Filippo Guerranti
PhD researcher | Machine Learning
I am a PhD researcher at TUM, focusing on machine learning for graphs and structured data. I am currently a Research Scientist Intern at NVIDIA, working on foundation models for structured data.
Outside research, I'm interested in endurance sports, science-based approaches to personal development, and the balance between mental and physical well-being. Read more about me and my approach to work and life.
Recent News
I have joined LeadTheFuture, a nonprofit STEM mentorship program, as a volunteer mentor. I will support students interested in machine learning, research, and career development.
On September 14, I started a Research Scientist Internship at NVIDIA, working on foundation models for structured data.
SAFT has been accepted for publication in Transactions on Machine Learning Research (TMLR). We introduce a structure-aware fine-tuning method that injects graph topology into pretrained large language models without modifying their architecture. Joint work with Rafiq Kamel, Simon Geisler, and Stephan Günnemann.
On May 1, I started a Research Scientist Internship at Microsoft Research Cambridge, working on parallel decoding for autoregressive LLMs to accelerate inference.
I attended the Spring Into Quant Finance 2026 program organized by G-Research in Palermo, selected from roughly 8,000 applicants (1.02% acceptance rate). The event brought together PhD students from across Europe for an introduction to quantitative finance and the role of machine learning, mathematics, and data science in systematic trading.