About

I'm a Ph.D. student in Computer Science at the University of Milano-Bicocca, advised by Professor Matteo Palmonari. My research focuses on efficient post-training of LLMs.

Before starting the Ph.D., I was a visiting student researcher at the Stanford NLP Group, supervised by Dan Jurafsky and Federico Bianchi, where I applied mechanistic interpretability techniques to LLM safety. I also took part in Neel Nanda's MATS training program. My work on sparse autoencoders and LLM safety has appeared at EMNLP, ICLR, and BlackboxNLP.

Earlier, I worked as a research fellow on NLP for the Italian legal domain, building information extraction and entity linking systems. I hold an M.Sc. in Data Science and a B.Sc. in Statistical and Economic Sciences, both from Milano-Bicocca.