Sharut Gupta

Sharut Gupta is a graduate student in electrical engineering and computer science, working to build more robust and intelligent AI systems. Her research addresses fundamental challenges in machine learning, including how models adapt to distribution shifts and how they can learn structured, interpretable representations with minimal supervision. As a MathWorks Fellow, Sharut will develop algorithms that endow AI models with world-modeling capabilities, enabling them to reason about unseen scenarios and adapt to dynamic environments. Her work uses MATLAB, Simulink, and toolboxes for symbolic math, statistics, and optimization to simulate controlled environments, formalize theoretical guarantees, and stress-test learning algorithms. These tools support a range of applications, from solving alignment problems in hyperbolic spaces to modeling rigid-body dynamics in simulated scenes. Sharut’s research could improve the reliability and safety of AI in real-world settings, from robotics and autonomous vehicles to scientific discovery. She plans to open-source tools that make this work accessible to the broader research and engineering communities.