Mechanistic Interpretability Meets Cognitive Linguistics: Modelling Locative Image Schemas in the Circuit Framework

Mattia Proietti, Afra Alishahi, Grzegorz Chrupała, Alessandro Lenci


Abstract
Large Language Models are often considered the best computational testbeds for linguistic theorisation at our disposal. However, their inner workings remain largely opaque, and the mechanisms behind their behaviour cannot always be easily connected with theoretical linguistic assumptions. Mechanistic Interpretability (MI) is surging as a specialised field to reverse engineer models’ internals and shed light on the causal relationships happening under the hood. Nevertheless, MI is predominantly focused on AI-Safety problems, and the attempts to understand linguistically motivated behaviours with these tools are still limited. In this work, we investigate whether an LLM, namely LlaMA-3.2-1b, has developed specialised mechanisms governing the selection of the locative preposition in simple copular clauses. To frame the problem as a next-token prediction objective, we introduce the Stranded Locative Preposition Selection task along with a small dataset aptly curated to test it. We make use of several MI tools to scan the model’s internals and relate their mechanisms to classic theory in Cognitive Linguistics, which assumes that the two basic locative prepositions in and on are the respective linguistic encoding of two different Image Schemas: Containment and Surface
Anthology ID:
2026.lrec-1.885
Volume:
Proceedings of the Fifteenth Language Resources and Evaluation Conference
Month:
May
Year:
2026
Address:
Palma de Mallorca, Spain
Editors:
Stelios Piperidis, Núria Bel, Henk van den Heuvel, Nancy Ide, Simon Krek, Antonio Toral
Venue:
LREC
SIG:
Publisher:
ELRA Language Resource Association
Note:
Pages:
11320–11331
Language:
External URL:
https://lrec.elra.info/lrec2026-main-885
DOI:
10.63317/4iwnnycar4sc
Bibkey:
Cite (ACL):
Mattia Proietti, Afra Alishahi, Grzegorz Chrupała, and Alessandro Lenci. 2026. Mechanistic Interpretability Meets Cognitive Linguistics: Modelling Locative Image Schemas in the Circuit Framework. In Proceedings of the Fifteenth Language Resources and Evaluation Conference, pages 11320–11331, Palma de Mallorca, Spain. ELRA Language Resource Association.
Cite (Informal):
Mechanistic Interpretability Meets Cognitive Linguistics: Modelling Locative Image Schemas in the Circuit Framework (Proietti et al., LREC 2026)
Copy Citation: