Tug-of-war between idioms’ figurative and literal interpretations in LLMs

A mechanistic interpretability study of how pretrained causal transformers represent and resolve the competition between an idiom’s literal and figurative readings.

Recommended citation: Oh, S., Huang, X., Pink, M., Hahn, M., & Demberg, V. (2026). "Tug-of-war between idioms’ figurative and literal interpretations in LLMs." Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), 2942–2958. https://aclanthology.org/2026.eacl-long.135/