{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T20:31:59Z","timestamp":1772137919123,"version":"3.50.1"},"reference-count":54,"publisher":"IOP Publishing","issue":"1","license":[{"start":{"date-parts":[[2025,3,28]],"date-time":"2025-03-28T00:00:00Z","timestamp":1743120000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2025,3,28]],"date-time":"2025-03-28T00:00:00Z","timestamp":1743120000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2025,3,31]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Machine learning has revolutionized the development of interatomic potentials over the past decade, offering unparalleled computational speed without compromising accuracy. However, the performance of these models is highly dependent on the quality and amount of training data. Consequently, the current scarcity of high-fidelity datasets (i.e. beyond semilocal density functional theory) represents a significant challenge for further improvement. To address this, this study investigates the performance of transfer learning (TL) across multiple fidelities for both molecules and materials. Crucially, we disentangle the effects of multiple fidelities and different configuration\/chemical spaces for pre-training and fine-tuning, in order to gain a deeper understanding of TL for chemical applications. This reveals that negative transfer, driven by noise from low-fidelity methods such as a density functional tight binding baseline, can significantly impact fine-tuned models. Despite this, the multi-fidelity approach demonstrates superior performance compared to single-fidelity learning. Interestingly, it even outperforms TL based on foundation models in some cases, by leveraging an optimal overlap of pre-training and fine-tuning chemical spaces.<\/jats:p>","DOI":"10.1088\/2632-2153\/adc222","type":"journal-article","created":{"date-parts":[[2025,3,18]],"date-time":"2025-03-18T18:56:20Z","timestamp":1742324180000},"page":"015071","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Multi-fidelity transfer learning for quantum chemical data using a robust density functional tight binding baseline"],"prefix":"10.1088","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6910-2142","authenticated-orcid":true,"given":"Mengnan","family":"Cui","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8473-8659","authenticated-orcid":true,"given":"Karsten","family":"Reuter","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0862-5289","authenticated-orcid":true,"given":"Johannes T","family":"Margraf","sequence":"additional","affiliation":[]}],"member":"266","published-online":{"date-parts":[[2025,3,28]]},"reference":[{"key":"mlstadc222bib1","doi-asserted-by":"publisher","first-page":"112","DOI":"10.1038\/s41929-022-00896-y","article-title":"Exploring catalytic reaction networks with machine learning","volume":"6","author":"Margraf","year":"2023","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib2","doi-asserted-by":"publisher","DOI":"10.1002\/anie.202219170","article-title":"Science-driven atomistic machine learning","volume":"62","author":"Margraf","year":"2023","journal-title":"Angew. Chem. Int. Ed."},{"key":"mlstadc222bib3","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1103\/PhysRevLett.98.146401","article-title":"Generalized neural-network representation of high-dimensional potential-energy surfaces","volume":"98","author":"Behler","year":"2007","journal-title":"Phys. Rev. Lett."},{"key":"mlstadc222bib4","doi-asserted-by":"publisher","first-page":"6365","DOI":"10.1039\/b905748j","article-title":"Optimal construction of a fast and accurate polarisable water potential based on multipole moments trained by machine learning","volume":"11","author":"Handley","year":"2009","journal-title":"Phys. Chem. Chem. Phys."},{"key":"mlstadc222bib5","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.104.136403","article-title":"Gaussian approximation potentials: the accuracy of quantum mechanics, without the electrons","volume":"104","author":"Bart\u00f3k","year":"2010","journal-title":"Phys. Rev. Lett."},{"key":"mlstadc222bib6","article-title":"SchNet: a continuous-filter convolutional neural network for modeling quantum interactions","volume":"vol 30","author":"Sch\u00fctt","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"mlstadc222bib7","doi-asserted-by":"publisher","first-page":"3678","DOI":"10.1021\/acs.jctc.9b00181","article-title":"PhysNet: a neural network for predicting energies, forces, dipole moments and partial charges","volume":"15","author":"Unke","year":"2019","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib8","article-title":"Directional message passing for molecular graphs","author":"Gasteiger","year":"2020"},{"key":"mlstadc222bib9","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms13890","article-title":"Quantum-chemical insights from deep tensor neural networks","volume":"8","author":"Sch\u00fctt","year":"2017","journal-title":"Nat. Commun."