About
Summary
As a PhD student at the University of Groningen, I work to advance our understanding of generative language models' inner workings, with the ultimate goal of making human-AI interaction more transparent and trustworthy. I focus, in particular, on machine translation applications since the interaction between automatic translation systems and human post-editors is a promising testbed for my studies.
I previously completed a master's degree in data science at the University of Trieste, where I used human gaze tracking data to augment natural language processing systems capabilities from human preferences. I have published several papers on deep learning and NLP at international venues such as ACL, ICLR and EMNLP, and I had the chance to apply my knowledge in industrial settings at the synthetic data startup Aindo and Amazon AWS AI Labs. I am passionate about positive social applications of new technologies and an advocate for good software development practices and open source collaboration.
Positions
Applied Scientist Intern Jul 2022 - Sep 2022
Summer internship in the Amazon Translate team. Manager: Georgiana Dinu. Research project on prompting large multilingual language models for attribute-controlled machine translation.
Doctoral Researcher Sep 2021 -
Center for Language and Cognition (CLCG), University of Groningen
Part of the NWO-funded project "InDeep: Interpreting Deep Learning Models for Text and Sound". Project focusing on user-centric interpretability perspectives for neural machine translation. Supervisors: Arianna Bisazza, Malvina Nissim and Grzegorz Chrupała.
Research Scientist Nov 2020 - Jul 2021
Research and development of innovative approaches in artificial intelligence. Worked on development, performance assessment, productionization and optimization for the following projects:
- Structured predictions for the biomedical domain using Transformers for sequence-to-sequence question answering.
- Experimental work in cross-linguistic transfer to Italian for pre-trained neural language models.
- Temporally-aware multi-interaction neural recommender systems for online clothes resellers.
- Graph and attention-based neural networks for molecular synthesis and molecular properties prediction.
- Controllable sketch-to-image synthesis for the fashion industry.Research and development of innovative approaches in artificial intelligence. Worked on development, performance assessment, productionization and optimization for the following projects: - Structured predictions for the biomedical domain using Transformers for sequence-to-sequence question answering. - Experimental work in cross-linguistic transfer to Italian for pre-trained neural language models. - Temporally-aware multi-interaction neural recommender systems for online clothes resellers. - Graph and attention-based neural networks for molecular synthesis and molecular properties prediction. - Controllable sketch-to-image synthesis for the fashion industry.
Research Assistant in Natural Language Processing Oct 2019 - Dec 2019
Institute for Computational Linguistics (ILC-CNR)
Working within the ItaliaNLP Lab on applications of machine learning and deep learning to psycholinguistics, with a focus on augmenting systems used to detect linguistic complexity with human cognitive data to improve their performances.
Machine Learning Intern Feb 2018 - Jun 2018
Development of a deep learning based handwritten character recognition system to improve a automatic document archiviation platform adopted by several international academic institutions.
Development of sentiment analysis platform to highlight inappropriate comments made by University of Montreal students to improve the quality of professors' feedback.
Research Assistant Oct 2015 - Aug 2017
Optimization and development of a C++ controller software for the production of tridimensional multi-layered textile products for the aeronautic industry.
Design and development of a verification procedure for a new software used for validating lab analysis on textile and geosynthetic products. Macro creation and Visual Basic programming for Microsoft Excel.
Optimization, debugging and development of an ASP.NET MVC website written in C# linked to a Microsoft SQL relational database. This website is used to manage research center's organizational aspects.
Education
University of Groningen 2021 -
Field of study: Natural Language Processing
Degree: Ph.D.
Part of the NWO-funded project "InDeep: Interpreting Deep Learning Models for Text and Sound". Project focusing on interpretability for neural machine translation. Supervisors: Arianna Bisazza, Malvina Nissim and Grzegorz Chrupala.
University of Trieste and SISSA 2018 - 2020
Field of study: Data Science and Artificial Intelligence
Degree: Master's degree in Data Science
Data science and Scientific Computing courses in collaboration with the International School of Advanced Studies (SISSA), University of Udine (UniUD) and the International Centre of Theoretical Physics (ICTP). Data Science for Social Sciences curriculum. Courses were taught and assessed in English.
Cégep of Saint Hyacinthe 2015 - 2018
Field of study: Management Informatics
Degree: Diplome d'études collégiales (DEC)
Located in Saint-Hyacinthe, Québec (CA). All the subjects are taught and assessed in French.
Professional interests
Mechanistic study of artificial neural networks
Keywords: Artificial Intelligence, computational cognitive science , Machine LearningFellowships
eScience Center Fellowship 2023 - 2024
Grants
Imminent Language Technology Innovator Grant 2023 - 2025
Languages
Italian, French, English, Spanish