David J. Lee

Research Engineer at Scale AI • GitHub • Google Scholar • LinkedIn • Email • New York, NY

I'm a research engineer at Scale AI, where I develop benchmarks to measure LLM capabilities. My recent work includes SEAL Showdown, a live LLM ranking using an in situ chat setting.

Prior to Scale, I was a graduate student at Cornell, where I worked with Kevin Ellis on program synthesis, focusing on synthetic data generation. In particular, I designed algorithms to generate diverse training data for code models, using ideas from novelty search.

Before that, I worked on filters, concurrency, and knot theory as an undergrad at Williams College. For my undergraduate thesis, I was fortunate to be advised by Shikha Singh and Sam McCauley.