Hi! Welcome to the website of the graph representation learning reading group at Mila - Quebec AI Institute. We cover papers a wide range of topics, spanning theory, methods, and (industrial) applications.

Feel free to join us (Zoom link) and discuss every at the interesection of graphs and machine learning!

Upcoming and past talks

Upcoming and past talks are listed below. If possible, we will make slides and video recordings available after the talk.

Date (Eastern Time) Presenter Topic Materials
September 23, 2021 @ 4:00 PM RG Organizers Design Space for Graph Neural Networks [paper]
October 7, 2021 @ 4:00 PM Alexander Tong, Yale University Diffusion Earth Mover’s Distance on graphs [1] [2] [video]
October 14, 2021 @ 4:00 PM Dominique Beaini, Valence Discovery Unlocking Deep Learning for Graphs [SAN] [DGN] [PNA]
October 21, 2021 @ 10:30 AM Benedek Rozemberczki, AstraZeneca PyTorch Geometric Temporal: Spatiotemporal Signal Processing with Neural Machine Learning Models [paper] [video]
October 28, 2021 @ 10:00 AM Petar Veličković, DeepMind Neuralising a Computer Scientist: The Story So Far [paper] [video]
November 4, 2021 @ 10:00 AM Matthias Fey, TU Dortmund University Pytorch Geometric 2.0 and Auto-Scaling GNNs [paper] [video]
November 11, 2021 @ 4:00 PM Uri Alon, Carnegie Mellon University How Attentive are Graph Attention Networks? [paper] [video]
November 18, 2021 @ 10:15 AM Bastian Rieck, Helmholtz Centre Munich Learning Topology-Based Graph Representations [1] [2] [3] [video]
November 25, 2021 @ 4:00 PM Joey Bose, Mila & McGill Introduction to Equivariant GNNs [Tensor Field Nets] [Schnet] [EGNN] [PAINN] [GemNet] [SpinConv] [video]
December 2, 2021 @ 10:00 AM Johannes Klicpera, Technical University of Munich Incorporating Directionality in GNNs [DimeNet] [GemNet] [Synthetic coordinates] [video]


If you are interested in presenting your work in the reading group, please reach out to one of the organizers listed above.