Agnese Lombardi
Agnese Lombardi /aɲˈɲe.ze/ as in gnocchi

PostDoc in Computational Psycholinguistics

About Me

I am currently a PostDoc researcher at IUSS Pavia and a member of the NepLab (PI: Valentina Bambini).

I completed my PhD at the Department of Philology, Literature, and Linguistics at the University of Pisa, where I was also a member of the CoLing Lab (I expect to be awarded the PhD title in May). My doctoral research was supervised by Alessandro Lenci and Domenica Romagno.

During my PhD, I was a visiting researcher at the University of Texas at Austin, working under the supervision of Kyle Mahowald.

I earned my BA in Modern Literature (2018) and MA in Linguistics (2021) from the University of Pisa. After graduating, I moved to Naples to work as a mobile application developer, but later left to pursue my PhD (2022).

I enjoy reading, playing football, traveling and cooking—but above all, I love pizza! Proudly from Campania (South of Italy) and even prouder to be a dialect speaker.

Download CV
Interests
  • Computational Linguistics
  • Psycholinguistics
  • Artificial Intelligence
Education
  • PhD Computational Psycholinguistics

    University of Pisa

  • MA in Linguistics

    University of Pisa

  • BSc in Humanities

    University of Pisa

📚 My Research

My primary research focus lies in computational psycholinguistics, with a strong theoretical foundation in semantics and its interface with both morphosyntax and pragmatics. From a computational perspective, I investigate Theory of Mind (ToM)-like abilities in large language models (LLMs), while also exploring other aspects of pragmatic competence.

I am deeply engaged in the ongoing debate on cognitive and emerging abilities in LLMs, advocating for the integration of cognitive science, psycholinguistics, and computational linguistics to advance all three fields. I believe that insights from cognitive science and psycholinguistics can help develop more fine-grained computational analyses of language models.

My broader interests include language processing theories, neural-symbolic networks, grounding, pLoT, experimental pragmatics, mechanistic interpretability, and LLM interpretability.

Publications
(2026). Conversational Implicatures Through the Lens of LLMs. Accepted.
(2026). LLMs and people both learn to form conventions – just not with each other.
(2026). ToM in LLM is not ToM, but a Pragmatic Effect.
Recent & Upcoming Talks
Teaching