I’m Rachit Singh, and I’m a senior at Harvard, doing research in variational inference, Indian Buffet processes, and language models. I work with Alexander Rush, Finale Doshi-Velez, and I’m part of the Harvard NLP research group. I frequently work with Jeffrey Ling.
I’m very interested in probabilistic programs, especially those like Edward, and I’m hoping to help build Pyro to be just as powerful, once I understand how it works. I like to write very fast code (I have a bit of CUDA + RTOS microcontroller work under my belt) in a variety of languages. Mostly these days that involves writing neural networks in PyTorch or Tensorflow, though I prefer to be Bayesian when possible.