Hi! My name is In Huh, and I am a first-year Ph.D. student in Electrical and Computer Engineering at Purdue University, working under the supervision of Prof. Muhammad Ashraful Alam.

My research focuses on geometric and topological methods for AI and machine learning, particularly their applications in modeling complex physical systems (e.g., semiconductor devices and materials). I am especially interested in leveraging deep learning methods enriched with topological insights to detect bifurcations, phase transitions, and catastrophic shifts hidden in seemingly regular data. Naturally, I am also drawn to deep learning models grounded in dynamical systems theory, such as neural ordinary differential equations (ODEs), and to problems in out-of-distribution (OOD) detection, since learning bifurcations can be viewed as an OOD problem with respect to the parameters of the underlying system.

My earlier research has centered on electron transport theory and its application to the mathematical modeling of low-power semiconductor devices, including tunnel field-effect transistors (TFETs) and negative-capacitance FETs (NCFETs).

In addition to my academic work, I am a staff research engineer in the Computational Science and Engineering (CSE) Team at Samsung Electronics. I am currently pursuing a Ph.D. at Purdue as part of a company-sponsored academic training program.

News

[May 2025] Our paper "Context-Informed Neural ODEs Unexpectedly Identify Broken Symmetries: Insights from the Poincaré-Hopf Theorem" is accepted to ICML 2025.

[Feb. 2025] Our paper "ML-Driven Compact Models for RRAMs: Addressing Variability and Simulation Efficiency" is accepted to IEEE EDL.

[Aug. 2024] I have started the Ph.D. program in Electrical and Computer Engineering at Purdue University.