[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84984-en":3,"doc-seo-84984-105":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":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},84984,7971461740909,"Levi","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Physics-Audited Agentic Discovery in Scientific Machine Learning","Agentic scientific machine learning uses LLM agents to discover surrogate models and pick one via an automated score, yet low error cannot guarantee that predicted mechanics fields satisfy physics-critical properties. Physics-Audited Agentic SciML (PA-SciML) introduces a verification-first workflow that freezes the scoring evaluator before search, turns physics rules into machine-checkable requirements, and audits each candidate’s outputs. It also performs targeted range-based searches for high-violation cases and, when enabled, adds probes and isolates modeling edits to attribute score gains. Numerical solid-mechanics examples show verified surrogates pass boundary, scaling, and causality checks, unlike error-only baselines.","arXiv :2607 .07379v 1 [ cs .AI] 8 Jul 2026  \nPHYSICS-AUDITED AGENTIC DISCOVERY IN SCIENTIFIC  \nMACHINE LEARNING  \nPREPRINT  \nDiab W. Abueidda 1 ,2 * , Bilal Ahmed 1 , Panos Pantidis 1 , Mostafa E. Mobasher 1†  \n1Civil and Urban Engineering Department, New York University Abu Dhabi, United Arab Emirates  \n2National Center for Supercomputing Applications, University of Illinois at Urbana-Champaign, United States of America  \nJuly 9, 2026  \nABSTRACT  \nIn agentic scientific machine learning (SciML), large language model (LLM) agents can discover surrogate models and select one by an automated score, typically an error metric. A low error, however, does not establish that the predicted fields satisfy the physics that matter for mechanics, such as boundary conditions, superposition, stiffness scaling, or causality. We introduce Physics-Audited Agentic SciML (PA-SciML), a verification-first workflow for agentic SciML discovery. The workflow fixes a scoring evaluator before search, derives reviewable machine-checkable physics requirements, checks each trained candidate on its outputs, and separately searches prescribed input ranges or measured load-history spans for high-violation cases without reference solution fields. A surrogate is reported as verified only under the stated checks. When enabled, the workflow also adds advisory numerical probes before training and tests one modeling change at a time to record which isolated edits are associated with score gains before reuse. In the reported computational-solid-mechanics numerical examples, the static elasticity run selects a surrogate with lower validation error than the error-only baseline while both selected models pass the common linear-elastic checks. In the transient elastodynamics run, an error-only baseline with similar mean error fails a stricter causality check by responding to future parts of the loading history, while the selected surrogate passes the stated checks.  \nThe main distinction is per-candidate physics evidence on predicted fields, not a richer aggregate score.  \nKeywords agentic discovery · scientific machine learning · large language model agents · physics-based verification · neural operators · trustworthy surrogate models  \n1 Introduction  \nComputational solid mechanics often requires repeated solutions to PDE-governed boundary-value and initial-boundaryvalue problems. Design sweeps, uncertainty quantification, inverse identification, and digital-twin workflows can require many evaluations of full displacement, stress, or history-dependent field responses. High-fidelity finite-element simulations remain the reference tool, but repeated solves can be too expensive when the geometry, material law, loading, or boundary data must be varied many times. Scientific machine learning (SciML) and neural-operator surrogates offer a complementary route: learn a map from problem inputs to solution fields so that a new case can be evaluated without a fresh finite-element solve [1–4] . The practical difficulty is that constructing a useful surrogate still depends on problem-specific choices of architecture, training strategy, physics-informed structure, and evaluation metric.  \nLarge language model (LLM) agents are beginning to automate parts of this workflow. One branch writes and runs simulation code for finite-element mechanics and computational fluid dynamics [5–8] . Another branch automates engineering design and design review [9] . A third branch, closest to this paper, automates SciML model discovery:  \n∗ Corresponding author: [abueidd2@illinois.edu](abueidd2@illinois.edu)[ ](abueidd2@illinois.edu)†[Corresponding author:](Corresponding author: mostafa.mobasher@nyu.edu)[ mostafa.mobasher@nyu.edu](Corresponding author: mostafa.mobasher@nyu.edu)  \nagents can propose surrogate designs, implement training code, critique candidates, and select models with limited human intervention [10–12], building on earlier automation of physics-informed network design [13, 14] . The","cbCaiiq5mSGNr631","https://ap.wps.com/l/cbCaiiq5mSGNr631","pdf",3674976,2,1,29,"English","en",105,"# Introduction\n## Problem motivation in computational solid mechanics\n## Limits of score-based surrogate selection\n## The PA-SciML verification-first workflow\n## Numerical distinctions across static and transient cases","[{\"question\":\"What problem does PA-SciML address in agentic SciML discovery?\",\"answer\":\"It addresses that selecting surrogates by low validation error does not provide direct, machine-checkable evidence that predicted fields satisfy physics requirements such as boundary conditions, superposition, scaling, and causality.\"},{\"question\":\"How does PA-SciML perform verification compared with error-only selection?\",\"answer\":\"PA-SciML fixes the scoring evaluator before search, derives reviewable physics requirements, and checks each trained candidate on its predicted outputs against those requirements, reporting a surrogate only when it passes the stated checks.\"},{\"question\":\"What difference do the experiments show between static elasticity and transient elastodynamics?\",\"answer\":\"In static elasticity, an audit-enabled selected surrogate achieves lower validation error while both selected models pass common linear-elastic checks. In transient elastodynamics, an error-only baseline with similar mean error fails a stricter causality check, while the selected surrogate passes.\"}]",1784200027,73,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"physics-audited-agentic-discovery-in-scientific-machine-learning","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,47,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":20},"https://docshare.wps.com/document/","Document",{"item":48,"name":12,"@type":43,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/physics-audited-agentic-discovery-in-scientific-machine-learning/84984/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-20","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does PA-SciML address in agentic SciML discovery?","Question",{"text":75,"@type":76},"It addresses that selecting surrogates by low validation error does not provide direct, machine-checkable evidence that predicted fields satisfy physics requirements such as boundary conditions, superposition, scaling, and causality.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does PA-SciML perform verification compared with error-only selection?",{"text":80,"@type":76},"PA-SciML fixes the scoring evaluator before search, derives reviewable physics requirements, and checks each trained candidate on its predicted outputs against those requirements, reporting a surrogate only when it passes the stated checks.",{"name":82,"@type":73,"acceptedAnswer":83},"What difference do the experiments show between static elasticity and transient elastodynamics?",{"text":84,"@type":76},"In static elasticity, an audit-enabled selected surrogate achieves lower validation error while both selected models pass common linear-elastic checks. In transient elastodynamics, an error-only baseline with similar mean error fails a stricter causality check, while the selected surrogate passes.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"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"]