Program

List of talks 

Aleksandr Kachur : Directed Connectome Analysis with Blurred Magnitude Homology
Anibal Medina-Mardones : Connectivity and global hardness in descent path search
António Leitão : It's All About Covers: Persistent Homology of Cover Refinements
Argyris Kalogeratos : Multi-view diffusion geometry using intertwined diffusion trajectories
Daniel Bennequin : Inner geometries of neural networks, natural or artifial
Edouard Oyallon :  DISCO: Learning to Discover Evolution Operators for Multi-Physics-Agnostic Prediction
Frederic Cazals : Two Topics in High-Dimensional Clustering
Jiayi Li : Basin Selection in Bayesian Learning: A Singular Learning Theory Perspective
Johannes Lutzeyer : Graph Representational Learning: When Does More Expressivity Hurt Generalisation?
Joseph Nardin-Gennequin : Topological Deep Learning for Protein Classification
Josué Tonelli Cueto : Tensor learning with orthogonal, Lorentz, and symplectic symmetries
Juan Pablo Vigneaux :  The geometry of syntax processing in transformer-based language models
Julien Tierny  : Topological Autoencoders++: Fast and Accurate Cycle-Aware Dimensionality Reduction
Levent Doğan : A Minimax Theorem for Convex Optimization on Moment Polytopes
Marzieh Eidi : Geometric learning in complex networks 
Matilde Marcolli  : System and Artificial Models of Language
Matthieu Montes : Protein shape similarity search comparative evaluation and benchmark development
Nina Otter : Algebra and geometry of (hyper)graphs
Olga Anosova : Geometric Data Science exposed thousands of exact duplicates in high-profile structural databases
Olivier Bisson : Log-Euclidean Geometry of Full-Rank Correlation Matrices
Santiago Velasco-Forero : Tropical Neural Networks: Galois Connections, Fixed-Point Operators, and Challenges of Learning from Data
Steve Oudot : Minimum Spanning Trees for Self-Supervised Learning
Tolga Birdal : TBA
Tony Lelièvre : How to combine generative AI models with enhanced sampling methods?
Ugur Canturk : Approximate Symmetries in Graph Neural Networks
Virginia Bolelli : On the Realization of Cyclic–Sequential Dynamics in Cortical Population Models
Yiannis Vlassopoulos : ReLU and Softplus Neural Nets as Zero-Sum, Turn-Based, Stopping Games

Schedule of the conference: 

Regular talks are 35 minutes long + 10 minutes of questions

Short talks are 20 minutes long + 5 minutes of questions