Works
Talks
Time
CV
Me
(???)
Talks
Pack your subgraphs: A journey into subgraphs for powerful Graph Neural Networks
Yves-Alexandre de Montjoye's Group Talks (Imperial College London)
tl;dr
Our latest two works on subgraphs for more expressive GNNs lined up to tell a coherent story. In the first part of the talk we show the design of a novel framework (ESAN) to process generic bags of subgraphs in an equivariant manner. We show that this approach effectively increases the expressive power of MPNNs. Then, we notice a surge in concurrent approaches which – sometimes unwittingly – make use of subgraphs for powerful graph representations. In the second part, a novel symmetry analysis allows to unify and better characterise these approaches when using a node-based selection policy. We prove an upper-bound on their expressive power and conceive a framework serving as a design space for equivariant node-based subgraph methods.
Based on: https://arxiv.org/abs/2206.11140, https://arxiv.org/abs/2110.02910
Pack your subgraphs: A journey into subgraphs for powerful Graph Neural Networks
Course on Geometric Deep Learning at Oxford University – Lecture
(slides)
tl;dr
Our latest two works on subgraphs for more expressive GNNs lined up to tell a coherent story. In the first part of the talk we show the design of a novel framework (ESAN) to process generic bags of subgraphs in an equivariant manner. We show that this approach effectively increases the expressive power of MPNNs. Then, we notice a surge in concurrent approaches which – sometimes unwittingly – make use of subgraphs for powerful graph representations. In the second part, a novel symmetry analysis allows to unify and better characterise these approaches when using a node-based selection policy. We prove an upper-bound on their expressive power and conceive a framework serving as a design space for equivariant node-based subgraph methods.
Based on: https://arxiv.org/abs/2206.11140, https://arxiv.org/abs/2110.02910
Exploring the practical and theoretical landscape of expressive Graph Neural Networks
Learning on Graphs Conference 2022
tl;dr
The tutorial reviews the most prominent expressive GNNs, categorises them into different families, and draws interesting connections between them. This is accomplished through a series of practical coding sessions and an organic overview of the literature landscape. We aim to convey the importance of studying the expressive power of GNNs and make this field more accessible to our community, especially practitioners and newcomers.
Co-presented with: Beatrice Bevilacqua, Haggai Maron
Understanding and Extending Subgraph GNNs by Rethinking their Symmetries
Learning on Graphs and Geometry Reading Group
tl;dr
Many concurrent works seem to share a similar underlying intuition: to represent a graph by processing subgraphs generated via some simple selection policy. Inspired by a novel symmetry analyses for the most prominent of these methods – those where subgraphs are in a 1:1 correspondence with nodes in the original graph – we prove an upper-bound on their expressive power and conceive a framework serving as a design space for future new equivariant node-based subgraph architectures.
Co-presented with: Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2206.11140
Pack your subgraphs: A journey into subgraphs for powerful Graph Neural Networks
Simone Scardapane's Group Talks (Sapienza)
tl;dr
We line up our latest two works on subgraphs for more expressive GNNs to tell a coherent story. In the first part of the talk we show the design of a novel framework (ESAN) to process generic bags of subgraphs in an equivariant manner. We show that this approach effectively increases the expressive power of MPNNs. Then, we notice a surge in concurrent approaches which – sometimes unwittingly – make use of subgraphs for powerful graph representations. In the second part, a novel symmetry analysis allows to unify and better characterise these approaches when using a node-based selection policy. We prove an upper-bound on their expressive power and conceive a framework serving as a design space for equivariant node-based subgraph methods.
Co-presented with: Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2206.11140, https://arxiv.org/abs/2110.02910
Pack your subgraphs: A journey into subgraphs for powerful Graph Neural Networks
Meta AI orgs Reading Meeting
tl;dr
We line up our latest two works on subgraphs for more expressive GNNs to tell a coherent story. In the first part of the talk we show the design of a novel framework (ESAN) to process generic bags of subgraphs in an equivariant manner. We show that this approach effectively increases the expressive power of MPNNs. Then, we notice a surge in concurrent approaches which – sometimes unwittingly – make use of subgraphs for powerful graph representations. In the second part, a novel symmetry analysis allows to unify and better characterise these approaches when using a node-based selection policy. We prove an upper-bound on their expressive power and conceive a framework serving as a design space for equivariant node-based subgraph methods.
Co-presented with: Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2206.11140, https://arxiv.org/abs/2110.02910
Pack your subgraphs: A journey into subgraphs for powerful Graph Neural Networks
Maks Ovsjanikov's Group Talks (École Polytechnique, France)
tl;dr
We line up our latest two works on subgraphs for more expressive GNNs to tell a coherent story. In the first part of the talk we show the design of a novel framework (ESAN) to process generic bags of subgraphs in an equivariant manner. We show that this approach effectively increases the expressive power of MPNNs. Then, we notice a surge in concurrent approaches which – sometimes unwittingly – make use of subgraphs for powerful graph representations. In the second part, a novel symmetry analysis allows to unify and better characterise these approaches when using a node-based selection policy. We prove an upper-bound on their expressive power and conceive a framework serving as a design space for equivariant node-based subgraph methods.
Co-presented with: Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2206.11140, https://arxiv.org/abs/2110.02910
Subgraphs for more expressive GNNs
Course on Geometric Deep Learning at the African Master's on Machine Intelligence (GDL100) – Seminar
tl;dr
Many concurrent works seem to share a similar underlying intuition: to represent a graph by processing subgraphs generated via some simple selection policy. We discuss these methods together and try to find a common characterisation, while showing that the architecture proposed in our previous work 'Equivariant Subgraph Aggregation Networks' subsumes most of them.