},{"key":"mlstadc222bib10","first-page":"6790","article-title":"GemNet: universal directional graph neural networks for molecules","volume":"vol 34","author":"Gasteiger","year":"2021","journal-title":"Advances Neural Information Processing Systems"},{"key":"mlstadc222bib11","doi-asserted-by":"publisher","first-page":"2453","DOI":"10.1038\/s41467-022-29939-5","article-title":"E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials","volume":"13","author":"Batzner","year":"2022","journal-title":"Nat. Commun."},{"key":"mlstadc222bib12","doi-asserted-by":"publisher","first-page":"333","DOI":"10.1039\/D2DD00008C","article-title":"NewtonNet: a Newtonian message passing network for deep learning of interatomic potentials and forces","volume":"1","author":"Haghighatlari","year":"2022","journal-title":"Digit. Discov."},{"key":"mlstadc222bib13","doi-asserted-by":"publisher","first-page":"7273","DOI":"10.1038\/s41467-021-27504-0","article-title":"SpookyNet: learning force fields with electronic degrees of freedom and nonlocal effects","volume":"12","author":"Unke","year":"2021","journal-title":"Nat. Commun."},{"key":"mlstadc222bib14","first-page":"11\u2009423","article-title":"MACE: higher order equivariant message passing neural networks for fast and accurate force fields","volume":"vol 35","author":"Batatia","year":"2022"},{"key":"mlstadc222bib15","article-title":"A foundation model for atomistic materials chemistry","author":"Batatia","year":"2024"},{"key":"mlstadc222bib16","doi-asserted-by":"crossref","first-page":"164","DOI":"10.1021\/acs.jctc.3c00710","article-title":"Transferable machine learning interatomic potential for bond dissociation energy prediction of drug-like molecules","volume":"20","author":"Gel\u017einyt\u0117","year":"2024","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib17","doi-asserted-by":"publisher","first-page":"10986","DOI":"10.1021\/acs.jctc.4c01157","article-title":"Efficient composite infrared spectroscopy: combining the double-harmonic approximation with machine learning potentials","volume":"20","author":"Pracht","year":"2024","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib18","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-024-01278-7","article-title":"Pretraining of attention-based deep learning potential model for molecular simulation","volume":"10","author":"Zhang","year":"2024","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib19","article-title":"Learning local equivariant representations for quantum operators","author":"Zhouyin","year":"2025"},{"key":"mlstadc222bib20","article-title":"How robust are modern graph neural network potentials in long and hot molecular dynamics simulations?","volume":"3","author":"Stocker","year":"2022","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstadc222bib21","doi-asserted-by":"publisher","first-page":"1345","DOI":"10.1109\/TKDE.2009.191","article-title":"A survey on transfer learning","volume":"22","author":"Pan","year":"2010","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"mlstadc222bib22","doi-asserted-by":"publisher","first-page":"28","DOI":"10.1039\/D2DD00102K","article-title":"Neural network potentials for chemistry: concepts, applications and prospects","volume":"2","author":"K\u00e4ser","year":"2023","journal-title":"Digit. Discov."},{"key":"mlstadc222bib23","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1109\/JPROC.2020.3004555","article-title":"A comprehensive survey on transfer learning","volume":"109","author":"Zhuang","year":"2021","journal-title":"Proc. IEEE"},{"key":"mlstadc222bib24","doi-asserted-by":"publisher","first-page":"2020","DOI":"10.1063\/5.0091698","article-title":"Erratum: an accurate and transferable machine learning potential for carbon","volume":"156","author":"Rowe","year":"2022","journal-title":"J. Chem. Phys."