Professional Summary
I am a PhD Student in AI and Cybersecurity at Scuola Superiore Sant’Anna. My research focuses on stabilizing Large Language Model alignment through Reinforcement Learning (e.g., GTPO) and detecting software vulnerabilities.
Education
PhD in AI and Cybersecurity
2024-10-01
Scuola Superiore Sant'Anna
Master in Physics of Complex Systems and Big Data
2021-09-01
2024-07-01
University of Tor Vergata
Bachelor in Physics
2018-09-01
2021-07-01
University of Tor Vergata
Interests
Deep Learning
Natural Language Processing (NLP)
Large Language Models (LLMs)
Reinforcement Learning (RL)
Policy based techniques (e.g. PPO, GRPO)
Physics of Complex Systems
Fractal Analysis
Cybersecurity
Videogames
Experience
AI Consultant
Netgroup S.p.A., Naples
Architected and trained a custom Transformer-based model from scratch leveraging proprietary tabular datasets to predict and classify cybersecurity attacks.
AI/NLP Research Intern
Serco Italia S.p.A, Rome
Engineered and fine-tuned a specialized LLM for deployment on resource-constrained hardware, implementing advanced optimization techniques to reduce model footprint and latency.
Head of Board & Community Manager
Element Gaming
Led community operations for a non-profit multigaming association, coordinating a network of 100+ affiliated content creators.
Education
PhD in AI and Cybersecurity
Scuola Superiore Sant'Anna
Current research focusing on LLM alignment, trajectory-based policy optimization (GTPO), and cyber threat intelligence.
Master in Physics of Complex Systems and Big Data
University of Tor Vergata
Thesis: “Exploring the frontier of chatbot technology”
Bachelor in Physics
University of Tor Vergata
Thesis: “Quantum computing and cosmology of the early universe”
Conference Papers
(2025).
Unmasking model behavior: How llms reason on vulnerability detection.
International Conference on Availability, Reliability and Security.
Preprints & Working Papers
(2025).
GTPO: Stabilizing Group Relative Policy Optimization via Gradient and Entropy Control.
arXiv preprint arXiv:2508.03772.
(2025).
Improving LLM Reasoning for Vulnerability Detection via Group Relative Policy Optimization.
arXiv preprint arXiv:2507.03051.
(2025).
TITAN: Graph-Executable Reasoning for Cyber Threat Intelligence.
arXiv preprint arXiv:2510.14670.