Julian Minder is a PhD student at the Data Science Lab (DLAB) at EPFL, supervised by Prof. Robert West and co-advised by Prof. Ryan Cotterell. His research centers on making AI systems more transparent and reliable, with a focus on interpretability and robustness in language models. He currently investigates how pretraining distributions shape model internals and how fine-tuning alters them. Julian earned his master’s degree in computer science from ETH Zurich in 2024, following earlier studies in computer science and neuroinformatics at the University of Zurich.