About Me

Hi, I’m Varun Sampath Kumar, a first-year PhD student at the Applied AI and Data Science Unit, The Maersk Mc-Kinney Møller Institute, University of Southern Denmark (SDU). I am fortunate to be supervised by Dr. Vinay Chakravarthi Gogineni and Dr. Esmaeil S. Nadimi. I completed my Integrated M.Sc. (B.Sc. + M.Sc.) in Data Science at PSG College of Technology (an autonomous institution under Anna University) in June 2025. During my undergraduate studies, I interned at KLA for 15 months, where I worked on self-supervised learning for wafer defect classification and contributed to data engineering pipelines involving Apache Iceberg, DuckDB, and PySpark. I also had the opportunity to collaborate with Dr. Poonthalir G on Federated Learning..

My current research focuses on developing novel machine unlearning algorithms for both vision (discriminative and generative) models and large language models (LLMs). In particular, I aim to design computationally efficient and retain data-free unlearning paradigms that can handle real-world complexities and scalability challenges.

Beyond unlearning, I am deeply interested in the memorization and generalization theory of deep neural networks, model compression, and mechanistic interpretability. My long-term goal is to contribute to building frameworks that enable AI systems to comply with emerging AI regulations while maintaining a balance between efficiency and efficacy.