Petar Veličković

Research Scientist


Throughout my academic career, I’ve been fortunate to contribute to a wide variety of research directions, including graph representation learning, computational biology, medical imaging, cross-modal learning, reinforcement learning, explainable AI, and deep learning on low-power/embedded systems. A detailed list of relevant publications is provided below.

Currently, my main research interests are on graph representation learning for algorithmic reasoning. This new and exciting direction seeks to understand and employ the expressive power of GNNs for modelling classical algorithms. The implications are potentially massive: expanding the space of application for classical algorithms, aiding novel discoveries in theoretical computer science, and rigorously benchmarking GNNs. For more details, you may consult these papers, and my talk on the area.

Lately, I have also been highly fascinated by the amazing things that happen when these methods interact synergistically with domain experts. The best example of this is our Nature paper on guiding human intuition with AI.

I would be happy to discuss any project idea at the intersection of the above areas; please check my contact page if you are interested!

Conference Publications

Journal Publications

Workshop Publications