I'm a Research Scientist at Google Brain where I work on artificial intelligence, large-scale language models and computer vision.
My research currently focuses on making large-scale language models cheaper to work with - via sparsity, adaptive computation and better distributed computing infrastructure.
In computer vision, I authored some of the pioneering work on Attention for Vision, proposed LambdaNetworks as a faster alternative and worked on simple vision baselines. At Google, I contributed to products such as AutoML and MuM. In the past, I proposed Neural Combinatorial Optimization with applications to AutoML and Ranking.
I advise a few select startups with their AI/ML needs and help with fashion/music projects (reach out if you're interested!)
Before Google, I spent wonderful years at Stanford as a grad student between the stats and the CS departments, after obtaining my M.S in Applied Math at Ecole Centrale Paris.
When I'm not thinking quantitavely you can probably find me doing something music related.