{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,14]],"date-time":"2026-01-14T20:18:47Z","timestamp":1768421927662,"version":"3.49.0"},"reference-count":123,"publisher":"IOP Publishing","issue":"1","license":[{"start":{"date-parts":[[2025,1,22]],"date-time":"2025-01-22T00:00:00Z","timestamp":1737504000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2025,1,22]],"date-time":"2025-01-22T00:00:00Z","timestamp":1737504000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/100017170","name":"Institut de Ci\u00e8ncies Fot\u00f2niques","doi-asserted-by":"crossref","award":["Internal \u201cQuantumGaudi\u201d project"],"award-info":[{"award-number":["Internal \u201cQuantumGaudi\u201d project"]}],"id":[{"id":"10.13039\/100017170","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Polish Ministry of Science and Higher Education","award":["Excellence initiative \u2013 research university"],"award-info":[{"award-number":["Excellence initiative \u2013 research university"]}]},{"name":"'la Caixa\u2019 Foundation","award":["ID 100010434"],"award-info":[{"award-number":["ID 100010434"]}]},{"DOI":"10.13039\/501100004281","name":"Narodowe Centrum Nauki","doi-asserted-by":"crossref","award":["2020\/36\/T\/ST2\/00588"],"award-info":[{"award-number":["2020\/36\/T\/ST2\/00588"]}],"id":[{"id":"10.13039\/501100004281","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000780","name":"European Union","doi-asserted-by":"crossref","award":["101113690"],"award-info":[{"award-number":["101113690"]}],"id":[{"id":"10.13039\/501100000780","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Ministry for Digital Transformation and of Civil Service of the Spanish Government","award":["QUANTUM ENIA project call - Quantum Spain project"],"award-info":[{"award-number":["QUANTUM ENIA project call - Quantum Spain project"]}]},{"name":"Fundaci\u00f3 Mir-Puig"},{"name":"Fundaci\u00f3 Cellex"},{"DOI":"10.13039\/501100011089","name":"Infrastruktura PL-Grid","doi-asserted-by":"crossref","award":["PLG\/2023\/016878"],"award-info":[{"award-number":["PLG\/2023\/016878"]}],"id":[{"id":"10.13039\/501100011089","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100018703","name":"HORIZON EUROPE European Innovation Council","doi-asserted-by":"crossref","award":["No 101080086 NeQSTGrant Agreement 101080086 \u2014 Ne"],"award-info":[{"award-number":["No 101080086 NeQSTGrant Agreement 101080086 \u2014 Ne"]}],"id":[{"id":"10.13039\/100018703","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100010663","name":"H2020 European Research Council","doi-asserted-by":"crossref","award":["AdG NOQIA"],"award-info":[{"award-number":["AdG NOQIA"]}],"id":[{"id":"10.13039\/100010663","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Barcelona Supercomputing Center MareNostrum","award":["FI-2023-3-0024"],"award-info":[{"award-number":["FI-2023-3-0024"]}]},{"DOI":"10.13039\/501100003030","name":"Ag\u00e8ncia de Gesti\u00f3 d\u2019Ajuts Universitaris i de Recerca","doi-asserted-by":"crossref","award":["AGAUR Grant No. 2021 SGR 01452"],"award-info":[{"award-number":["AGAUR Grant No. 2021 SGR 01452"]}],"id":[{"id":"10.13039\/501100003030","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100020314","name":"QuantERA","doi-asserted-by":"crossref","award":["DYNAMITE PCI2022-132919"],"award-info":[{"award-number":["DYNAMITE PCI2022-132919"]}],"id":[{"id":"10.13039\/501100020314","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100000893","name":"Simons Foundation","doi-asserted-by":"crossref","award":["Flatiron Institute"],"award-info":[{"award-number":["Flatiron Institute"]}],"id":[{"id":"10.13039\/100000893","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100011033","name":"Agencia Estatal de Investigaci\u00f3n","doi-asserted-by":"crossref","award":["CEX2019-000910-S\/10.13039\/501100011033"],"award-info":[{"award-number":["CEX2019-000910-S\/10.13039\/501100011033"]}],"id":[{"id":"10.13039\/501100011033","id-type":"DOI","asserted-by":"crossref"}]}],"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 (ML) is a promising tool for the detection of phases of matter. However, ML models are also known for their black-box construction, which hinders understanding of what they learn from the data and makes their application to novel data risky. Moreover, the central challenge of ML is to ensure its good generalization abilities, i.e. good performance on data outside the training set. Here, we show how the informed use of an interpretability method called class activation mapping, and the analysis of the latent representation of the data with the principal component analysis can increase trust in predictions of a neural network (NN) trained to