Photo of İsmail İlkan Ceylan

İsmail İlkan Ceylan

Principal Investigator

Associate Professor, TU Wien · Adjunct PI, AITHYRA

İsmail is an Associate Professor at TU Wien, an Adjunct PI at AITHYRA, and an Academic Member at the University of Oxford, where he was previously Assistant Professor. His research investigates how AI can learn and reason reliably in relational domains, with a particular emphasis on scientific discovery and mathematical reasoning. He received his PhD from TU Dresden; his dissertation won the E. W. Beth Dissertation Prize.

Alicja Maksymiuk

PhD Student

University of Oxford · since 2024

Alicja is a DPhil student at the University of Oxford working on deep learning for molecular simulation dynamics. [TODO: extend bio.]

Co-supervised with: Michael Bronstein

Angelo Gnazzo

PhD Student

TU Wien · since 2026

Angelo is a PhD student at TU Wien working on geometric generative learning. [TODO: extend bio.]

Ben Finkelshtein

PhD Student

University of Oxford · since 2022

Ben is a DPhil student at the University of Oxford working on expressivity and generalization in graph machine learning. [TODO: extend bio.]

Co-supervised with: Michael Bronstein

Kacper Wyrwal

PhD Student

University of Oxford · since 2026

Kacper is a DPhil student at the University of Oxford, starting in fall 2026. He completed his MSc thesis on topological flow matching in the group. [TODO: confirm role / extend bio.]

Co-supervised with: Michael Bronstein, Alexander Tong

Katarina Petrović

PhD Student

University of Oxford · since 2024

Katarina is a DPhil student at the University of Oxford working on geometric deep learning for dynamical systems. [TODO: extend bio.]

Co-supervised with: Michael Bronstein

Krzysztof Olejniczak

PhD Student

University of Oxford · since 2025

Krzysztof is a DPhil student at the University of Oxford working on AI for mathematics. He completed his MSc thesis in the group on graph neural networks for answering conjunctive queries, which was awarded the Gibbs Prize. [TODO: extend bio.]

Co-supervised with: Bernardo Cuenca Grau

Photo of Louis Tichelman

Louis Tichelman

PhD Student

TU Wien & AITHYRA · since 2026

Louis is a PhD student at TU Wien and AITHYRA working on graph foundation models and relational learning for the Life Sciences. He previously studied Maths in Bonn and Aachen.

Co-supervised with: Michael Bronstein

Nacho Fernandez Mullin

PhD Student

University of Oxford · since 2025

Nacho is a DPhil student at the University of Oxford working on geometric deep learning for structural steel design. [TODO: extend bio.]

Co-supervised with: Barbara Rossi

Oscar Davis

PhD Student

University of Oxford · since 2023

Oscar is a DPhil student at the University of Oxford working on discrete generative modeling. His MSc thesis in the group, on information-theoretic perspectives for graph neural networks, was awarded the Hoare Project Prize. [TODO: extend bio.]

Co-supervised with: Michael Bronstein

Xingyue Huang

PhD Student

University of Oxford · since 2023

Xingyue is a DPhil student at the University of Oxford working on knowledge graph foundation models. [TODO: extend bio.]

Co-supervised with: Michael Bronstein

Ayhan Suleymanzade

External Collaborator

PhD student at EPFL

Ayhan is a PhD student at EPFL and an external collaborator of the group. He works on how models can reason more efficiently by spending computation in a latent space, from continuous reasoning with multiplexed tokens (MUX) in language models to looped flows that train recurrent computation with denoising objectives.

Long-term collaborators

Alumni — where are they now

Andrius Vaičenavičius

DPhil, 2018–2021

Explanations for query answers in ontologies

Now: [TODO: current position]

Emily Jin

DPhil, University of Oxford

Graph machine learning over biomedical knowledge graphs for drug discovery

Now: [TODO: current position]

Ralph Abboud

DPhil, 2018–2022

Learning and inference over relational data

Now: [TODO: current position]

Tommaso Salvatori

DPhil, 2018–2022

Learning and memorization via predictive coding

Now: [TODO: current position]