Caterina Caccavella

Caterina Caccavella is a PhD student in Neuro AI at ETH Zurich and the Zurich University of Applied Sciences (ZHAW). After earning a bachelor's degree in Information Engineering and a master's degree in Neuroinformatics, she focused her research path on studying the mechanisms underlying artificial intelligence. She is currently conducting a research period as a visiting researcher at the Forschungszentrum Jülich and RWTH Aachen University.

His work is situated at the intersection of artificial intelligence and neuroscience, a field that aims to understand how natural and artificial systems learn, represent, and interpret the world. His interests range from computer vision and understanding the surrounding environment to the study of models inspired by principles observed in biological systems.

In particular, it explores alternative approaches to traditional deep learning architectures, with the aim of developing more efficient, adaptive systems capable of generalizing to new contexts. An important theme of the research is embodiment, the idea that intelligence emerges from the continuous interaction between perception, action, and the environment, similar to what happens in living organisms.

At the same time, it focuses on understanding how the internal representations of artificial intelligence models develop during learning and which mechanisms allow them to build an increasingly abstract and structured understanding of the world. In this regard, particular attention is given to the interpretability of the models, considered a fundamental element not only for understanding the functioning of intelligent systems but also for making them more transparent and reliable.