Nina Nusbaumer

Home page

About Me

I am a PhD student specializing in computational psycholinguistics and machine learning. My research focuses on human-inspired sentence processing in (large) language models. I am particularly interested in exploring the extent to which these models align with human cognition, with the broader goal of leveraging them to deepen our understanding of human language and language impairments. I work at the Laboratoire de Linguistique Formelle (LLF), CNRS, at Université Paris Cité under the supervision of Benoît Crabbé and Guillaume Wisniewski. My PhD is part of the ANR-funded project COMPO where, together with other research groups, we explore ways to embed inductive biases related to compositionality and memory constraints in language models, in order to design models that are more efficient, using less data and fewer computational resources.

News

• December 2025: I’ll be presenting our new benchmark dataset for evaluating language models against human processing behavior at the Computational Psycholinguistic Meeting!
  Utrecht, Netherlands
• July 2025: I’ll be presenting our paper Évaluer la capacité des transformeurs à distinguer les significations compositionnelles et idiomatiques d’une même expression at TALN!
  Marseille, France
• June 2025: I’m co-organizing the 2025-edition of the REtreat for Students in Computational Linguistics (REST-CL)
  Tarragona, Spain

Education

• PhD in Computational Psycholinguistics
  LLF, CNRS, Université Paris Cité, 🇫🇷
  Bridging cognitive science and machine learning, my doctoral work focuses on modeling human-like sentence processing in language models.
• MA (Res) in Computational Linguistics
  Université Paris Cité, 🇫🇷
  This program strengthened my technical foundation in NLP and machine learning, laying the groundwork for my current research.
• MA (Res) in Multilingual Communication
  Université de Liège, 🇧🇪 & Universidad Complutense de Madrid, 🇪🇸
  Focused on English, German, and Spanish linguistics, philology, and discourse analysis, this program also offered training in communication strategy and entrepreneurial skills — broadening both my academic and professional horizons.
• BA in Philological Studies
  Aix-Marseille Université, 🇫🇷 & Universität Leipzig,🇩🇪
  I began with a strong foundation in the structure, history, and diversity of languages, developed through immersion in both Romance and Germanic philologies.

Research Stays

• Visiting researcher at Computational Linguistics and Linguistic Theory (COLT), Universitat Pompeu Fabra, Barcelona, 🇪🇸 (January 2025 - )

Research Interests

• Human and artificial language processing
• Language model interpretability
• Psycholinguistics and cognitive neuroscience of language

Peer-reviewed publications

• [Published] Nusbaumer, N., Wisniewski, G., Crabbé, B., 2025. Évaluer la capacité des transformeurs à distinguer les significations compositionnelles et idiomatiques d’une même expression, TALN pdf.

Conference talk and poster presentations (with peer-reviewed abstract or paper)

• [Poster] Nusbaumer, N., Wisniewski, G., Crabbé, B., 2025. Évaluer la capacité des transformeurs à distinguer les significations compositionnelles et idiomatiques d’une même expression, TALN
• [Presentation] Shafiabadi, N., Nusbaumer, N., Bonami, O., 2024. Assessing the import of morphology in gender assignment: A computational study of French, LLcD.

Contact

• Email: first.last@gmail.com
• 
• 
•