[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119990-en":3,"doc-seo-119990-105":30,"detail-sidebar-cat-0-en-105":83},{"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},119990,549758252649,"Ivy","https://ap-avatar.wpscdn.com/avatar/8000253669c5317157?_k=1778319167496531819",6,"Technology","Refining fast simulation using machine learning - CMS FastSim post-hoc corrections with regression neural network","CMS requires accurate yet efficient event simulation for Phase 2 conditions, where FastSim is about 10x faster than FullSim based on Geant4 and full reconstruction, but shows up to ~20% discrepancies in some final analysis observables. This work introduces a machine learning refinement strategy using a regression neural network trained with multiple loss functions to apply post-hoc corrections without event reweighting. The refined outputs improve agreement with FullSim and strengthen observable correlations, enabling broader, higher-accuracy use of FastSim.","Refining fast simulation using machine learning  \nSamuel Bein1 , Patrick Connor1,2 , Kevin Pedro3 , Peter Schleper1 , and Moritz Wolf1 , ∗(on behalf of the CMS Collaboration)  \n1University of Hamburg, Institut für Experimentalphysik, Germany  \n2Center for Data and Computing in Natural Sciences, Hamburg, Germany  \n3Fermi National Accelerator Laboratory, Batavia, IL, USA  \nAbstract. At the CMS experiment, a growing reliance on the fast Monte Carlo application (FastSim) will accompany the high luminosity and detector granularity expected in Phase 2 . The FastSim chain is roughly 10 times faster than the application based on the Geant4 detector simulation and full reconstruction referred to as FullSim. However, this advantage comes at the price of decreased accuracy in some of the final analysis observables. In this contribution, a machine learning-based technique to refine those observables is presented. We employ a regression neural network trained with a sophisticated combination of multiple loss functions to provide post-hoc corrections to samples produced by the FastSim chain. The results show considerably improved agreement with the FullSim output and an improvement in correlations among output observables  \nand external parameters. This technique is a promising replacement for existing correction factors, providing higher accuracy and thus contributing to the wider usage of FastSim.  \n1 Introduction  \nSimulating particle collisions, the subsequent detector response, and the reconstruction of the final state are crucial for modern high energy physics. For the purpose of simulating events in the CMS detector [1], the collaboration largely relies on a simulation chain based on Geant4 [2, 3], referred to as FullSim, which gives an accurate representation of the truth [4, 5] . However, this requires a considerable amount of computing power. Therefore, another simulation chain has been established, which uses approximations to speed up the process by roughly a factor of 10 [6–8] . This application, known as FastSim, provides output with the same format and structure as FullSim. Looking towards the future with higher LHC luminosity and increased CMS detector granularity [9], FastSim will only gain in importance as the collaboration strives to keep the computing needs within budget in Phase 2 [10, 11] .  \nThe output of the FastSim chain is generally in good agreement with the FullSim chain, but discrepancies on the order of up to 20% are observed in some analysis observables. Traditionally, differences between simulation samples (or differences between simulation and data) are treated with dedicated correction factors or weights that are derived either by physics object groups or by individual users carrying out analyses. These corrections or weights are applied to events or physics objects to correct certain biases, for example in the transverse  \n∗[e-mail: moritz](e-mail: moritz.wolf@cern.ch)[.](e-mail: moritz.wolf@cern.ch)[wolf@cern](e-mail: moritz.wolf@cern.ch)[.](e-mail: moritz.wolf@cern.ch)[ch](e-mail: moritz.wolf@cern.ch)  \n© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 ([https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/)).  \nmomentum or selection efficiency of jets, photons, or leptons. One approach that goes beyond these traditional corrections in terms of sophistication and accuracy is the application of weights that are derived using machine learning-based methods, such as the DCTR (deep neural networks using classification for tuning and reweighting) approach [12] . While the accuracy of simulated variables, as well as correlations among the variables, is improved compared to the unweighted sample, the use of weights reduces the statistical power, undermining the advantage of fast simulation applications. In contrast to reweighting, the aim of our method is to change the values o","cbCaifkbxjhoxPye","https://ap.wps.com/l/cbCaifkbxjhoxPye","pdf",1956628,1,9,"English","en",105,"# Introduction\n## FastSim vs FullSim and observed discrepancies\n## Limitations of traditional correction factors\n# Data sample\n## Supersymmetry gluino pair production and event generation\n## Jet matching and NanoAOD high-level observables","[{\"question\":\"How is the training target and comparison established between FastSim and FullSim?\",\"answer\":\"Both FastSim and FullSim simulate the same events up to NanoAOD production. After reconstruction, jets are matched between generator-level and reconstructed jets (GEN, FullSim, FastSim) using a ∆R-based criterion and clustering with anti-kT (R=0.4).\"}]","Refining fast simulation using machine learning - CMS FastSim post-hoc corrections with regression neural network | PDF",1785727517,23,{"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":78,"head_meta":80,"extra_data":82,"updated_unix":28},"refining-fast-simulation-using-machine-learning-cms-fastsim-post-hoc-corrections-with-regression-neural-network","",{"@graph":36,"@context":77},[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/technology/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/refining-fast-simulation-using-machine-learning-cms-fastsim-post-hoc-corrections-with-regression-neural-network/119990/",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","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71],{"name":72,"@type":73,"acceptedAnswer":74},"How is the training target and comparison established between FastSim and FullSim?","Question",{"text":75,"@type":76},"Both FastSim and FullSim simulate the same events up to NanoAOD production. After reconstruction, jets are matched between generator-level and reconstructed jets (GEN, FullSim, FastSim) using a ∆R-based criterion and clustering with anti-kT (R=0.4).","Answer","https://schema.org",{"og:url":52,"og:type":79,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":81,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":84},[85,89,93,97,102,105,110,115,119,122,126],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":86,"show_sort_weight":87,"slug":88},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":90,"show_sort_weight":91,"slug":92},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Exam",70,"exam",{"id":98,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":103,"slug":104},50,"technology",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},7,"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},8,"Research & Report",30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":98,"slug":129},19,"General","general"]