Leonardo Cavallaro

Leonardo Cavallaro is a student in Master's Degree in Computer Engineering at Politecnico di Torino.

During his period at ELLIS Alicante, his research is centred on AI Safety and Mechanistic Interpretability under the supervision of Diego Miguel Lozano.

His work specifically focuses on auditing the internal representations and latent activation dynamics of Large Language Models under syntactic variations to identify safety vulnerabilities and build more robust, aligned, and trustworthy AI systems.