Giulia Rambelli
I am a Post-doc for the Institute of Language, Communication, and the Brain (ILCB) at Aix-Marseille University (France), hosted by the TALEP research group (LIS – Laboratoire d’Informatique et Systèmes) and Laboratoire Parole et Langage.
In my project "ANALOGY – Modeling Analogical Generalization in Children's Creative Compound Formation through Computational Models", I investigate the role of analogical mechanisms in linguistic generalization, with a particular focus on linguistic productivity. Taking a computational perspective on how children comprehend and produce novel noun–noun compounds, the research aims to shed light on the cognitive and linguistic processes underlying lexical innovation and the formation of new linguistic combinations. I am collaborating with prof. Abdellah Fourtassi and prof. Laurent Prévot.
In my project "ANALOGY – Modeling Analogical Generalization in Children's Creative Compound Formation through Computational Models", I investigate the role of analogical mechanisms in linguistic generalization, with a particular focus on linguistic productivity. Taking a computational perspective on how children comprehend and produce novel noun–noun compounds, the research aims to shed light on the cognitive and linguistic processes underlying lexical innovation and the formation of new linguistic combinations. I am collaborating with prof. Abdellah Fourtassi and prof. Laurent Prévot.
I am a computational and cognitive linguist working at the intersection of artificial intelligence, theoretical linguistics, and human communication. My research investigates how meaning is constructed, generalized, and innovated in language, both by humans and (large) language models, through the study of compositionality, analogy, and conceptual abstraction. I combine deep linguistic expertise with computational methods to advance our understanding of computational models and language.
News
- I’m really happy to share that, from January 2027, I’ll be starting a permanent position as a Lecturer in AI, Language and Human communication at UCL Institute of Education.
- 26.10.2026 - I was invited to join the CIFAR Horizon Networks meeting in Barcelona (Spain) in the Learning through conversation track.
- 29.07.2026 - I gave an oral presentation "Input to Productivity: Relational Abstraction in Compound Formation with Child-Directed Speech Language Models" at ESLP in Lancaster.
- 12.01.2026 - I was invited to give a talk at the Institute for Natural Language Processing (IMS) at the University of Stuttgart.
Latest publications
- Sommerauer, Pia, Rambelli, Giulia, Caselli, Tommaso (2025). Simulating Identity, Propagating Bias: Abstraction and Stereotypes in LLM-Generated Text. In *Findings of EMNLP 2025 (Suzhou, China)*
- Andrea Pedrotti, Giulia Rambelli, Caterina Villani, Marianna Bolognesi (2025). How Humans and LLMs Organize Conceptual Knowledge: Exploring Subordinate Categories in Italian. In *ACL 2025 (Vienna, Austria)*
- Giulia Rambelli (2024). Constructions and Compositionality: Cognitive and Computational Explorations. Cambridge University