I am currently a research scientist at Google DeepMind, where I do research on really large-scale machine learning models. A recurring theme of my work is to explore quality-resource tradeoffs by applying fundamental ideas from randomized algorithms (i.e. make it faster with math) and compression (i.e. make it smaller with math).
I completed my PhD at Rice University, where I was advised by Anshumali Shrivastava. Previously, I attended Florida Atlantic University High School in Boca Raton FL, where I concurrently attended high school and college. I've also worked at Google Brain, Amazon AWS and A9, Adobe Research, and several startups.
My goal for this blog is to make my work more accessible to practitioners and researchers outside my area. Most research articles are written in a formal style for an expert audience. This format is a great way to deliver detailed technical information, but I think that pictures and informal articles can convey intuition more effectively.
My definition of fun includes real analysis and combinatorics, but I'm also into cycling, dance and cooking. Feel free to contact me at benjamin [dot] ray [dot] coleman [at] gmail with questions, suggestions or feedback.