classify quantum phases. In particular, we show that we can ensure better out-of-distribution (OOD) generalization in the complex classification problem by choosing such an NN that, in the simplified version of the problem, learns a known characteristic of the phase. We also discuss the characteristics of the data representation learned by a network that are predictors of its good OOD generalization. We show this on an example of the topological Su\u2013Schrieffer\u2013Heeger model with and without disorder, which turned out to be surprisingly challenging for NNs trained in a supervised way. This work is an example of how the systematic use of interpretability methods can improve the performance of NNs in scientific problems.<\/jats:p>","DOI":"10.1088\/2632-2153\/ad9079","type":"journal-article","created":{"date-parts":[[2024,11,8]],"date-time":"2024-11-08T22:58:41Z","timestamp":1731106721000},"page":"015014","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Characterizing out-of-distribution generalization of neural networks: application to the disordered Su\u2013Schrieffer\u2013Heeger model"],"prefix":"10.1088","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2600-7473","authenticated-orcid":true,"given":"Kacper","family":"Cybi\u0144ski","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0835-1644","authenticated-orcid":false,"given":"Marcin","family":"P\u0142odzie\u0144","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1792-8043","authenticated-orcid":true,"given":"Micha\u0142","family":"Tomza","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0210-7800","authenticated-orcid":true,"given":"Maciej","family":"Lewenstein","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4996-2561","authenticated-orcid":true,"given":"Alexandre","family":"Dauphin","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9498-1732","authenticated-orcid":true,"given":"Anna","family":"Dawid","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"266","published-online":{"date-parts":[[2025,1,22]]},"reference":[{"key":"mlstad9079bib1","article-title":"Introduction to latent variable energy-based models: a path towards autonomous machine intelligence","author":"Dawid","year":"2023"},{"key":"mlstad9079bib2","doi-asserted-by":"publisher","first-page":"692","DOI":"10.1038\/s41570-023-00516-8","article-title":"Ab initio quantum chemistry with neural-network wavefunctions","volume":"7","author":"Hermann","year":"2023","journal-title":"Nat. Rev. Chem."},{"key":"mlstad9079bib3","doi-asserted-by":"publisher","first-page":"631","DOI":"10.1140\/epjp\/s13360-024-05311-y","article-title":"Neural-network quantum states for many-body physics","volume":"139","author":"Medvidovi\u0107","year":"2024","journal-title":"Eur. Phys. J. Plus"},{"key":"mlstad9079bib4","doi-asserted-by":"publisher","DOI":"10.1088\/2058-9565\/ad7168","article-title":"From architectures to applications: a review of neural quantum states","volume":"9","author":"Lange","year":"2024","journal-title":"Quantum Sci. Technol."},{"key":"mlstad9079bib5","article-title":"Modern applications of machine learning in quantum sciences","author":"Dawid","year":"2023"},{"key":"mlstad9079bib6","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevA.107.010101","article-title":"Artificial intelligence and machine learning for quantum technologies","volume":"107","author":"Krenn","year":"2023","journal-title":"Phys. Rev. A"},{"key":"mlstad9079bib7","doi-asserted-by":"publisher","first-page":"202","DOI":"10.1049\/qtc2.12094","article-title":"Advances in artificial intelligence and machine learning for quantum communication applications","volume":"5","author":"Mafu","year":"2024","journal-title":"IET Quantum Commun."},{"key":"mlstad9079bib8","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevA.98.012315","article-title":"Separability-entanglement classifier via machine learning","volume":"98","author":"Lu","year":"2018","journal-title":"Phys. Rev. A"},{"key":"mlstad9079bib9","doi-asserted-by":"publisher","first-page":"936","DOI":"10.1140\/epjp\/s13360-023-04546-5","article-title":"Entanglement classification via witness operators generated by support vector machine","volume":"138","author":"Sanavio","year":"2023","journal-title":"Eur. Phys. J. Plus"},{"key":"mlstad9079bib10","doi-asserted-by":"publisher","first-page":"431","DOI":"10.1038\/nphys4035","article-title":"Machine learning phases of matter","volume":"13","author":"Carrasquilla","year":"2017","journal-title":"Nat. Phys."