PinnedMichael BronsteininTowards Data ScienceThe Road to Biology 2.0 Will Pass Through Black-Box DataFuture bio-AI breakthroughs will arise from novel high-throughput low-cost AI-specific “black-box” data modalities.Mar 182Mar 182
Michael BronsteininTowards Data ScienceCo-operative Graph Neural NetworksA new message-passing paradigm where every node can choose to either ‘listen’, ‘broadcast’, ‘listen & broadcast’ or ‘isolate’.Dec 6, 20232Dec 6, 20232
Michael BronsteininTowards Data ScienceTopological Generalisation with Advective Diffusion TransformersA new diffusion-based continuous GNN model offers better generalisation capabilitiesOct 19, 20231Oct 19, 20231
Michael BronsteininTowards Data ScienceDynamically rewired delayed message passing GNNsDynamic rewiring and delayed message passing mechanisms offer a tradeoff between standard MPNNs and graph TransformersJun 19, 2023Jun 19, 2023
Michael BronsteininTowards Data ScienceDirection Improves Graph LearningHow a wise use of direction when doing message passing on heterophilic graphs can result in very significant gains.Jun 8, 20234Jun 8, 20234
Michael BronsteininTowards Data ScienceHyperbolic Deep Reinforcement LearningMany RL problems have hierarchical tree-like nature. Hyperbolic geometry offers a powerful prior for such problems.Apr 30, 20235Apr 30, 20235
Michael BronsteininTowards Data ScienceLearning Network GamesHow to learn the network underlying the interactions of players in social applications, economics, and beyond.Apr 20, 2023Apr 20, 2023
Michael BronsteininTowards Data ScienceGraph Neural Networks as gradient flowsGNNs derived as gradient flows minimising a learnable energy that describes attractive and repulsive forces between graph nodes.Oct 14, 20222Oct 14, 20222
Michael BronsteininTowards Data ScienceTowards Geometric Deep Learning IV: Chemical Precursors of GNNsIn the last post in the “Towards Geometric Deep Learning” series, we look at early prototypes of GNNs in the field of chemistry.Jul 25, 20222Jul 25, 20222
Michael BronsteininTowards Data ScienceTowards Geometric Deep Learning III: First Geometric ArchitecturesIn the third post of our series “Towards Geometric Deep Learning” we look at the first “geometric” architectures: Neocognitron and CNNsJul 18, 2022Jul 18, 2022