[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117607-en":3,"doc-seo-117607-105":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},117607,962075114765,"Quinn","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","AutoPot - Automated and massively parallelized construction of Machine-Learning Potentials","Machine-learning potentials (MLIPs) enable quantum-accurate atomistic modeling, but near-quantum performance depends on training neighborhoods matching those encountered during simulation. Building a universal training set is largely infeasible, so methods like active learning and fine-tuning add needed configurations yet create complex, hard-to-implement, interpret, and reproduce training workflows. AutoPot automates construction and archiving of MLIPs using a parallel simulation engine and an event-based workflow manager, enabling flexible integration of existing codes and on-the-fly dataset selection from MD trajectories and ab initio calculations.","arXiv :2601 .01185v1 [physics .comp-ph] 3 Jan 2026  \nAutoPot: Automated and massively parallelized construction of  \nMachine-Learning Potentials  \nM. Hodapp 1 and G. Anciaux2  \n1 Christian Doppler Laboratory for Digital material design guidelines for mitigation of alloy  \nembrittlement, Materials Center Leoben Forschung GmbH (MCL), Leoben (AT)  \n2 Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne (CH)  \nJanuary 6, 2026  \nAbstract  \nMachine-learning potentials (MLIPs) have been a breakthrough for computational physics in bringing the accuracy of quantum mechanics to atomistic modeling. To achieve near-quantum accuracy, it is necessary that neighborhoods contained in the training set are rather close to the ones encountered during a simulation. Yet, constructing a single training set that works well for all applications is, and likely will remain, infeasible, so, one strategy is to supplement training protocols for MLIPs with additional learning methods, such as active learning, or fine-tuning. This strategy, however, yields very complex training protocols that are difficult to implement efficiently, and cumbersome to interpret, analyze, and reproduce.  \nTo address the above difficulties, we propose AutoPot, a software for automating the construction and archiving of MLIPs. AutoPot is based on BlackDynamite, a software operating parametric tasks, e.g., running simulations, or single-point ab initio calculations, in a highly-parallelized fashion, and Motoko, an event-based workflow manager for orchestrating interactions between the tasks. The initial version of AutoPot supports selection of training configurations from large training candidate sets, and on-the-fly selection from molecular dynamics simulations, using Moment Tensor Potentials as implemented in MLIP-2, and single-point calculations of the selected training configurations using VASP. Another strength of AutoPot is its flexibility: BlackDynamite tasks and orchestrators are Python functions to which own existing code can be easily added and manipulated without writing complex parsers. Therefore, it will be straightforward to add other MLIP and ab initio codes, and manipulate the Motoko orchestrators to implement other training protocols.  \n1 Introduction  \nMachine-learning interatomic potentials (MLIPs) [1, 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13] have revolutionized computational materials science by overcoming the two persisting limitations of empirical interatomic potentials:  \n• MLIPs are able to predict mechanisms that influence defect motion, stability of crystal structures, etc., of a hypothetical large-scale quantum-mechanical simulation, enabling quantum-accurate multiscale simulations up to the continuum level (e.g., [14]) .  \n• MLIPs are able to predict the influence of variations of the chemistry on material properties that, e.g., enables exploring trends in the composition of complex concentrated alloys (e.g., [15, 16 , 17]) .  \nDespite these success stories, achieving predictive accuracy requires a carefully designed training set containing atomic neighborhoods that are not too far from those encountered during a simulation of the material behavior, which is a time-consuming task. Random sampling of training configuration before running a simulation is often not accurate enough, or requires a lot of training configurations [18, 19 , 20 , 21 , 22 , 23], so, protocols for constructing MLIP training sets are complemented with advanced sampling algorithms, e.g., active learning [24] from simulations that resemble the underlying problem of interest [25, 26 , 27 , 28 , 29] . Another related problem is fine-tuning a MLIP that has been pre-trained on a (generally very large) dataset using additional problem-specific configurations (e.g., [30]) .  \nSuch algorithms already have a certain complexity in that they require communication between several software codes, i.e., the code that implements the potential, the code that runs the ab initio calcu","cbCaiesJObgl16Zu","https://ap.wps.com/l/cbCaiesJObgl16Zu","pdf",6332390,1,18,"English","en",105,"# Introduction\n## MLIPs and training set challenges\n## Active learning and fine-tuning\n## AutoPot overview and workflow design\n## Configuration datasets, reproducibility, and FAIR","[{\"question\":\"Why is constructing an MLIP training set so challenging?\",\"answer\":\"Near-quantum accuracy requires training neighborhoods close to those seen during simulation. A single training set for all applications is generally infeasible, making data selection time-consuming and error-prone.\"},{\"question\":\"What problem does AutoPot address?\",\"answer\":\"AutoPot automates and archives the construction of machine-learning potentials while avoiding overly complex, difficult-to-reproduce training protocols generated by approaches like active learning and fine-tuning.\"},{\"question\":\"How does AutoPot orchestrate large-scale training workflows?\",\"answer\":\"AutoPot uses BlackDynamite to run parametric tasks in highly parallel fashion and Motoko as an event-based workflow manager to coordinate simulations and ab initio calculations, including dynamic selection from MD trajectories.\"}]","AutoPot - Automated and massively parallelized construction of Machine-Learning Potentials | PDF",1785677256,45,{"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},"autopot-automated-and-massively-parallelized-construction-of-machine-learning-potentials","",{"@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/autopot-automated-and-massively-parallelized-construction-of-machine-learning-potentials/117607/",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","2026-08-05","2026-08-02",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},"Why is constructing an MLIP training set so challenging?","Question",{"text":76,"@type":77},"Near-quantum accuracy requires training neighborhoods close to those seen during simulation. A single training set for all applications is generally infeasible, making data selection time-consuming and error-prone.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What problem does AutoPot address?",{"text":81,"@type":77},"AutoPot automates and archives the construction of machine-learning potentials while avoiding overly complex, difficult-to-reproduce training protocols generated by approaches like active learning and fine-tuning.",{"name":83,"@type":74,"acceptedAnswer":84},"How does AutoPot orchestrate large-scale training workflows?",{"text":85,"@type":77},"AutoPot uses BlackDynamite to run parametric tasks in highly parallel fashion and Motoko as an event-based workflow manager to coordinate simulations and ab initio calculations, including dynamic selection from MD trajectories.","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},[94,98,102,106,111,116,121,124,129,132,136],{"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":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]