[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117888-en":3,"doc-seo-117888-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},117888,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","Finding defects in glasses through machine learning","Structural defects govern the kinetic, thermodynamic, and mechanical properties of glasses, especially two-level systems (TLS) that control low-temperature physics yet are extremely difficult to identify directly in simulations. A machine learning approach is presented to efficiently explore glass potential energy landscapes and rapidly predict quantum splitting between pairs of amorphous configurations. The method shifts effort toward collecting TLS rather than abundant non-tunneling defects, and the model is interpreted to provide microscopic physical insight into TLS identification and characterization.","Article [https://doi.org/10.1038/s41467-023-39948-7](https://doi.org/10.1038/s41467-023-39948-7)  \nFinding defects in glasses through machine learning  \nReceived: 8 February 2023  \n\n| Accepted: 20 June 2023 |\n| --- |\n|  |\n| Check for updates |\n\nSimone Ciarella 1,7 , Dmytro Khomenko2,3,7 , Ludovic Berthier 4,5, Felix C. Mocanu1, David R. Reichman 2, Camille Scalliet 6 & Francesco Zamponi1  \nStructural defects control the kinetic, thermodynamic and mechanical properties of glasses. For instance, rare quantum tunneling two-level systems (TLS) govern the physics of glasses at very low temperature. Due to their extremely low density, it is very hard to directly identify them in computer simulations. We introduce a machine learning approach to efficiently explore the potential energy landscape of glass models and identify desired classes of defects. We focus in particular on TLS and we design an algorithm that is able to rapidly predict the quantum splitting between any two amorphous configurations produced by classical simulations. This in turn allows us to shift the computational effort towards the collection and identification of a larger number of TLS, rather than the useless characterization of non-tunneling defects which are much more abundant. Finally, we interpret our machine learning model to understand how TLS are identified and characterized, thus giving direct physical insight into their microscopic nature.  \nWhen a glass-forming liquid is cooled rapidly, its viscosity increases dramatically and it eventually transforms into an amorphous solid, called a glass, whose physical properties are profoundly different from those of ordered crystalline solids1. At even lower temperature, around 1 K, the specific heat of a disordered solid is much larger than that ofits crystalline counterpart as it scales linearly rather than cubically with temperature. Similarly, the temperature evolution of the thermal conductivity in glasses is quadratic, rather than cubic2–11. A theoretical framework rationalizing such anomalous behavior was provided by Anderson, Halperin, and Varma12 and by Phillips13,14. They argued that the energy landscape of amorphous solids contains many nearlydegenerate minima, connected by the localized motion of a few atoms, that can act as tunneling defects, called two-level systems (TLS)15–19. Since then, localized structural defects have been understood to playa crucial role in many other glass properties20. Understanding the microscopic origin of such localized defects and how to control their  \ndensity and physical properties is a major goal not only to improve our fundamental understanding of amorphous solids, but also for technological applications, such as optimizing the performance of certain quantum devices21,22.  \nThe development of particle-swap computer algorithms23,24 has allowed the creation of computer glasses at unprecedentedly low temperatures. Combined with potential energy landscape exploration algorithms25–34, this provides a powerful method to investigate the nature of defects in materials prepared under conditions comparable to experimental studies35,36. These tools have enabled direct numerical observation of TLS35,37, confirming the experimental result7,8,10,11,38 that the density of tunneling defects is strongly depleted as the kinetic stability of a glass increases. Similar results have been obtained for a different kind of defect, namely soft vibrational modes39. The direct detection of TLS revealed some of their microscopic features, namely that fewer atoms participate in the TLS of stable glasses35. It was also  \n1Laboratoire de Physique de l’École Normale Supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université de Paris, 75005 Paris, France. 2Department of Chemistry, Columbia University,3000 Broadway, New York, NY 10027, USA. 3Dipartimento di Fisica, Sapienza Universitàdi Roma, P.le A. Moro 2, I-00185 Rome, Italy. 4Yusuf Hamied Department of Chemistry, University of Cambridge, L","cbCaid1BhqK77tz5","https://ap.wps.com/l/cbCaid1BhqK77tz5","pdf",1836763,1,11,"English","en",105,"# Introduction\n## Motivation: structural defects and TLS\n## Challenges in direct identification\n# Proposed machine learning approach\n## Predicting quantum splitting between amorphous configurations\n## Efficient defect identification and reduced waste\n# Interpretation and physical insight\n## Understanding how TLS are recognized","[{\"question\":\"Why are two-level systems (TLS) important in glasses?\",\"answer\":\"TLS govern the physics of glasses at very low temperatures. They connect nearly degenerate energy minima through localized atomic motion, leading to observable anomalous properties.\"},{\"question\":\"What limitation makes direct simulation-based TLS detection difficult?\",\"answer\":\"Direct landscape exploration is computationally expensive because it requires finding energy paths between very large numbers of inherent-structure pairs to determine whether they form tunneling defects.\"},{\"question\":\"How does the machine learning method improve TLS identification?\",\"answer\":\"It rapidly predicts the quantum splitting between pairs of amorphous configurations produced by classical simulations, enabling focus on tunneling TLS rather than the far more abundant non-tunneling defects.\"}]","Finding defects in glasses through machine learning | PDF",1785680182,28,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"finding-defects-in-glasses-through-machine-learning","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/finding-defects-in-glasses-through-machine-learning/117888/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why are two-level systems (TLS) important in glasses?","Question",{"text":75,"@type":76},"TLS govern the physics of glasses at very low temperatures. They connect nearly degenerate energy minima through localized atomic motion, leading to observable anomalous properties.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What limitation makes direct simulation-based TLS detection difficult?",{"text":80,"@type":76},"Direct landscape exploration is computationally expensive because it requires finding energy paths between very large numbers of inherent-structure pairs to determine whether they form tunneling defects.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the machine learning method improve TLS identification?",{"text":84,"@type":76},"It rapidly predicts the quantum splitting between pairs of amorphous configurations produced by classical simulations, enabling focus on tunneling TLS rather than the far more abundant non-tunneling defects.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]