Large Language Models Behave (Almost) as Rational Speech Actors: Insights from Metaphor Understanding
Abstract
What are the inner workings of large language models? Can they perform pragmatic inference? This paper attempts to characterize from a mathematical angle the processes of large language models involved in metaphor understanding. Specifically, we show that GPT2-XL model’s reasoning mechanisms can be well predicted within the Rational Speech Act framework for metaphor understanding, which has already been used to grasp the principles of human pragmatic inference in dealing with figurative language. Our research contributes to the field of explainability and interpretability of large language models and highlights the usefulness of adopting a Bayesian model of human cognition to gain insights into the pragmatics of conversational agents.
Cite
Text
Carenini et al. "Large Language Models Behave (Almost) as Rational Speech Actors: Insights from Metaphor Understanding." NeurIPS 2023 Workshops: InfoCog, 2023.Markdown
[Carenini et al. "Large Language Models Behave (Almost) as Rational Speech Actors: Insights from Metaphor Understanding." NeurIPS 2023 Workshops: InfoCog, 2023.](https://mlanthology.org/neuripsw/2023/carenini2023neuripsw-large/)BibTeX
@inproceedings{carenini2023neuripsw-large,
title = {{Large Language Models Behave (Almost) as Rational Speech Actors: Insights from Metaphor Understanding}},
author = {Carenini, Gaia and Bodot, Louis and Bischetti, Luca and Schaeken, Walter and Bambini, Valentina},
booktitle = {NeurIPS 2023 Workshops: InfoCog},
year = {2023},
url = {https://mlanthology.org/neuripsw/2023/carenini2023neuripsw-large/}
}