},{"key":"mlstad9079bib11","doi-asserted-by":"publisher","first-page":"312","DOI":"10.1016\/j.aop.2018.02.018","article-title":"Applications of neural networks to the studies of phase transitions of two-dimensional Potts models","volume":"391","author":"Li","year":"2018","journal-title":"Ann. Phys., NY"},{"key":"mlstad9079bib12","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.99.062107","article-title":"Vector field divergence of predictive model output as indication of phase transitions","volume":"99","author":"Sch\u00e4fer","year":"2019","journal-title":"Phys. Rev. E"},{"key":"mlstad9079bib13","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.102.053306","article-title":"Mapping distinct phase transitions to a neural network","volume":"102","author":"Bachtis","year":"2020","journal-title":"Phys. Rev. E"},{"key":"mlstad9079bib14","doi-asserted-by":"publisher","DOI":"10.1103\/physrevresearch.3.023016","article-title":"Revealing the phase diagram of Kitaev materials by machine learning: cooperation and competition between spin liquids","volume":"3","author":"Liu","year":"2021","journal-title":"Phys. Rev. Res."},{"key":"mlstad9079bib15","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.108.024113","article-title":"Learning by confusion approach to identification of discontinuous phase transitions","volume":"108","author":"Richter-Laskowska","year":"2023","journal-title":"Phys. Rev. E"},{"key":"mlstad9079bib16","doi-asserted-by":"publisher","first-page":"435","DOI":"10.1038\/nphys4037","article-title":"Learning phase transitions by confusion","volume":"13","author":"Van Nieuwenburg","year":"2017","journal-title":"Nat. Phys."},{"key":"mlstad9079bib17","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.96.022140","article-title":"Unsupervised learning of phase transitions: from principal component analysis to variational autoencoders","volume":"96","author":"Wetzel","year":"2017","journal-title":"Phys. Rev. E"},{"key":"mlstad9079bib18","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.120.176401","article-title":"Discriminative cooperative networks for detecting phase transitions","volume":"120","author":"Liu","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib19","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.97.013306","article-title":"Unsupervised machine learning account of magnetic transitions in the Hubbard model","volume":"97","author":"Ch\u0149g","year":"2018","journal-title":"Phys. Rev. E"},{"key":"mlstad9079bib20","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.99.104106","article-title":"Automated discovery of characteristic features of phase transitions in many-body localization","volume":"99","author":"Huembeli","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib21","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.125.170603","article-title":"Unsupervised phase discovery with deep anomaly detection","volume":"125","author":"Kottmann","year":"2020","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib22","doi-asserted-by":"publisher","first-page":"025","DOI":"10.21468\/SciPostPhys.11.2.025","article-title":"Unsupervised mapping of phase diagrams of 2D systems from infinite projected entangled-pair states via deep anomaly detection","volume":"11","author":"Kottmann","year":"2021","journal-title":"SciPost Phys."},{"key":"mlstad9079bib23","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevResearch.3.033052","article-title":"Interpretable and unsupervised phase classification","volume":"3","author":"Arnold","year":"2021","journal-title":"Phys. Rev. Res."},{"key":"mlstad9079bib24","doi-asserted-by":"publisher","DOI":"10.1088\/1367-2630\/ac9c7a","article-title":"Unsupervised learning of Rydberg atom array phase diagram with Siamese neural networks","volume":"24","author":"Patel","year":"2022","journal-title":"New J. Phys."},{"key":"mlstad9079bib25","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.104.L140202","article-title":"Detecting ergodic bubbles at the crossover to many-body localization using neural networks","volume":"104","author":"Szodra","year":"2021","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib26","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.105.224205","article-title":"Unsupervised detection of decoupled subspaces: many-body scars and beyond","volume":"105","author":"Szodra","year":"2022","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib27","doi-asserted-by":"publisher","first-page":"8823","DOI":"10.1038\/s41598-017-09098-0","article-title":"Machine learning quantum phases of matter beyond the fermion sign problem","volume":"7","author":"Broecker","year":"2017","journal-title":"Sci. Rep."