Alex Serrano

Alex Serrano Terre

I am a Student Researcher on the AGI Safety and Alignment team at Google DeepMind.

Previously, I was a Visiting Researcher at the Center for Human-Compatible AI at UC Berkeley, advised by Scott Emmons and Prof. Stuart Russell, and a Research Fellow at MATS, advised by Scott Emmons, Erik Jenner and David Lindner. I've been a Research Fellow at AI Safety Camp, working on fundamental interpretability research supervised by Kola Ayonrinde and a Research Fellow at SPAR, working on white-box control protocols supervised by Aryan Bhatt. I studied Data Science and Mathematics as an undergraduate at CFIS (Polytechnic University of Catalonia).

I am interested in making AI models safer and more robust, and developing effective monitoring measures that can robustly scale to frontier language models. If you are interested in collaborating, please reach out.

Email: mail [at] alexserrano [dot] org
Google Scholar

Publications

Open Source Projects