Alessandro Trenta personal page
Hello! My name is Alessandro Trenta and I am a PhD student in Artificial Intelligence at University of Pisa, under the supervision of prof. Davide Bacciu and Andrea Cossu. My research focuses on the topics of Learning and Transferring Physical Systems, particularly studying the core information for Physical Knowledge. In our Lab, we also develop Graph Neural Networks capable of long-range propagation by exploiting dynamics inspired from physical systems. Currently, I am a Visiting PhD student at the University of Amsterdam, working on Causality for dynamical systems under the supervision of prof. Sara Magliacane. I am a M.S. graduate in Mathematics at University of Pisa, where I focused on Probability, Statistics, Calculus and Machine Learning. I graduated with a thesis on Reinforcement Learning for Solving Functional Equations in CFTs under the supervision of prof. Davide Bacciu. I also hold a Master in Mathematics at Scuola Normale Superiore, where I worked with prof. Stefano Marmi on Information Theory and Market Portfolios. Previously, I was a Data Scientist at Electra Vehicles inc., where I worked on fault detection for electric batteries and smart charging.
News
- 2026-02 Two papers accepted at ICLR 2026! ComPhy: Composing Physical Models with end-to-end Alignment and Improving Long-Range Interactions in Graph Neural Simulators via Hamiltonian Dynamics. See you in Rio de Janeiro :BR:
- 2026-01 Learning and Transferring Physical Models through Derivatives, is accepted at TMLR!
- 2025-12 SONAR - Long Range Graph Propagation through Information Waves is accepted at NeurIPS 2025! I will be presenting the work at EurIPS in Copenhagen (DK).
Useful Links
- My Master Thesis: PDF
- My Bachelor Thesis: PDF
- My solutions for written exams of the Meccanica Razionale course at University of Pisa: Drive
- Notes by Giacomo Mezzedimi: Personal page
- Notes by Simone Cappellini: Personal page
- Notes for Elementi di Teoria degli insiemi by Luca Tonelli: Notes
- Notes for Analisi 3 by Emanuele Angilè et al.: Notes
