CALAMITA: Challenge the Abilities of LAnguage Models in ITAlian
About
Federico Borazio
PhD Candidate - Department of Enterprise Engineering / University of Rome Tor Vergata, Italy
Applied Scientist II Intern - Amazon AGI, California, U.S.A.
Federico Borazio is a PhD candidate in the Data Science program at the University of Rome Tor Vergata, where he is a member of the SAG Art research group. His doctoral research is supervised by Roberto Basili and Danilo Croce.
His main research interests include Large Language Models fine-tuning and adaptation for downstream and knowledge-intensive tasks, question answering in the biomedical domain, retrieval-augmented generation, and multimodal dialogue systems. He has contributed to projects on LLM benchmarking, human-robot interaction, and AI applications for epidemic intelligence, publishing in venues such as ACL Findings, ECIR, LREC-COLING, and the Italian Journal of Computational Linguistics.
He is also part of the CALAMITA Data and Evaluation Team, contributing to the benchmarking of Large Language Models on native Italian tasks.
In addition to his academic work, he collaborates with Reveal S.r.l., an industry partner focused on AI and NLP driven solutions.
Currently, he is an Applied Scientist Intern at Amazon in Los Angeles, working on LLM experimentation and neural retrieval for local search projects with the Alexa Local Search team supervised by Alessandro Moschitti.
Publications
Training Multi-Modal LLMs through Dialogue Planning for HRI
Adapting LLMs for Domain-Specific Retrieval: A Case Study in Nuclear Safety
Integrating AI and IR Paradigms for Sustainable and Trustworthy Accurate Access to Large Scale Biomedical Information
UniTor at BioASQ 2025: Modular Biomedical QA with Synthetic Snippets and Multiple Task Answer Generation.
Challenging the Abilities of Large Language Models in Italian: a Community Initiative
MM-IGLU-IT: Multi-modal Interactive Grounded Language Understanding in Italian
Semi-automatic topic discovery and classification for epidemic intelligence via large language models
ArTLLaMA: Adaptating LLaMA to Performative Art Applications
Intelligent Natural Language Processing for Epidemic Intelligence
UniTor at DeSegMa-It: Analyzing Supervision and Encoder Representations for Italian Machine-Generated Text Detection
UniTor at PFB: Constrained Agentic Prompting and Self-Consistency for Financial Multi-Choice QA
No publications match the current filter.
Conferences & summer schools
Conference talks, posters and awards, plus summer schools attended, across Europe. Select a marker for the papers and venue.
Athens, Greece
Summer schoolJuly 2024
ESSAI & ACAI 20242nd European Summer School on Artificial Intelligence (ESSAI) and 21st Advanced Course on Artificial Intelligence (ACAI), 15–26 July 2024. Attended.
- Talk
- Poster
- Award
- Summer school
- Delft, Netherlands — ECIR 2026, March 2026. Presented by Prof. Danilo Croce. Paper: Integrating AI and IR Paradigms for Sustainable and Trustworthy Accurate Access to Large Scale Biomedical Information
- Bari, Italy — EVALITA 2026, February 2026. Honorable Mention for UniTor at DeSegMa-It. Papers: UniTor at DeSegMa-It: Analyzing Supervision and Encoder Representations for Italian Machine-Generated Text Detection; UniTor at PFB: Constrained Agentic Prompting and Self-Consistency for Financial Multi-Choice QA
- Rome, Italy — Hyperstage 2025, November 2025. Presented by Elisa Passone. Paper: RAGArTLLaMA
- Madrid, Spain — BioASQ @ CLEF 2025, September 2025. Paper: UniTor at BioASQ 2025: Modular Biomedical QA with Synthetic Snippets and Multiple Task Answer Generation
- Vienna, Austria — ACL 2025, July 2025. Paper: Training Multi-Modal LLMs through Dialogue Planning for HRI
- Pisa, Italy — CLiC-it 2024, December 2024. Paper: CALAMITA: Challenge the Abilities of LAnguage Models in ITAlian
- Bolzano, Italy — AIxIA 2024, November 2024. Paper: MM-IGLU-IT: Multi-modal Interactive Grounded Language Understanding in Italian
- Lucca, Italy — ECIR 2025, April 2025. Paper: Adapting LLMs for Domain-Specific Retrieval: A Case Study in Nuclear Safety
- Torino, Italy — LREC-COLING 2024, May 2024. Paper: Semi-automatic topic discovery and classification for epidemic intelligence via large language models
- Athens, Greece — ESSAI & ACAI 2024, July 2024. 2nd European Summer School on Artificial Intelligence (ESSAI) and 21st Advanced Course on Artificial Intelligence (ACAI), 15–26 July 2024. Attended.
- Siena, Italy — Digital Heritage 2025, September 2025. Presented by Elisa Passone. Paper: ArTLLaMA: Adaptating LLaMA to Performative Art Applications
Reveal s.r.l. - Industry Projects Collaboration
Matchmaking for Innovation - 2024/25
Simple Knowledge Organization Systems for the Decommissioning of Nuclear Facilities - 2024
Italian Network of Epidemic Intelligence - 2023/24
Teaching Assistant
Contacts
- 📧 Email: borazio@ing.uniroma2.it
- 🏛️ Office: Via del Politecnico 1, Information Engineering Building, Block D, 2nd Floor, Room 20