Published a preprint on matrix sign methods and their application to the Muon algorithm titled 'The Polar Express: Optimal Matrix Sign Methods and Their Application to the Muon Algorithm'; Published a paper on attention mechanisms at ICLR titled 'Quality over Quantity in Attention Layers: When Adding More Heads Hurts'.
Research Experience
Was a research engineer at Reservoir Labs (now Qualcomm) working on modeling congestion control in communication networks; Spent summers at Adobe Research, Polymathic, and the Weizmann Institute, with the latter involving phylogenetic inference under the guidance of Boaz Nadler and Ariel Jaffe.
Education
Fourth-year PhD student in Computer Science at NYU, advised by Joan Bruna and Chris Musco; Undergraduate at Yale, advised by Dan Spielman.
Background
Research interests include linear algebra, approximation theory, and deep learning. Broadly interested in scientific computing and optimization as well.
Miscellany
Helps organize the ML-NYC speaker series at the Flatiron Institute.