Dataset and materials for “Assessing LLMs’ pragmatic competence in irony: recognition, comprehension, and production tasks”

This repository contains the study materials, prompts, and derived data used in a study that evaluates large language models’ (LLMs) pragmatic competence in irony across three modules: irony recognition, irony comprehension, and irony production. The dataset is intended to support transparency, replication, and secondary analyses.