},{"key":"mlstad9079bib28","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.100.224202","article-title":"Neural network setups for a precise detection of the many-body localization transition: finite-size scaling and limitations","volume":"100","author":"Th\u00e9veniaut","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib29","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.99.121104","article-title":"Machine learning of quantum phase transitions","volume":"99","author":"Dong","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib30","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevD.101.094507","article-title":"Towards novel insights in lattice field theory with explainable machine learning","volume":"101","author":"Bl\u00fccher","year":"2020","journal-title":"Phys. Rev. D"},{"key":"mlstad9079bib31","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.120.066401","article-title":"Machine learning topological invariants with neural networks","volume":"120","author":"Zhang","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib32","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.102.054512","article-title":"Deep learning of topological phase transitions from entanglement aspects","volume":"102","author":"Tsai","year":"2020","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib33","doi-asserted-by":"publisher","first-page":"087","DOI":"10.21468\/SciPostPhysCore.6.4.087","article-title":"Identifying Chern numbers of superconductors from local measurements","volume":"6","author":"Baireuther","year":"2023","journal-title":"SciPost Phys. Core"},{"key":"mlstad9079bib34","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.97.134109","article-title":"Identifying quantum phase transitions with adversarial neural networks","volume":"97","author":"Huembeli","year":"2018","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib35","doi-asserted-by":"publisher","first-page":"790","DOI":"10.1038\/s41567-019-0512-x","article-title":"Identifying topological order through unsupervised machine learning","volume":"15","author":"Rodriguez-Nieva","year":"2019","journal-title":"Nat. Phys."},{"key":"mlstad9079bib36","doi-asserted-by":"publisher","DOI":"10.1088\/1367-2630\/ab7771","article-title":"Unsupervised identification of topological phase transitions using predictive models","volume":"22","author":"Greplova","year":"2020","journal-title":"New J. Phys."},{"key":"mlstad9079bib37","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/abcc43","article-title":"Unsupervised interpretable learning of topological indices invariant under permutations of atomic bands","volume":"2","author":"Balabanov","year":"2021","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstad9079bib38","doi-asserted-by":"publisher","first-page":"1086","DOI":"10.1016\/j.fmre.2022.12.016","article-title":"Unsupervised learning of interacting topological phases from experimental observables","volume":"4","author":"Yu","year":"2023","journal-title":"Fundam. Res."},{"key":"mlstad9079bib39","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.108.205152","article-title":"Clustering neural quantum states via diffusion maps","volume":"108","author":"Teng","year":"2023","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib40","doi-asserted-by":"publisher","first-page":"917","DOI":"10.1038\/s41567-019-0554-0","article-title":"Identifying quantum phase transitions using artificial neural networks on experimental data","volume":"15","author":"Rem","year":"2019","journal-title":"Nat. Phys."},{"key":"mlstad9079bib41","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevA.102.033326","article-title":"Visualizing strange metallic correlations in the two-dimensional fermi-hubbard model with artificial intelligence","volume":"102","author":"Khatami","year":"2020","journal-title":"Phys. Rev. A"},{"key":"mlstad9079bib42","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/abffe7","article-title":"Unsupervised machine learning of topological phase transitions from experimental data","volume":"2","author":"K\u00e4ming","year":"2021","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstad9079bib43","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevResearch.5.013026","article-title":"Machine learning discovery of new phases in programmable quantum simulator snapshots","volume":"5","author":"Miles","year":"2023","journal-title":"Phys. Rev. Res."