Anuj Karpatne

Hello everyone! I am an Associate Professor in the Department of Computer Science at Virginia Tech (VT) where I lead the Knowledge-guided Machine Learning (KGML) Lab. I also serve as a Faculty Fellow and Dean’s Fellow in the College of Engineering at VT, and on leadership roles in center-scale research projects including: co-PI & KGML lead of the $15M NSF HDR Imageomics Institute, Jump Thrust co-lead of the $18M NSF PIPP COMPASS Center, site PI & KGML co-lead of a $10M USDA SAS project, site PI and ML lead of the $1.1M NSF Eco-KGML project, and lead PI of a $1M NSF Medium project.

What is KGML? The goal of KGML is to incorpoate scientific knowledge in AI/ML models to make them scientifically grounded, interpretable, and generalizable even on out-of-distribution data. Starting with a vision article in 2017, KGML has grown into a vibrant research community spanning multiple directions, from solving partial differential equations (PDEs) to discovering governing equations, generating scientific hypotheses, and building digital twins combining physics-based and ML models. Today, KGML is leading a new paradigm in AI for Science while also advancing the Science of AI driven by the needs of problems in science and engineering. To learn more about KGML, see: KGML book, lecture recording of a tutorial on KGML, recent perspective article, and keynote talk slides from a recent KGML workshop sponsored by Schmidt Sciences.

Focus of KGML Lab: Our lab is advancing the frontiers of KGML by rethinking the assumptions that go into AI/ML models inspired by scientific knowledge. Our work has resulted in novel formulations for knowledge-guided (KG)-representation learning, KG-loss functions, KG-neural architectures, KG-interpretable AI, and more recently, KG-generative AI, KG-foundation models, and KG-agentic AI (see Publications for details). Our work is highly inter-disciplinary in nature and we have been fortunate to collaborate with an amazing network of scientists spanning environmental sciences (aquatic sciences, agriculture, remote sensing, and geophysics), biological sciences (organismal biology, biodiversity science, virology, and mechanobiology), and physical sciences (fluid dynamics, quantum mechanics, and electromagnetism), with generous support from NSF and USDA (see Funding for details on grants and collaborations).

To learn more about our research, please visit the KGML-Lab Website.

Slides from Recent Talks:

1. Navigating Research Landscape in KGML: Problems, Methods, and Emerging Opportunities, keynote talk at the KGML workshop held at the University of Michigan in August 2025 and sponsored by Schmidt Sciences AI in Science program.
2. Foundations and Frontiers for Generative AI in Science and Engineering, opening session of the GenAI4Science Workshop held at the Univesity of Minnesota in August 2025, co-presented with Aryan Deshwal and Jana Doppa.