Muntasir Wahed

I am a Ph.D. candidate in Computer Science at the University of Illinois Urbana-Champaign (Siebel School of Computing and Data Science), advised by Dr. Ismini Lourentzou. My research focuses on trustworthy and multi-modal machine learning, specifically aiming to advance the robustness and real-world reliability of foundation models. I specialize in addressing fundamental weaknesses in AI systems, such as failures under long-tail distributions, adversarial interactions, and missing inputs, to build models that are safe, grounded, and technically rigorous. My research has been published in premier venues, including CVPR, NeurIPS, EMNLP, WACV, and CIKM.

I have significant industry research experience, most recently as an Intern at Google DeepMind on the Gemini App team. Previously, I spent three summers at IBM Research (Almaden Lab), where my work resulted in two U.S. patents and multiple publications. I co-led the winning team for the Amazon NOVA AI Challenge, developing methodologies for LLM safety and code security that led to publications at NeurIPS 2025 and EMNLP 2025. Earlier, I was part of a finalist team in the Amazon Alexa Prize TaskBot Challenge 2, contributing to the design of a multimodal conversational agent deployed to thousands of users.

Previously, I earned my M.S. in Computer Science from Virginia Tech and my B.Sc. in Computer Science & Engineering from the University of Dhaka. My technical expertise spans large language models, vision-language models, and robust ML, backed by strong technical skills in Python, C, and full-stack development.