Go directly to page contentent
TEAM LADE

Francesco Ortu

Researcher at Laboratory of Data Engineering

I’m a PhD student at LADE and the University of Trieste, supervised by Alberto Cazzaniga. My research interests lie at the intersection between geometrical and causal tools for interpreting LLMs and their applications in improving AI healthcare systems, with a focus on rare diseases.

Previously, I studied Data Science & Artificial Intelligence at the University of Trieste and Mathematics at the University of Rome, La Sapienza.

Sono uno studente di dottorato presso il gruppo LADE e l’Università di Trieste, sotto la supervisione di Alberto Cazzaniga. I miei interessi di ricerca risiedono nell’intersezione tra strumenti geometrici e causali per l’interpretazione dei LLM e le loro applicazioni nel miglioramento delle applicazioni sanitarie che usano IA, con un focus sulle malattie rare.

In precedenza, ho studiato Data Science & Artificial Intelligence all’Università di Trieste e Matematica all’Università di Roma, La Sapienza.

 

Research Interests 
  • Interpretability of deep learning models
  • NLP in healthcare
  • NLP4Good
Experience & Education
  • MSc in Data Science & Scientific Computing, University of Trieste, 2024
  • BSc in Mathematics, University of Rome “La Sapienza”, 2020
Latest Publications
01/07/2026
When Seeing Overrides Knowing: Disentangling Knowledge Conflicts in Vision-Language Models
Abstract Vision-language models (VLMs) increasingly combine visual and textual information to perform complex tasks. However,…
Go to the news When Seeing Overrides Knowing: Disentangling Knowledge Conflicts in Vision-Language Models
01/03/2026
Preserving Historical Truth: Detecting Historical Revisionism in Large Language Models
Abstract Large language models (LLMs) are increasingly consulted for historical information by citizens, journalists, and…
Go to the news Preserving Historical Truth: Detecting Historical Revisionism in Large Language Models
28/11/2025
Are LLMs Good Safety Agents or a Propaganda Engine?
Abstract Large Language Models (LLMs) are trained to refuse to respond to harmful content. However,…
Go to the news Are LLMs Good Safety Agents or a Propaganda Engine?