Cristian Bodnar

Senior Researcher @ Microsoft Research. Previously @ Google Brain, Google X, Twitter Cortex.

cris_profile.jpg

Hi there! I am a Senior Researcher at Microsoft Research in the AI4Science team where I am working at intersection of deep learning and partial differential equations (PDEs). On the ML side, my research spans a range of topics such as geometric & topological deep learning, graph neural networks and neural differential equations.

Previously, I finished my PhD at the University of Cambridge, supervised by Prof Pietro Liò and supported by a Microsoft Research PhD Fellowship (2021). I’ve also spent significant time in industry as a research intern at Microsoft Research (2022), Twitter Cortex (2021), Google Brain (2020), and as an AI Resident at Google X (2019). In 2019, I graduated with distinction the MPhil in Advanced Computer Science at Cambridge with a Best MPhil Student Award.

news

Mar 2023 I have joined the AI4Science team at Microsoft Research as a Senior Researcher to work on hard problems at the intersection of machine learning and the natural sciences.
Feb 2023 Successfully defended my PhD thesis on Topological Deep Learning. My examiners were Prof Max Welling and Prof José Miguel Hernández-Lobato.
Jan 2023 In August, I will be giving an invited talk at the Mathematics of Geometric Deep Learning Minisymposium at the 10th International Congress on Industrial and Applied Mathematics.
Sep 2022 I mentored a research project on Sheaf Neural Networks at the London Geometry and Machine Learning Summer School. The paper was accepted to the NeurReps workshop at NeurIPS 2022.
Jul 2022 I was a lecturer at the First Italian Summer School of Geometric Deep Learning alongside Michael Bronstein, Francesco Di Giovanni, Pim de Haan and Maurice Weiler. My lectures cover a large part of my upcoming PhD thesis on Topological Deep Learning and argue why topology should play a central role in GDL. The recordings and the slides are available online.

selected papers

  1. thesis2.png
    Topological Deep Learning: Graphs, Complexes, Sheaves
    Cristian Bodnar
    PhD Thesis, University of Cambridge, 2022
    Website
  2. parallel_transport.png
    Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs
    Cristian Bodnar, Francesco Di Giovanni, Benjamin Paul Chamberlain, and 2 more authors
    In Advances in Neural Information Processing Systems, 2022
    arXiv Blog Code
  3. cwn.png
    Weisfeiler and Lehman Go Cellular: CW Networks
    Cristian Bodnar*, Fabrizio Frasca*, Nina Otter, and 4 more authors
    Advances in Neural Information Processing Systems, 2021
    arXiv Blog Code
  4. mpsn.png
    Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
    Cristian Bodnar*, Fabrizio Frasca*, Yuguang Wang*, and 4 more authors
    In International Conference on Machine Learning, 2021
    arXiv Blog Code
  5. ndp_exp.png
    Neural ODE Processes
    Alexander Norcliffe*, Cristian Bodnar*, Ben Day*, and 2 more authors
    In International Conference on Learning Representations, 2021
    arXiv Code
  6. sonode.gif
    On Second Order Behaviour in Augmented Neural ODEs
    Alexander Norcliffe, Cristian Bodnar, Ben Day, and 2 more authors
    Advances in Neural Information Processing Systems, 2020
    arXiv Code
  7. q2opt.gif
    Quantile QT-Opt for Risk-Aware Vision-Based Robotic Grasping
    Cristian Bodnar, Adrian Li, Karol Hausman, and 2 more authors
    In Robotics: Science and Systems (Best Systems Paper Finalist), 2020
    arXiv Website