},{"key":"mlstad9079bib44","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.130.203401","article-title":"Machine learning the phase diagram of a strongly interacting Fermi gas","volume":"130","author":"Link","year":"2023","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib45","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.94.195105","article-title":"Discovering phase transitions with unsupervised learning","volume":"94","author":"Wang","year":"2016","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib46","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.102.134213","article-title":"Topological quantum phase transitions retrieved through unsupervised machine learning","volume":"102","author":"Che","year":"2020","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib47","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.124.185501","article-title":"Unsupervised manifold clustering of topological phononics","volume":"124","author":"Long","year":"2020","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib48","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.103.075106","article-title":"Visualizing quantum phases and identifying quantum phase transitions by nonlinear dimensional reduction","volume":"103","author":"Yang","year":"2021","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib49","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.121.255702","article-title":"Extrapolating quantum observables with machine learning: inferring multiple phase transitions from properties of a single phase","volume":"121","author":"Vargas-Hern\u00e1ndez","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib50","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.100.174408","article-title":"Identification of emergent constraints and hidden order in frustrated magnets using tensorial kernel methods of machine learning","volume":"100","author":"Greitemann","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib51","doi-asserted-by":"publisher","DOI":"10.1088\/1361-648X\/abbe7b","article-title":"The view of TK-SVM on the phase hierarchy in the classical kagome Heisenberg antiferromagnet","volume":"33","author":"Greitemann","year":"2021","journal-title":"J. Phys.: Condens. Matter"},{"key":"mlstad9079bib52","doi-asserted-by":"publisher","DOI":"10.1080\/23746149.2023.2202331","article-title":"Topological data analysis and machine learning","volume":"8","author":"Leykam","year":"2023","journal-title":"Adv. Phys. X"},{"key":"mlstad9079bib53","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.124.226401","article-title":"Unsupervised machine learning and band topology","volume":"124","author":"Scheurer","year":"2020","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib54","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.104.104426","article-title":"Quantitative and interpretable order parameters for phase transitions from persistent homology","volume":"104","author":"Cole","year":"2021","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib55","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.105.195115","article-title":"Unsupervised learning of topological phase diagram using topological data analysis","volume":"105","author":"Park","year":"2022","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib56","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.122.040504","article-title":"Quantum machine learning in feature Hilbert spaces","volume":"122","author":"Schuld","year":"2019","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib57","doi-asserted-by":"publisher","first-page":"4919","DOI":"10.1038\/s41467-022-32550-3","article-title":"Generalization in quantum machine learning from few training data","volume":"13","author":"Caro","year":"2022","journal-title":"Nat. Commun."},{"key":"mlstad9079bib58","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.130.220603","article-title":"Model-independent learning of quantum phases of matter with quantum convolutional neural networks","volume":"130","author":"Liu","year":"2023","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib59","doi-asserted-by":"publisher","DOI":"10.1088\/1367-2630\/abc463","article-title":"Phase detection with neural networks: interpreting the black box","volume":"22","author":"Dawid","year":"2020","journal-title":"New J. Phys."},{"key":"mlstad9079bib60","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevResearch.2.033499","article-title":"Discovering symmetry invariants and conserved quantities by interpreting siamese neural networks","volume":"2","author":"Wetzel","year":"2020","journal-title":"Phys. Rev. Res."},{"key":"mlstad9079bib61","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/ac338d","article-title":"Hessian-based toolbox for reliable and interpretable machine learning in physics","volume":"3","author":"Dawid","year":"2021","journal-title":"Mach. Learn.: Sci. Technol."