About Me
Hello there, I’m Sara!
I’m a second-year PhD student in Natural Language Processing at the University of Trieste (Italy), where I work within the AI Lab under the supervision of Luca Bortolussi and Gabriele Sarti.I’m currently visiting the NLP group of the Center for Language and Cognition (CLCG).
My current research focuses on understanding the reasoning abilities of large language models (LLMs), identifying their limitations, and exploring ways to encourage generalization that goes beyond memorization. I’m particularly interested in Mechanistic Interpretability, Reinforcement Learning paradigms and Neuro-Symbolic integration. You can read more in my scholar personal profile. 🧠
I hold a Bachelor’s degree in Statistics and a Master’s degree in Artificial Intelligence. My Master’s thesis explored how multi-agent systems can enhance the performance of extremely compact models on Question Answering tasks and received a Mention of Honor from the Italian Association of Computational Linguistics (AILC). 🤝
Outside of research, I’m the vice-president of the AI Student Society, and previously I worked as a research intern at Rachael S.r.l..
- Reasoning in LLMs
- Reinforcement Learning approaches for NLP
- Neuro-Symbolic Integration
- Multi-Agent LLMs
- Generalization in LLMs
PhD in Applied Data Science and Artificial Intelligence
University of Trieste
MSc in Data Science and Scientific Computing (LM-44)
University of Trieste
BSc in Statistics (L-41)
University of Trieste
I’m interested in understanding the generalization abilities of large language models (LLMs) — identifying their limitations and exploring ways to encourage reasoning beyond memorization.
My work sits at the intersection of Neuro-Symbolic AI, Reinforcement Learning for NLP, and multi-agent systems. I’m particularly fascinated by how symbolic structure can guide and constrain neural learning.
Feel free to reach out if you’d like to collaborate! 😃