LLM-Generated Pair-Programming Dialogue Responses

This dataset contains LLM-generated responses to pair-programming dialogues between students. The dataset includes the human ground-truth responses. The responses were rated by human volunteers according to Coherence, Collaborativeness, and Humanness. The researchers also annotated coreference in the LLM responses; coreference annotations in the human responses were obtained from a larger dataset (see citations below) and were annotated by a team of trained annotators.Details about the dataset collection can be found in the following publications:* Domingo, C., Piwek, P., Stoyanchev, S., Wermelinger, M., Adhikari, K., \& Doddipatla, R. S. (2025). Human ratings of LLM response generation in pair-programming dialogue. In L. Flek, S. Narayan, L. H. Phuong, \& J. Pei (Eds), Proceedings of the 18th International Natural Language Generation Conference (pp. 41–59). Association for Computational Linguistics. https://aclanthology.org/2025.inlg-main.3/* Domingo, Cecilia (2026). Automated Processing of Referring Acts in Pair-Programming Dialogue. PhD Thesis. The Open University. Milton Keynes, England.