Lukas Galke Poech

Assistant Professor
University of Southern Denmark (SDU)
Faculty of Science
Department of Mathematics and Computer Science (IMADA)
Data Science Section
Center for Machine Learning
Affiliate member of the Pioneer Centre for Artificial Intelligence (P1)

Advancing AI safety through interpretability-informed control

My research advances AI safety by investigating how and why advanced AI systems behave the way they do, and by developing interpretability-informed methods for retaining human control. I lead the AI Safety & Interpretability Lab at SDU, where we work on interpretability and multi-agent safety.