Summary
Ana Marasović is an Assistant Professor at the University of Utah with 11 years of experience at the intersection of NLP, multimodality, and explainable AI, focused on building trustworthy and controllable language technology. She earned her PhD in Computational Linguistics from Heidelberg University researching deep learning for discourse-oriented tasks with limited and noisy labels, and has since held research roles at the University of Washington and the Allen Institute for AI. Her work centers on interpretability questions—why models make particular predictions, how to change their outputs, and how inputs affect behavior—using techniques like multi-task and adversarial learning. Trained originally in mathematics, she brings a strong quantitative foundation to problems of interaction and user confidence in NLP systems. Raised in Omiš, Croatia, she combines European academic rigor with US research leadership, often bridging theory and practical evaluation in applied language technologies.
11 years of coding experience
6 years of employment as a software developer
Deep Learning and Reinforcement Learning Summer Schools
Doctor of Philosophy (Ph.D.), Computational Linguistics, Doctor of Philosophy (Ph.D.), Computational Linguistics at Heidelberg University
High School, Mathematics, High School, Mathematics at 3. gimnazija, Split
The Machine Learning Summer School 2016
Master's Degree, Applied mathematics, Master's Degree, Applied mathematics at University of Zagreb, Faculty of Science
The 2nd Summer School on Integrating Vision and Language: Deep Learning
Croatian, English