Petar Veličković

Petar Veličković is an Affiliated Lecturer at the University of Cambridge's Department of Computer Science and Technology and an Associate of Clare Hall. He holds a PhD from Trinity College, Cambridge, and is a Senior Staff Research Scientist at Google DeepMind. His work focuses on geometric deep learning, particularly graph representation learning and its applications in algorithmic reasoning and mathematical conjectures.

Education: PhD in Computer Science (University of Cambridge, Trinity College), supervised by Pietro Liò.

Research Interests: Geometric Deep Learning, Graph Neural Networks, Symmetry-Invariant Architectures, and Applications in Real-World Systems (e.g., Google Maps Travel-Time Predictions and Mathematical Theorem Discovery).

Notable Contributions: First author of Graph Attention Networks and Deep Graph Infomax, which have been widely adopted in machine learning. His work has been featured in Nature, Science, and major tech publications.

Awards: Recognized as an ELLIS Scholar in the Geometric Deep Learning Program.

Teaching: Co-teaches an MPhil course on geometric deep learning at the University of Cambridge.