I am Hugo Attali, a postdoctoral researcher at CentraleSupélec, Université Paris-Saclay, working with Prof. Fragkiskos Malliaros. Previously, I completed my Ph.D. at LIPN (Sorbonne Paris Nord University). My current research focuses on Graph Neural Networks and, more broadly, the field of Graph Representation Learning.
NEWS :
July 2027 : I successfully defended my thesis!!
June 2026: I gave a talk at the Learning on Graphs (LoG) Conference 2025 — Paris Local Meetup, held on Monday, June 15th. My talk was titled "When Should GNNs Look Further? Rethinking Rewiring and Non-Local Dependencies."
June 2026: I presented a seminar titled "Better Understanding Rewiring and Over-Squashing in Graph Neural Networks" to the Data Science and Mining (DaSciM) team at the Computer Science Laboratory (LIX) of École Polytechnique.
March 2025: During my one-month research stay at the University of Seville, I gave a seminar on message passing in Graph Neural Networks and its limitations at the lenguajes y sistemas informaticos laboratory.
October 2024 : I presented a seminar titled "Improving Graph Neural Networks through Graph Rewiring" to the Data Science and Mining (DaSciM) team at the Computer Science Laboratory (LIX) of École Polytechnique.