I’m passionate about advancing the frontier of multimodal foundation models through data-centric approaches. From rethinking the way we curate fine-tuning datasets to enabling foundation models to create societal impact in domains such as medicine and remote sensing, I firmly believe that higher-quality data is the key to substantially improving the capability of ML models. Dive into my projects, explore my research, and please feel free to reach out to collaborate.

To pursue my research interests, I’m currently affiliated with the Stanford ML Group under Prof. Andrew Ng, and Stanford Trustworthy AI Research under Prof. Sanmi Koyejo.

A bit about my background

Originally from Lahore, Pakistan, I had the honor and privilege of attending Pomona College (Chirp Chirp!) in sunny Claremont, California for my bachelor’s degree, where I completed a double-major in Mathematics and Economics. For my Mathematics undergraduate thesis, I had the fortune of being advised and mentored by Prof. Edray Goins (more about the project in the Portfolio section).

Building upon my rigorous undergraduate training in Mathematics and a year-long experience as an Analyst at Analysis Group, I pursued my M.S. in Statistics at Stanford University (Go Card!).

News

1. I am helping organize and serving as a Reviewer for the Muslims in ML Workshop (MusIML) at NeurIPS 2024.
2. I am serving as a Reviewer for ICLR 2025.
3. I served as a Reviewer for the Socially Responsible Language Modeling Research (SoLaR) Workshop at NeurIPS 2024.
4. Our work on Many Shot In-Context Learning in Multimodal Foundation Models was accepted to the 1st In-Context Learning Workshop at ICML 2024. More on the paper in the Publications section.