},{"key":"mlstadc222bib25","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-021-00639-w","article-title":"AtomSets as a hierarchical transfer learning framework for small and large materials datasets","volume":"7","author":"Chen","year":"2021","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib26","doi-asserted-by":"publisher","first-page":"305","DOI":"10.1109\/JAS.2022.106004","article-title":"A survey on negative transfer","volume":"10","author":"Zhang","year":"2023","journal-title":"IEEE\/CAA J. Autom. Sinica"},{"key":"mlstadc222bib27","article-title":"Overcoming data scarcity with transfer learning","author":"Hutchinson","year":"2017"},{"key":"mlstadc222bib28","doi-asserted-by":"publisher","first-page":"3564","DOI":"10.1021\/acs.chemmater.9b01294","article-title":"Graph networks as a universal machine learning framework for molecules and crystals","volume":"31","author":"Chen","year":"2019","journal-title":"Chem. Mater."},{"key":"mlstadc222bib29","doi-asserted-by":"publisher","first-page":"13406","DOI":"10.1021\/acsnano.0c05267","article-title":"Machine learning-enabled design of point defects in 2D materials for quantum and neuromorphic information processing","volume":"14","author":"Frey","year":"2020","journal-title":"ACS Nano"},{"key":"mlstadc222bib30","doi-asserted-by":"publisher","first-page":"4192","DOI":"10.1021\/acs.jctc.0c00121","article-title":"Extending the applicability of the ANI deep learning molecular potential to sulfur and halogens","volume":"16","author":"Devereux","year":"2020","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib31","first-page":"187","article-title":"Skill demonstration transfer for learning from demonstration","author":"Fitzgerald","year":"2015"},{"key":"mlstadc222bib32","article-title":"MACE-OFF23: transferable machine learning force fields for organic molecules","author":"Kov\u00e1cs","year":"2023"},{"key":"mlstadc222bib33","doi-asserted-by":"publisher","first-page":"24\u2009906","DOI":"10.1021\/acsami.9b02174","article-title":"Multifidelity information fusion with machine learning: a case study of dopant formation energies in hafnia","volume":"11","author":"Batra","year":"2019","journal-title":"ACS Appl. Mater. Interfaces"},{"key":"mlstadc222bib34","doi-asserted-by":"publisher","DOI":"10.1063\/5.0158919","article-title":"Comparison of multifidelity machine learning models for potential energy surfaces","volume":"159","author":"Goodlett","year":"2023","journal-title":"Int. J. Chem. Phys."},{"key":"mlstadc222bib35","doi-asserted-by":"publisher","first-page":"2087","DOI":"10.1021\/acs.jctc.5b00099","article-title":"Big data meets quantum chemistry approximations: the delta-machine learning approach","volume":"11","author":"Ramakrishnan","year":"2015","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib36","doi-asserted-by":"publisher","first-page":"4586","DOI":"10.1021\/acs.jctc.2c00343","article-title":"A hybrid machine learning approach for structure stability prediction in molecular co-crystal screenings","volume":"18","author":"Wengert","year":"2022","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib37","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-022-00947-9","article-title":"A multi-fidelity machine learning approach to high throughput materials screening","volume":"8","author":"Fare","year":"2022","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib38","doi-asserted-by":"publisher","first-page":"1517","DOI":"10.1038\/s41467-024-45566-8","article-title":"Transfer learning with graph neural networks for improved molecular property prediction in the multi-fidelity setting","volume":"15","author":"Buterez","year":"2024","journal-title":"Nat. Commun."},{"key":"mlstadc222bib39","doi-asserted-by":"publisher","first-page":"4913","DOI":"10.1039\/D3SC00841J","article-title":"Physics-inspired machine learning of localized intensive properties","volume":"14","author":"Chen","year":"2023","journal-title":"Chem. Sci."},{"key":"mlstadc222bib40","doi-asserted-by":"publisher","DOI":"10.1038\/s41524-024-01339-x","article-title":"Learning together: towards foundation models for machine learning interatomic potentials with meta-learning","volume":"10","author":"Allen","year":"2024","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib41","doi-asserted-by":"publisher","DOI":"10.1063\/1.4812323","article-title":"Commentary: the materials project: a materials genome approach to accelerating materials innovation","volume":"1","author":"Jain","year":"2013","journal-title":"APL Mater."