Mohammad Taha Zakir | Carnegie Mellon University

Hello, I’m Taha! ✨

I am a current Master’s student in the inaugural AI Systems Management cohort at Carnegie Mellon University on the Fulbright Scholarship.

Broadly, my research centers on AI for Social Good, Responsible AI, and Interpretability. I am currently investigating red teaming, an area focused on adversarial testing of large language models (LLMs) to identify and understand their vulnerabilities. This project is carried out under the mentorship of Professor Ramayya Krishnan and Professor Beibei Li. My work examines what makes certain adversarial prompts succeed while others fail, revealing the underlying factors that shape an attack’s effectiveness and where state-of-the-art LLMs break down.

I earned my Bachelors in Computer Science from LUMS in 2023, where I worked with Professors Agha Ali Raza, Maryam Mustafa (LUMS) and Mustafa Naseem (University of Michigan-Ann Arbor) on AI applications for social good.

Check out my Google Scholar for my publications and my CV for my experiences.

Misc 🍿

In my free time, I make and play social/multiplayer tabletop or digital games. I also love reading, with my current favourite story being The Princess Bride by William Goldman. I used to be a competitive Scrabble player. The color scheme of this site was inspired by Solarized.