Ana Lucic

Ana Lucic is an Assistant Professor in Artificial Intelligence at the University of Amsterdam, with a joint appointment between the Institute for Logic, Language and Computation and the Informatics Institute. Her research focuses on interpretable machine learning applications for scientific discovery and societal impact.

• Formerly at Microsoft Research AI for Science and Partnership on AI
• PhD in Explainable Machine Learning from University of Amsterdam (2022)
• BSc/MSc in Mathematics from McMaster University

Research Highlights:Develops mechanistic interpretability methods for deep learning architectures. Created Aurora, a foundation model for Earth system forecasting outperforming traditional operational models in air quality prediction and tropical cyclone tracking. Pioneers Clifford-Steerable CNNs for geophysical data analysis. Actively hiring PhD students for AI transparency research.

Collaborative Networks:Contributions to ELLIS Summer School and ICML workshops. Collaborates with Microsoft Research AI for Science team on climate-related ML projects. Involved in organizing TerraBytes workshop at ICML 2025.

Recent Advancements:Key role in publishing Aurora model in Nature (2025), demonstrating superior performance in Earth system forecasting. Supervises Ege Erdogan, new PhD student focused on mechanistic interpretability. Actively contributes to open-source AI development through GitHub repositories and technical discussions.