AI Engineer at UBS

John Russell
Himawan

I am an ML engineer and ML researcher.

Based in Singapore, I study how language models learn, forget, and follow instructions, with a focus on trustworthy machine learning.

John Russell Himawan in a snowy landscape

About me

I graduated from NUS with Honours (Highest Distinction), majoring in Computer Science and minoring in Statistics and Entrepreneurship. My bachelor’s thesis explored machine unlearning. At NUS, I taught algorithms, machine learning, and databases, and served as Vice President of the NUS Entrepreneurship Society and Google Developer Group @ NUS.

Outside of work, I read about personal growth and mythology, and watch and play basketball.

News

  1. My co-first-authored paper, DIAL, is now on arXiv. It lets users adjust how closely a language model follows an instruction

  2. I joined UBS as an AI Engineer through the Graduate Talent Programme (GTP)

  3. My first-authored paper, Watershed, was part of the AI4GOOD Workshop at ICML. We built a benchmark for tracing how training data influences language models

  4. My co-authored paper, WaterDrum, appeared at ICLR. We used text watermarks to test whether models had actually forgotten specific training data

  5. I joined the GLOW.AI lab at NUS as a Research Assistant working with Prof. Bryan Low Kian Hsiang

  6. I graduated from NUS with Honours (Highest Distinction), majoring in Computer Science and minoring in Statistics and Entrepreneurship

Projects

All projects