Kehang Han

Hi there, I am a Senior Research Engineer at Google DeepMind, with a current focus on generative modeling and discrete diffusion (see our latest work MD4 and AR2Diff).

Previously, I completed my PhD on complex systems, chemistry and ML at MIT. Then I spent one year in Google Brain as an AI resident (2020-21), working with Science team and Reliable Deep Learning team.

I’m broadly interested in deep learning, probabilistic ML, language modeling, HPC and IoT. In the past I have made contributions to projects: Gemini, RT-2, T5X, Plex and RMG.

If you’d like to connect, feel free to message me via email (the envelope icon below).