BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:Europe/Stockholm X-LIC-LOCATION:Europe/Stockholm BEGIN:DAYLIGHT TZOFFSETFROM:+0100 TZOFFSETTO:+0200 TZNAME:CEST DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20230831T095754Z LOCATION:Dischma DTSTART;TZID=Europe/Stockholm:20230626T163000 DTEND;TZID=Europe/Stockholm:20230626T183000 UID:submissions.pasc-conference.org_PASC23_sess163@linklings.com SUMMARY:MS2A - Machine Learning Techniques for Modeling Under-Resolved Phe nomena in Massively Parallel Simulations: Algorithms/Frameworks/Applicatio ns (Part 2/2) DESCRIPTION:Minisymposium\n\nOver the last ten years there has been a prof usion of scalable packages, such as TensorFlow, Pytorch, and ONNX, that al so run on heterogeneous computing platforms. The ability of these tools to ingest massive amounts of training data and make predictions, makes them an obvious choice as tools for scientific machine learning (SciML). In our presentations we demonstrate how researchers in diverse areas of scientif ic inquiry employ these tools, creatively, to model small scale phenomena in coarse grain simulations, which are then used as predictive tools. Also , with the availability of exa-scale computing platforms, it is becoming c lear that storing several petabytes of training data, for machine (ML) mod els, is not a viable option. We present ongoing research in the area of in -situ ML, where the simulation code and ML and deep learning (DL) framewor k are run together, to generate, and use, the streaming simulation data, t o train the model, and make predictions using coarser simulations. Our pre sentations also explore the performance of in-situ machine learning framew orks, and the portability of the generated ML models to simulation codes t hat are different from the one that was used to train the model.\n\nGenera tive Modeling and Smarter Sampling for Lattice Gauge Theories\n\nIn this w ork we describe how recent advancements in generative modeling have contri buted to simulations in lattice gauge theory, and discuss some ongoing wor k in this direction. In particular, we are interested in generating indepe ndent (lattice) gauge configurations, distributed according to the de...\n \n\nSam Foreman, James Osborn, and Xiao-Yong Jin (Argonne National Laborat ory)\n---------------------\nMiniGAP: A Proxy App for ML Prediction of Mol ecular Properties\n\nOne could argue that AI/ML has disrupted chemistry an d materials science, as well as other disciplines, in part, due to the syn ergy of accessible databases, accelerated computing resources, and communi ty-supported and open codes. Three common elements found in AI/ML in chemi stry research are data co...\n\n\nAlvaro Vazquez-Mayagoitia and Murat Kece li (Argonne National Laboratory)\n---------------------\nPredictive Scale- Bridging Simulations through Active Learning\n\nThroughout computational s cience, there is a growing need to utilize the continual improvements in r aw computational horsepower to achieve greater physical fidelity through s cale-bridging over brute-force increases in the number of mesh elements. F or instance, quantitative predictions of transport i...\n\n\nTimothy Germa nn (Los Alamos National Laboratory)\n---------------------\nPhysics-Inform ed Machine Learning for Reduced Lagrangian Modeling of Turbulence: Lagrang ian LES\n\nObtaining accurate numerical solutions of turbulent flows with Direct Numerical Simulation (DNS) is intractable for most practical applic ations. Thus, building efficient, accurate, and generalizable reduced-orde r models for turbulent flows remains of great interest, however, this dema nds new and cre...\n\n\nMichael Woodward (Los Alamos National Laboratory, The University of Arizona); Yifeng Tian and Chris Fryer (Los Alamos Nation al Laboratory); Misha Stepanov (The University of Arizona); Daniel Livescu (Los Alamos National Laboratory); and Misha Chertkov (The University of A rizona)\n\nDomain: Computer Science, Machine Learning, and Applied Mathema tics 
\n\nSession Chairs: Riccardo Balin (Argonne National Laboratory) and Ramesh Balakrishnan (Argonne National Laboratory) END:VEVENT END:VCALENDAR