Based on: https://arxiv.org/abs/2110.02910
Graph Representation Learning on Simplicial and Cellular Complexes
Dagstuhl Seminars – Graph Embeddings: Theory meets Practice
tl;dr
I gather together our works on message-passing on Simplicial and Cellular Complexes and present them together showing how these approaches in turn allow for more expressive graph representations.
Based on: https://arxiv.org/abs/2106.12575, https://arxiv.org/abs/2103.03212
Equivariant Subgraph Aggregation Networks
Author Interviews with Zak Jost
tl;dr
We explore the idea of modeling a graph as a bag of subgraphs generated by simple, domain agnostic selection policies, such as node-deletion, edge-deletion, ego-networks. We formally show that this approach lead to the design Graph Neural Networks which can be made strictly more powerful than standard Message Passing ones. We prove this by defining a novel WL variant and by constructing layers equivariant to the emerging symmetry group.
Co-presented with: Derek Lim, Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2110.02910
Subgraph Networks
Third Nepal Winter School in AI
tl;dr
Many concurrent works seem to share a similar underlying intuition: to represent a graph by processing subgraphs generated via some simple selection policy. We discuss these methods together and try to find a common characterisation, while showing that the architecture proposed in our previous work 'Equivariant Subgraph Aggregation Networks' subsumes most of them.
Based on: https://arxiv.org/abs/2110.02910
Equivariant Subgraph Aggregation Networks
Learning on Graphs and Geometry Reading Group
tl;dr
We explore the idea of modeling a graph as a bag of subgraphs generated by simple, domain agnostic selection policies, such as node-deletion, edge-deletion, ego-networks. We formally show that this approach lead to the design Graph Neural Networks which can be made strictly more powerful than standard Message Passing ones. We prove this by defining a novel WL variant and by constructing layers equivariant to the emerging symmetry group.
Co-presented with: Derek Lim, Beatrice Bevilacqua
Based on: https://arxiv.org/abs/2110.02910
Graph Convolutional Network for Disease Prediction with Imbalanced Data – Anees Kazi
London Machine Learning Meetup (Moderator)
tl;dr
I moderate Anees Kazi's talk at the London Machine Learning Meetup.
Weisfeiler and Lehman Go Cellular: CW Networks
Learning on Graphs and Geometry Reading Group
tl;dr
We extend message-passing from graphs and Simplicial Complexes to Cellular Complexes, combinatorial topological spaces generalising these two. By lifting graphs to Cellular Complexes we can naturally model molecules and unlock a series of advantages w.r.t. standard pairwise message-passing.
Co-presented with: Cristian Bodnar
Based on: https://arxiv.org/abs/2106.12575
Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
Course on Geometric Deep Learning at the African Master's on Machine Intelligence (GDL100) – Seminar
(slides)
tl;dr
We re-define colour refinement and message-passing on Simplicial Complexes, topological generalisation of graphs. This allows to model higher-order interactions, model meso-scale structures and increase the expressive power of Graph Neural Networks.
Based on: https://arxiv.org/abs/2103.03212
The expressive power of GNNs by the WL test
London Geometry and Machine Learning Summer School 2021 – Tutorial
tl;dr
What is the current state of affair in provably more expressive Graph Neural Networks? Giorgos Bouritsas and I lead and moderate this interactive discussion on approaches to go beyond 1-WL discriminating power. Ultimately, we would like to tackle the following question: 'Is the Weisfeiler-Leman hierarchy suitable to study the representational power of learning models on graphs? Do we necessarily need to go beyond message-passing?'
Co-presented with: Giorgos Bouritsas
Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
TopoNets 2021 – Networks Beyond Pairwise Interactions
tl;dr
We re-define colour refinement and message-passing on Simplicial Complexes, topological generalisation of graphs. This allows to model higher-order interactions, model meso-scale structures and increase the expressive power of Graph Neural Networks.
Co-presented with: Cristian Bodnar
Based on: https://arxiv.org/abs/2103.03212
Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
Artificial Intelligence Research Group Talks (Cambridge University)
tl;dr
We re-define colour refinement and message-passing on Simplicial Complexes, topological generalisation of graphs. This allows to model higher-order interactions, model meso-scale structures and increase the expressive power of Graph Neural Networks.
Co-presented with: Cristian Bodnar, Yu Guang Wang
Based on: https://arxiv.org/abs/2103.03212
Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
Graph Journal Club (Valence AI)
tl;dr
We re-define colour refinement and message-passing on Simplicial Complexes, topological generalisation of graphs. This allows to model higher-order interactions, model meso-scale structures and increase the expressive power of Graph Neural Networks.
Co-presented with: Cristian Bodnar, Yu Guang Wang
Based on: https://arxiv.org/abs/2103.03212
Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
Math Machine Learning seminar MPI MIS + UCLA
tl;dr
We re-define colour refinement and message-passing on Simplicial Complexes, topological generalisation of graphs. This allows to model higher-order interactions, model meso-scale structures and increase the expressive power of Graph Neural Networks.
Co-presented with: Cristian Bodnar, Yu Guang Wang
Based on: https://arxiv.org/abs/2103.03212
Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting
Graph Journal Club (Valence AI)
tl;dr
We study how isomorphism counting of small substructures can improve the expressive power of Message Passing Neural Networks, while retaining equivariance, a good inductive bias, and their tractable fwd-pass computational complexity.
Co-presented with: Giorgos Bouritsas
Based on: https://arxiv.org/abs/2006.09252
My personal website has been realised through Jekyll and Github Pages. The theme — slightly customised — is by orderedlist. Drop me a message if you'd fancy a chat!