"January 27th, 2026"},"versionCreatedAt":"2025-10-27 11:32:27","video":"","vorDoi":"10.1038/s41524-026-01977-3","vorDoiUrl":"https://doi.org/10.1038/s41524-026-01977-3","workflowStages":[]},"version":"v1","identity":"rs-7716756","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7716756","identity":"rs-7716756","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}