The Researcher#
I am Alessandro Zirilli, a PhD student in Computer Science at Sapienza University of Rome (🇮🇹). I have a bachelor’s in Mathematics and a master’s in Computer Science, both from Sapienza. However, during my master’s I was adopted by TUM (🇩🇪), where I spent half of the program as an exchange student, and where I also did my master’s thesis research with Prof. Laura Leal-Taixé.
My research started with Reinforcement Learning, especially applied to games, with my bachelor’s thesis “Deep Learning for Hex”, Computer Vision, with my master’s thesis “Efficient SAM2”, and Model Merging, which is also the topic of my first research paper “MASS”, accepted at ICLR 2026 (actually it’s MoErging, but they are close enough).
Recently I’ve been working on making AI inference private using Fully Homomorphic Encryption (FHE), a mathematical construct that allows computations directly on ciphertexts. Check out “HEAT”, accepted at two NeurIPS 2026 workshops, and “Perseus”, paper coming soon!
The Person#
I’m 24, and I was born and raised in Rome. I am a curious person (who would have guessed it, a curious researcher?!), and I love understanding how things work. I love mathematical formalism and its abstraction, which ease connections between apparently distant things.
I’m also interested in a lot of other things; some examples are: wristwatches, sports (almost any, but especially 🥋⛷️🏋🏻♂️🏊), languages (🇮🇹🇬🇧🇫🇷, my Duolingo is set to 🇩🇪 now), creating websites (like this amazing one 👨💻), but also photography, cooking, history, travelling… I mean, I get passionate about things easily.
I spread my time between a lot of different things, but I always try to keep a balance. 🧘