},{"key":"mlstad9079bib62","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevX.12.031044","article-title":"Replacing neural networks by optimal analytical predictors for the detection of phase transitions","volume":"12","author":"Arnold","year":"2022","journal-title":"Phys. Rev. X"},{"key":"mlstad9079bib63","article-title":"Closed-form interpretation of neural network classifiers with symbolic regression gradients","author":"Wetzel","year":"2024"},{"key":"mlstad9079bib64","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.96.184410","article-title":"Machine learning of explicit order parameters: from the Ising model to SU(2) lattice gauge theory","volume":"96","author":"Wetzel","year":"2017","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib65","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.99.060404","article-title":"Probing hidden spin order with interpretable machine learning","volume":"99","author":"Greitemann","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib66","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.99.104410","article-title":"Learning multiple order parameters with interpretable machines","volume":"99","author":"Liu","year":"2019","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib67","doi-asserted-by":"publisher","first-page":"3905","DOI":"10.1038\/s41467-021-23952-w","article-title":"Correlator convolutional neural networks as an interpretable architecture for image-like quantum matter data","volume":"12","author":"Miles","year":"2021","journal-title":"Nat. Commun."},{"key":"mlstad9079bib68","article-title":"Plex: towards reliability using pretrained large model extensions","author":"Tran","year":"2022"},{"key":"mlstad9079bib69","article-title":"A survey on domain adaptation theory: learning bounds and theoretical guarantees","author":"Redko","year":"2022"},{"key":"mlstad9079bib70","doi-asserted-by":"publisher","first-page":"2921","DOI":"10.1109\/CVPR.2016.319","article-title":"Learning deep features for discriminative localization","author":"Zhou","year":"2016"},{"key":"mlstad9079bib71","doi-asserted-by":"publisher","DOI":"10.5821\/dissertation-2117-379455","article-title":"Quantum many-body physics with ultracold atoms and molecules: exact dynamics and machine learning","author":"Dawid-\u0141\u0119kowska","year":"2022"},{"key":"mlstad9079bib72","doi-asserted-by":"publisher","author":"Asb\u00f3th","year":"2016","DOI":"10.1007\/978-3-319-25607-8"},{"key":"mlstad9079bib73","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.113.046802","article-title":"Topological criticality in the chiral-symmetric AIII class at strong disorder","volume":"113","author":"Mondragon-Shem","year":"2014","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib74","doi-asserted-by":"publisher","first-page":"929","DOI":"10.1126\/science.aat3406","article-title":"Observation of the topological Anderson insulator in disordered atomic wires","volume":"362","author":"Meier","year":"2018","journal-title":"Science"},{"key":"mlstad9079bib75","doi-asserted-by":"publisher","first-page":"24","DOI":"10.1038\/s41534-020-0253-9","article-title":"Topological phases of a dimerized Fermi\u2013Hubbard model for semiconductor nano-lattices","volume":"6","author":"Le","year":"2020","journal-title":"npj Quantum Inf."},{"key":"mlstad9079bib76","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevResearch.3.013148","article-title":"Topological properties of the long-range Kitaev chain with Aubry-Andr\u00e9-Harper modulation","volume":"3","author":"Fraxanet","year":"2021","journal-title":"Phys. Rev. Res."},{"key":"mlstad9079bib77","doi-asserted-by":"publisher","first-page":"591","DOI":"10.1088\/0143-0807\/31\/3\/016","article-title":"Exact diagonalization: the Bose-Hubbard model as an example","volume":"31","author":"Zhang","year":"2010","journal-title":"Eur. J. Phys."},{"key":"mlstad9079bib78","article-title":"PyTorch: an imperative style, high-performance deep learning library","author":"Paszke","year":"2019"},{"key":"mlstad9079bib79","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1111\/j.2517-6161.1952.tb00104.x","article-title":"Rational decisions","volume":"14","author":"Good","year":"1952","journal-title":"J. R. Stat. Soc. B"},{"key":"mlstad9079bib80","doi-asserted-by":"publisher","first-page":"1492","DOI":"10.1103\/PhysRev.109.1492","article-title":"Absence of diffusion in certain random lattices","volume":"109","author":"Anderson","year":"1958","journal-title":"Phys. Rev."