@inproceedings{kunz-kuhlmann-2024-properties,
    title = "Properties and Challenges of {LLM}-Generated Explanations",
    author = "Kunz, Jenny  and
      Kuhlmann, Marco",
    editor = "Blodgett, Su Lin  and
      Cercas Curry, Amanda  and
      Dev, Sunipa  and
      Madaio, Michael  and
      Nenkova, Ani  and
      Yang, Diyi  and
      Xiao, Ziang",
    booktitle = "Proceedings of the Third Workshop on Bridging Human--Computer Interaction and Natural Language Processing",
    month = jun,
    year = "2024",
    address = "Mexico City, Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/moar-dois/2024.hcinlp-1.2/",
    doi = "10.18653/v1/2024.hcinlp-1.2",
    pages = "13--27",
    abstract = "The self-rationalising capabilities of large language models (LLMs) have been explored in restricted settings, using task-specific data sets.However, current LLMs do not (only) rely on specifically annotated data; nonetheless, they frequently explain their outputs.The properties of the generated explanations are influenced by the pre-training corpus and by the target data used for instruction fine-tuning.As the pre-training corpus includes a large amount of human-written explanations ``in the wild'', we hypothesise that LLMs adopt common properties of human explanations.By analysing the outputs for a multi-domain instruction fine-tuning data set, we find that generated explanations show selectivity and contain illustrative elements, but less frequently are subjective or misleading.We discuss reasons and consequences of the properties' presence or absence. In particular, we outline positive and negative implications depending on the goals and user groups of the self-rationalising system."
}