Abhijith Sharma

portfolio_photo_low.png

Namaste` 🙏 Hello 👋

I am a Machine Learning Research Associate II at MBZUAI. I am working at the intersection of probabilistic generative modelling, specifically GFlowNets and CyberSecurity.

Previously, I was an Research Associate at University of Waterloo Research with AVL, where I developed adversarial testing framework for autonomous vehicles under Dr. Sebastian Fischmesiter and Dr. Nasser Azad.

I earned my master’s in computer science at the University of British Columbia. I researched safeguarding vision models against physical adversaries (Thesis) as a MITACS Accelerate intern with Troj AI Inc under the supervision of Dr. Apurva Narayan.

During my undergrad, I was also privileged to be part of COEP-SAT-2, a satellite initiative, where I worked on the priliminary control system design. I received my bachelor’s in Instrumentation and Control Engineering with the Gold Medal at the College of Engineering, Pune, India. I also worked shortly as a Decision Analytics Associate at ZS Associates

Outside of work, I enjoy playing squash, swimming, and delving into history to understand its impact on contemporary society.

news

Mar 18, 2025 My recent work on On the Privacy Risks of Spiking Neural Networks: A Membership Inference Analysis got accepted at the ICLR 2025 ICBINB Workshop as a Spotlight Paper. The source code can be found on this GitHub Link.
Dec 20, 2024 I’ve moved to Abu Dhabi this winter to join MBZUAI (Mohamed bin Zayed University of Artificial Intelligence) as a Machine Learning Research Associate II.
Nov 05, 2024 My recent work on Autonomous Vehicle Assessment by Testing Adversarial-patches in Real-time (AVATAR) got accepted at the IEEE Transactions on Intelligent Vehicles. The details can be found on this project page.
COEP Logo UBC Logo University of Waterloo Logo MBZUAI Logo