},{"key":"mlstad9079bib81","article-title":"Spurious correlations in machine learning: a survey","author":"Ye","year":"2024"},{"key":"mlstad9079bib82","article-title":"Examples are not enough, learn to criticize! Criticism for interpretability","volume":"vol 29","author":"Kim","year":"2016"},{"key":"mlstad9079bib83","author":"Molnar","year":"2023"},{"key":"mlstad9079bib84","doi-asserted-by":"publisher","first-page":"1316","DOI":"10.1109\/ICCV48922.2021.00137","article-title":"Towards better explanations of class activation mapping","author":"Jung","year":"2021"},{"key":"mlstad9079bib85","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.03825","article-title":"Smoothgrad: removing noise by adding noise","author":"Smilkov","year":"2017"},{"key":"mlstad9079bib86","doi-asserted-by":"publisher","first-page":"336","DOI":"10.1007\/s11263-019-01228-7","article-title":"Grad-CAM: visual explanations from deep networks via gradient-based localization","volume":"128","author":"Selvaraju","year":"2020","journal-title":"Int. J. Comput. Vis."},{"key":"mlstad9079bib87","doi-asserted-by":"publisher","first-page":"839","DOI":"10.1109\/WACV.2018.00097","article-title":"Grad-CAM++: generalized gradient-based visual explanations for deep convolutional networks","author":"Chattopadhay","year":"2018"},{"key":"mlstad9079bib88","doi-asserted-by":"publisher","first-page":"972","DOI":"10.1109\/WACV45572.2020.9093360","article-title":"Ablation-CAM: visual explanations for deep convolutional network via gradient-free localization","author":"Desai","year":"2020"},{"key":"mlstad9079bib89","article-title":"Axiom-based grad-CAM: towards accurate visualization and explanation of CNNs","author":"Fu","year":"2020"},{"key":"mlstad9079bib90","doi-asserted-by":"publisher","first-page":"111","DOI":"10.1109\/CVPRW50498.2020.00020","article-title":"Score-CAM: score-weighted visual explanations for convolutional neural networks","author":"Wang","year":"2020"},{"key":"mlstad9079bib91","article-title":"Visualizing higher-layer features of a deep network","author":"Erhan","year":"2009"},{"key":"mlstad9079bib92","article-title":"Deep inside convolutional networks: visualising image classification models and saliency maps","author":"Simonyan","year":"2014"},{"key":"mlstad9079bib93","doi-asserted-by":"publisher","first-page":"427","DOI":"10.1109\/CVPR.2015.7298640","article-title":"Deep neural networks are easily fooled: high confidence predictions for unrecognizable images","author":"Nguyen","year":"2015"},{"key":"mlstad9079bib94","doi-asserted-by":"publisher","DOI":"10.23915\/distill.00007","article-title":"Feature visualization","volume":"2","author":"Olah","year":"2017","journal-title":"Distill"},{"key":"mlstad9079bib95","doi-asserted-by":"publisher","first-page":"2131","DOI":"10.1109\/TPAMI.2018.2858759","article-title":"Interpreting deep visual representations via network dissection","volume":"41","author":"Zhou","year":"2019","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"mlstad9079bib96","doi-asserted-by":"publisher","DOI":"10.23915\/distill.00015","article-title":"Activation atlas","volume":"4","author":"Carter","year":"2019","journal-title":"Distill"},{"key":"mlstad9079bib97","doi-asserted-by":"publisher","first-page":"1926","DOI":"10.1109\/TNNLS.2022.3185901","article-title":"From clustering to cluster explanations via neural networks","volume":"35","author":"Kauffmann","year":"2024","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"mlstad9079bib98","doi-asserted-by":"publisher","first-page":"559","DOI":"10.1080\/14786440109462720","article-title":"LIII. On lines and planes of closest fit to systems of points in space","volume":"2","author":"Pearson","year":"1901","journal-title":"London, Edinburgh Dublin Phil. Mag. J. Sci."},{"key":"mlstad9079bib99","doi-asserted-by":"publisher","first-page":"073","DOI":"10.21468\/SciPostPhys.11.3.073","article-title":"A supervised learning algorithm for interacting topological insulators based on local curvature","volume":"11","author":"Molignini","year":"2021","journal-title":"SciPost Phys."},{"key":"mlstad9079bib100","doi-asserted-by":"publisher","first-page":"005","DOI":"10.21468\/SciPostPhys.14.1.005","article-title":"Unsupervised and supervised learning of interacting topological phases from single-particle correlation functions","volume":"14","author":"Tibaldi","year":"2023","journal-title":"SciPost Phys."},{"key":"mlstad9079bib101","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.06083","article-title":"Towards deep learning models resistant to adversarial attacks","author":"Madry","year":"2018"},{"key":"mlstad9079bib102","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevB.105.155128","article-title":"Deep learning for disordered topological insulators through their entanglement spectrum","volume":"105","author":"Ur\u00eda-\u00c1lvarez","year":"2022","journal-title":"Phys. Rev. B"},{"key":"mlstad9079bib103","article-title":"Robustness may be at odds with accuracy","author":"Tsipras","year":"2020"},{"key":"mlstad9079bib104","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.12518288","article-title":"Github repository: interpreting_NNs_for_topological_phases_of_matter (Version arXiv 1.0)","author":"Cybi\u0144ski","year":"2024"},{"key":"mlstad9079bib105","article-title":"The well: a large-scale collection of diverse physics simulations for machine learning","author":"Ohana","year":"2024"},{"key":"mlstad9079bib106","article-title":"The multimodal universe: enabling large-scale machine learning with 100 TB of astronomical scientific data","author":"Angeloudi","year":"2024"},{"key":"mlstad9079bib107","doi-asserted-by":"publisher","first-page":"1698","DOI":"10.1103\/PhysRevLett.42.1698","article-title":"Solitons in polyacetylene","volume":"42","author":"Su","year":"1979","journal-title":"Phys. Rev. Lett."