Advisors: Prof. Richard Zemel, Prof. Carl Vondrick.

Princeton Visual AI Lab with Prof. Olga Russakovsky. Worked on vision+language models for visual question answering, introducing new tasks and modalities in this setting.

Princeton Lab for Intelligent Probabilistic Systems: Worked with Prof. Ryan Adams to bring new mathematical tools into neural network computation. https://www.pacm.princeton.edu/news/2021-undergraduate-program-prize-winners-announced

Built ML models and visualizations to make Search Ads more aware and responsive to user tasks, and to identify commercial opportunities for ads growth.

Built an end-to-end pipeline to automatically generate labels for training data, enabling better IT/HR support with larger, more accurate recommendation/chatbot models.

Developed a deep learning model to design antibodies that bind better to their targets. Named as one of 40 Finalists nationwide in the Regeneron Science Talent Search.

Developed a novel deep learning-based method to outline tumors in medical images, which performed nearly as well as human radiologists. Published a paper in the leading medical imaging journal IEEE Trans. Medical Imaging (doi: 10.1109/TMI.2016.2628084)