.jpg)
Tuomas works on the white-box methods/applied interpretability pod and is focused on creating safer AI systems through understanding their internal mechanisms. Tuomas earned his PhD in computer science from UC San Diego advised by Prof. Lily Weng where he focused on building improved automated interpretability methods and more rigorous interpretability evaluations. Before that he earned his Bachelor’s in Computer Science and Engineering and in Philosophy from MIT. Overall, he is driven to better understand AI and use this knowledge to improve AI safety.
