Kefan Xiao(萧克凡)

About & Interests

I am a Research Engineer at Google DeepMind, focused on pushing the boundaries of large language models through both model and data scaling. I have a long-standing interest in how machines learn and reason, beginning with my research as a 20%er on project Meena at Google Brain in 2019.

Throughout my career, I have contributed to several milestone projects:

• Google DeepMind: I was a founding member of the Gemini program and have been working across the entire Gemini series. Previously, I helped build PaLM-Coder and PaLM 2 through mid-training.
• Anthropic: I focused on pre-training data and modeling research.

Outside of research, I enjoy weightlifting, backpacking, and most outdoor activities. Recently, I've been spending much of my free time swimming.

Publications (Google Scholar)