• InterACT lab, building bridging systems on top of Community Notes.• Created Python library to simulate user studies with LLMs.• Previously worked on building a YouTube recommendation system with more user control.

• Published a convolutional neural network pruning paper in NeurIPS building upon a previous ICLR submission• Updated previous structured pruning methods and fed them into an adversarial attack pipeline as a baseline against a novel dense structured kernel pruning method.• Currently contributing to a knowledge graph-based knowledge editing method and efficient training of transformers for recommender systems tasks.

• Backend developer in a software development team to support Computer Science Mentors' website• Worked with Django to create a word-of-the-day feature for section attendance and analyzed 4 semesters' worth of student and mentor information, both numerical and written responses, using pandas to inform better teaching practices and to evaluate the trajectory of the organization.

Led a project called Lunchpad• Novel transformer architecture and latent diffusion models to fancify pictures of food with a generated recipePreviously worked on Ditto• Using GANs to create pixel art in the style of Pokémon

• Teaches a combined lab and discussion section and holds weekly office hours• Updates course content, including creating new test questions and improving programming projects, labs, and discussions.

• Taught an exam prep section to prepare students for the midterm and final and held office hours weekly to answer student questions and solidify conceptual understanding.

• Contributed to a paper published in Finance Research Letters on revenue diversification during COVID-19, reviewing literature on historic effects of non-interest income on risk and profitability.