},{"key":"mlstad9079bib108","article-title":"Deep residual learning for image recognition","author":"He","year":"2015"},{"key":"mlstad9079bib109","article-title":"Frustratingly easy domain adaptation","author":"Daum\u00e9 III","year":"2009"},{"key":"mlstad9079bib110","article-title":"Return of frustratingly easy domain adaptation","author":"Sun","year":"2015"},{"key":"mlstad9079bib111","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-49409-8_35","article-title":"Deep coral: correlation alignment for deep domain adaptation","author":"Sun","year":"2016"},{"key":"mlstad9079bib112","doi-asserted-by":"publisher","first-page":"3681","DOI":"10.1609\/aaai.v33i01.33013681","article-title":"Interpretation of neural networks is fragile","volume":"vol 33","author":"Ghorbani","year":"2019"},{"key":"mlstad9079bib113","article-title":"Thermometer encoding: one hot way to resist adversarial examples","author":"Buckman","year":"2018"},{"key":"mlstad9079bib114","doi-asserted-by":"publisher","first-page":"581","DOI":"10.1109\/BigData50022.2020.9378171","article-title":"PyHessian: neural networks through the lens of the Hessian","author":"Yao","year":"2020"},{"key":"mlstad9079bib115","doi-asserted-by":"publisher","first-page":"16057","DOI":"10.1109\/CVPR46437.2021.01580","article-title":"Adversarial laser beam: effective physical-world attack to DNNs in a blink","author":"Duan","year":"2021"},{"key":"mlstad9079bib116","article-title":"Adversarial attacks and defenses in machine learning-powered networks: a contemporary survey","author":"Wang","year":"2023"},{"key":"mlstad9079bib117","article-title":"Explanations can be manipulated and geometry is to blame","volume":"vol 32","author":"Dombrowski","year":"2019"},{"key":"mlstad9079bib118","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2021.108194","article-title":"Towards robust explanations for deep neural networks","volume":"121","author":"Dombrowski","year":"2022","journal-title":"Pattern Recognit."},{"key":"mlstad9079bib119","article-title":"UMAP: uniform manifold approximation and projection for dimension reduction","author":"McInnes","year":"2020"},{"key":"mlstad9079bib120","doi-asserted-by":"publisher","first-page":"2881","DOI":"10.1162\/neco_a_01434","article-title":"Parametric UMAP embeddings for representation and semisupervised learning","volume":"33","author":"Sainburg","year":"2021","journal-title":"Neural Comput."},{"key":"mlstad9079bib121","doi-asserted-by":"publisher","DOI":"10.1101\/689851","article-title":"A novel metric reveals previously unrecognized distortion in dimensionality reduction of scRNA-seq data","author":"Cooley","year":"2019"},{"key":"mlstad9079bib122","doi-asserted-by":"publisher","DOI":"10.23915\/distill.00002","article-title":"How to use t-SNE effectively","volume":"1","author":"Wattenberg","year":"2016","journal-title":"Distill"},{"key":"mlstad9079bib123","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1011288","article-title":"The specious art of single-cell genomics","volume":"19","author":"Chari","year":"2023","journal-title":"PLoS Comput. Biol."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,1,22]],"date-time":"2025-01-22T07:24:51Z","timestamp":1737530691000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad9079"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1,22]]},"references-count":123,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,1,22]]},"published-print":{"date-parts":[[2025,3,31]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/ad9079","relation":{},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,1,22]]},"assertion":[{"value":"Characterizing out-of-distribution generalization of neural networks: application to the disordered Su\u2013Schrieffer\u2013Heeger model","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":"2024-06-26","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2024-11-08","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-01-22","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}