Jason Yim

Hi! I am a 2nd year EECS PhD candidate at MIT CSAIL advised by coffee connoisseur Tommi Jaakkola and running enthusiast Regina Barzilay. Previously, I was as a research engineer at DeepMind and obtained my B.S. from Johns Hopkins University in Computer Science and Applied Mathematics. My research is partially funded by NSF graduate research fellowship (GRFP).

My research aims to develop machine learning methods in scientific domains such as biology and chemistry. I am passionate about applications to medicine, environmentalism, and advancing scientific knowledge.

I enjoy learning the science behind each problem and developing practical algorithms that reflect the underlying structure and mechanism. To this end, I do not limit myself to specific techniques but instead look for the right foundation in mathematics, statistics, optimization and aim to develop novel machine learning methods from them. I have research experience in the following:

• Applications: medical imaging, protein structure modeling, protein fitness optimization, de novo protein design.

• Methods: geometric deep learning, generative models (diffusion & score-based models), discrete optimization, Riemannian manifolds.

In addition, I value strong research engineering of moving fast and efficiently (i.e. minimizing bugs).

See my publications page for my recent work.