[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119835-en":3,"doc-seo-119835-105":30,"detail-sidebar-cat-0-en-105":92},{"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":20,"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},119835,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","New Results on Machine Learning Based Distinguishers - Accepted IEEE Access full version","Machine Learning (ML) is widely applied across disciplines, including differential distinguishing attacks on symmetric key ciphers. This work investigates multiple ML-based distinguishers for ASCON, SPECK, SKINNY, and SIMECK, extending the differential distinguishing model for unkeyed and round-reduced variants. New distinguishers are demonstrated for SPECK-32, SPECK-128, ASCON, SIMECK-32, SIMECK-64, and SKINNY-128, using neural networks and support vector machines. Experiments include PyTorch and TensorFlow/Keras and evaluate different activation choices and input-difference tuples, including practical data complexity.","New Results on Machine Learning Based Distinguishers  \nAnubhab Baksi 1 , Jakub Breier2 , Vishnu Asutosh Dasu3 , Xiaolu Hou4 , Hyunji Kim5 , and Hwajeong Seo5  \n1 Nanyang Technological University, Singapore; [anubhab001@e.ntu.edu.sg](anubhab001@e.ntu.edu.sg)  \n2 Silicon Austria Labs, Austria; [jakub.breier@gmail.com](jakub.breier@gmail.com)  \n3 Pennsylvania State University, Pennsylvania, USA; [vishnu98dasu@gmail.com](vishnu98dasu@gmail.com)  \n4 Slovak University of Technology in Bratislava, Slovakia; [xiaolu.hou@stuba.sk](xiaolu.hou@stuba.sk)  \n5 Hansung University, Seoul, South Korea; [khj1594012@gmail.com](khj1594012@gmail.com), [hwajeong84@gmail.com](hwajeong84@gmail.com)  \nAbstract. Machine Learning (ML) is almost ubiquitously used in multiple disciplines nowadays. Recently, we have seen its usage in the realm of differential distinguishers for symmetric key ciphers. In this work, we explore the possibility of a number of ciphers with respect to various ML-based distinguishers.  \nWe show new distinguishers on the unkeyed and round reduced version of SPECK-32, SPECK-128, ASCON, SIMECK-32, SIMECK-64 and SKINNY-128 . We explore multiple avenues in the process. In summary, we use neural network as well as support vector machine in various settings (such as varying the activation function), apart from experimenting with a number of input difference tuples. Among other results, we showa distinguisher of 8-round SPECK-32 that works with practical data complexity (most of the experiments take a few hours on a personal computer) .  \nKeywords: speck · ascon · simeck · skinny · distinguisher · machine learning · differential  \n1 Introduction  \nMachine learning (ML) is becoming ubiquitous in multiple research areas in computer science. Naturally, there have been a number of attempts to use of ML in cryptography, particularly fitting it to work with the well-known differential attack model. In fact, ML tools typically have native support for classification problem, which is similar to the distinguisher model where one attempts to classify the CIPHER from RANDOM. Particularly after Gohr’s work on SPECK-32 [23], the proper application of ML seems to be growing quite fast, as many research works including (but not limited to) [3, 11 , 15 , 16 , 18 , 19 , 22 , 26 , 28 , 30] have taken interest in this.  \nIn this work, we humbly attempt to extend the ML-assisted differential attack model. As the starting point, we adopt the differential distinguishing model presented in [7, Section 3.1] . We apply the concept of [7] to new ciphers, namely ASCON [21], SPECK [12], SKINNY [13], and SIMECK [31] . While SPECK-32 has been the major, if not the only, focus of the previous works (a trend initiated/popularised by [23]); rest ciphers have never been analyzed with respect to ML-assisted attacks, to the best of our knowledge.  \nWe carry out experiment with the two major Neural Network (NN) library, PyTorch and TensorFlow/Keras. Further, we explore the applicability of the Support Vector Machine (SVM), thus supplementing the NN which is the only ML tool used in the existing literature up to this point. More details on ML are deferred till Section 2.2.  \n1.1 Contributions  \nWe argue the traditional analysis of the differential distinguisher (that does not involve ML tools), in all likelihood, has been underestimating the attacker’s true power, who is free to use ML tools. Unlike some of the recent works (most notably, by Gohr [23]), where it is assumed the attacker is an expert in machine learning (thus is capable of designing a special purposed ML architecture), here we assume the other way around. We show, how the attacker is able to achieve the task of distinguishing cipher by using very simple ML tools – the parameters of which are decided arbitrarily. Even with that, we easily beat the non-ML based analysis, and yield same (if not better) result compared to a specialized ML architecture.  \nOur results, which are detailed in Section 5, can be summari","cbCaijIIAubpDOCY","https://ap.wps.com/l/cbCaijIIAubpDOCY","pdf",580610,1,20,"English","en",105,"# Introduction\n## Contributions\n## Novelty and Advancement of State-of-the-art","[{\"question\":\"What does the paper focus on regarding machine learning in cryptography?\",\"answer\":\"It studies how ML-based distinguishers can be used within the differential distinguishers framework for symmetric key ciphers.\"},{\"question\":\"Which ciphers and security settings are analyzed?\",\"answer\":\"The paper reports new distinguishers for unkeyed and round-reduced versions of SPECK-32, SPECK-128, ASCON, SIMECK-32, SIMECK-64, and SKINNY-128.\"},{\"question\":\"How are distinguishers implemented and what ML methods are compared?\",\"answer\":\"The work uses neural networks (including different activation functions and input-difference tuples) and also explores support vector machines with different kernels, using PyTorch and TensorFlow/Keras.\"}]","New Results on Machine Learning Based Distinguishers - Accepted IEEE Access full version | PDF",1785726556,50,{"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":87,"head_meta":89,"extra_data":91,"updated_unix":28},"new-results-on-machine-learning-based-distinguishers-accepted-ieee-access-full-version","",{"@graph":36,"@context":86},[37,54,69],{"@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/new-results-on-machine-learning-based-distinguishers-accepted-ieee-access-full-version/119835/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04","2026-08-03",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What does the paper focus on regarding machine learning in cryptography?","Question",{"text":76,"@type":77},"It studies how ML-based distinguishers can be used within the differential distinguishers framework for symmetric key ciphers.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Which ciphers and security settings are analyzed?",{"text":81,"@type":77},"The paper reports new distinguishers for unkeyed and round-reduced versions of SPECK-32, SPECK-128, ASCON, SIMECK-32, SIMECK-64, and SKINNY-128.",{"name":83,"@type":74,"acceptedAnswer":84},"How are distinguishers implemented and what ML methods are compared?",{"text":85,"@type":77},"The work uses neural networks (including different activation functions and input-difference tuples) and also explores support vector machines with different kernels, using PyTorch and TensorFlow/Keras.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,115,120,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":29,"slug":114},6,"Technology","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":21,"slug":126},9,"Religion & Spirituality","religion-spirituality",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":21,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":107,"slug":137},19,"General","general"]