Kayo Yin is a PhD student at UC Berkeley advised by Jacob Steinhardt and Dan Klein. She aims to make AI systems reliable, accessible, and aligned with human values. She is currently working on better understanding the internal representations and mechanisms of large language models, as well as developing language models for signed languages. Formerly, she received her M.S. in Language Technologies from Carnegie Mellon University, B.S. in Mathematics and Computer Science from École Polytechnique, and has spent summers working at Microsoft Research and DeepMind.