The University of Amsterdam offers a PhD position focusing on mechanistic interpretability of machine learning models. The research will involve developing novel techniques for analyzing information flow in deep neural networks, creating evaluation frameworks, and connecting empirical findings with theoretical models. The candidate will work with AI models like transformers and GNNs, aiming to improve AI interpretability, reliability, and safety. Applicants should have a Master's degree in AI, Computer Science, or related fields, along with a solid background in machine learning and excellent Python skills.
