معرفی
Alessio Gravina is a Researcher at the University of Pisa's Department of Computer Science. His work focuses on graph representation learning, neural differential equations, and dynamic graphs. He holds a PhD in Computer Science from the University of Pisa (2024) and has conducted research visits at institutions including Stanford University, Cambridge, and Huawei Research. Gravina's research emphasizes long-range information propagation in graphs, non-dissipative architectures, and applications in temporal graph learning. His recent work includes the CTAN model for continuous-time dynamic graphs and the GRAMA framework for adaptive graph learning.
Education includes a BSc/MSc from the University of Pisa (cum laude) and an Erasmus+ exchange at University College Dublin. His awards include the Best Student Paper Award at DLG-AAAI 2023 and 1st place in the Fujitsu AI-NLP Challenge (2018). He has contributed to drug repurposing research via deep graph networks and pioneered methods like SWAN and Port-Hamiltonian architectures to combat oversquashing and vanishing gradients in GNNs.
Gravina's collaborations span cybersecurity (Huawei), neuroscience (Stanford), and medical AI. His lab focuses on integrating dynamical systems theory with graph neural networks. He actively publishes in top venues like ICLR, ICML, and NeurIPS workshops, with a focus on open-source tools (e.g., GRAMA GitHub repository).