},{"key":"mlstadc222bib42","doi-asserted-by":"crossref","first-page":"11","DOI":"10.1038\/s41597-022-01882-6","article-title":"SPICE, a dataset of drug-like molecules and peptides for training machine learning potentials","volume":"10","author":"Eastman","year":"2023","journal-title":"Sci. Data"},{"key":"mlstadc222bib43","doi-asserted-by":"publisher","first-page":"1546","DOI":"10.1021\/acs.jctc.8b00832","article-title":"Boosting quantum machine learning models with a multilevel combination technique: Pople diagrams revisited","volume":"15","author":"Zaspel","year":"2019","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib44","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1038\/s41597-021-00812-2","article-title":"QM7-x, a comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules","volume":"8","author":"Hoja","year":"2021","journal-title":"Sci. Data"},{"key":"mlstadc222bib45","doi-asserted-by":"publisher","first-page":"6158","DOI":"10.1063\/1.478522","article-title":"Toward reliable density functional methods without adjustable parameters: the PBE0 model","volume":"110","author":"Adamo","year":"1999","journal-title":"Int. J. Chem. Phys."},{"key":"mlstadc222bib46","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.108.236402","article-title":"Accurate and efficient method for many-body van der Waals interactions","volume":"108","author":"Tkatchenko","year":"2012","journal-title":"Phys. Rev. Lett."},{"key":"mlstadc222bib47","article-title":"Overfitting in neural nets: backpropagation, conjugate gradient and early stopping","volume":"vol 13","author":"Caruana","year":"2000","journal-title":"Advances in Neural Information Processing Systems"},{"key":"mlstadc222bib48","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41524-024-01264-z","article-title":"Complexity of many-body interactions in transition metals via machine-learned force fields from the TM23 data set","volume":"10","author":"Owen","year":"2024","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib49","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.90.104108","article-title":"Accuracy and transferability of Gaussian approximation potential models for tungsten","volume":"90","author":"Szlachta","year":"2014","journal-title":"Phys. Rev. B"},{"key":"mlstadc222bib50","doi-asserted-by":"publisher","first-page":"47","DOI":"10.1038\/s41524-024-01500-6","article-title":"Systematic softening in universal machine learning interatomic potentials","volume":"11","author":"Deng","year":"2025","journal-title":"npj Comput. Mater."},{"key":"mlstadc222bib51","doi-asserted-by":"crossref","first-page":"1546","DOI":"10.1021\/acs.jctc.8b00832","article-title":"Boosting quantum machine learning models with a multilevel combination technique: Pople diagrams revisited","volume":"15","author":"Zaspel","year":"2019","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib52","doi-asserted-by":"publisher","first-page":"1875","DOI":"10.1126\/sciadv.adf0873","article-title":"Accurate global machine learning force fields for molecules with hundreds of atoms","volume":"9","author":"Chmiela","year":"2023","journal-title":"Sci. Adv."},{"key":"mlstadc222bib53","doi-asserted-by":"publisher","first-page":"5276","DOI":"10.1021\/acs.jctc.4c00228","article-title":"Obtaining robust density functional tight-binding parameters for solids across the periodic table","volume":"20","author":"Cui","year":"2024","journal-title":"J. Chem. Theory Comput."},{"key":"mlstadc222bib54","doi-asserted-by":"publisher","DOI":"10.1063\/5.0222649","article-title":"Small basis set density functional theory method for cost-efficient, large-scale condensed matter simulations","volume":"161","author":"Keller","year":"2024","journal-title":"Int. J. Chem. Phys."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,28]],"date-time":"2025-03-28T09:23:03Z","timestamp":1743153783000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/adc222"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,28]]},"references-count":54,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,3,28]]},"published-print":{"date-parts":[[2025,3,31]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/adc222","relation":{"has-preprint":[{"id-type":"doi","id":"10.26434\/chemrxiv-2024-9734b","asserted-by":"object"}]},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,3,28]]},"assertion":[{"value":"Multi-fidelity transfer learning for quantum chemical data using a robust density functional tight binding baseline","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2025 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2024-11-21","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-03-18","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-03-28","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}