Aryan Sharma

I'm an undergrad at Yale studying computer science and math. I'm currently interning at Two Sigma and leading engineering for CourseTable, the course selection tool used by all of Yale's 6,500+ undergrads.

My research interests are in machine learning, with a focus on improving model reliability. Under the Supervised Program for Alignment Research, I studied hierarchical representations in transformers. At Yale, I work with Arman Cohan on mechanistic interpretability and Scott Shapiro on reinforcement learning with verifiable rewards. Most recently, our work was accepted to the ICML 2026 Compositional Learning Workshop.