Abhirup Mukherjee

pfp-ms-grad.png

Hello and welcome to my page!

I’m a Ph.D. student in Computer Science at The University of Texas at Dallas (UTD), working with Prof. Blaine Hoak as a Graduate Research Assistant on problems at the intersection of adversarial machine learning and computer security. Broadly, I’m interested in interpretable and robust AI models in security-critical applications.

Previously, I earned my master’s in Computer Science from UTD while working a part-time, on-campus student assistant job at UTD’s high performance computing center. And prior to that, I was a product development engineer at Comviva Technologies. Before venturing out into the industry, I earned my bachelor’s in Electronics and Communication Engineering (ECE) from SRM Institute of Science and Technology.

When I’m not working, I enjoy playing the guitar, listening to music, attending metal concerts, rewatching the Michael Bay live action Transformers movie trilogy, playing role-playing games (RPGs) (Soulsborne games, The Witcher franchise, etc.), and resistance training.

news

Aug 2026 Thrilled to start as a computer science Ph.D. student and Graduate Research Assistant in the Department of Computer Science, advised by Prof. Blaine Hoak!
Jun 2026 My first first-author paper “Activation Steering for Smart Contract Vulnerability Detection: A Calibration Mechanism Bounded by Pre-Training Knowledge” got accepted at the 2026 IEEE International Conference on Intelligence and Security Informatics (ISI), which was held during August 12-14, 2026 in Cambridge, United Kingdom! Huge thanks to my co-authors Prof. Latifur Khan and Prof. Bhavani Thuraisingham for their contributions, support, and guidance!
May 2026 I graduated with my master’s in computer science from UTD after defending my thesis Activation Steering for Smart Contract Vulnerability Detection! Huge thanks to my advisor Prof. Latifur Khan and committee members Prof. Sergey Bereg and Prof. Yapeng Tian!

selected publications

  1. Conference
    Activation steering methodology: leave-one-category-out prompting, CDM and CPS steering vector extraction, and steered inference
    Activation Steering for Smart Contract Vulnerability Detection: A Calibration Mechanism Bounded by Pre-Training Knowledge
    Abhirup Mukherjee, Latifur Khan, and Bhavani Thuraisingham
    ISI’26: IEEE International Conference on Intelligence and Security Informatics