Two PhD Positions in AI for Science at the University of Amsterdam

Two four-year PhD projects make foundation models useful and understandable for science, within the Horizon Europe project EVELYN led at the University of Amsterdam by Prof. Evangelos Kanoulas and Dr. Ana Lucic.

Position 1 (Information Retrieval Lab) develops parameter-efficient fine-tuning and retrieval-augmented methods that adapt multimodal foundation models to scientific tasks such as satellite detection maps, fast simulation surrogates, gap-filled sensor records and ranked explanations, using EVELYN environmental use cases (air quality, PFAS, microplastics, soil and water contamination, oil-spill and plastic detection), with evaluation against strong specialist baselines.

Position 2 focuses on mechanistic interpretability: post-hoc techniques that reverse-engineer how trained deep networks (for example transformers and graph neural networks) make predictions, and frameworks for testing whether those explanations are correct.

Applications close on 15 January 2027.

MSc in artificial intelligence, computer science, engineering, mathematics, physics or a related discipline; demonstrable background in machine learning; excellent Python software engineering skills; fluent English.