Large Language Models (LLMs) and chatbots can increasingly be used as tools for creating clear, human-readable textual explanations of AI decisions. LLMs can translate complex model outputs into understandable narratives that highlight key features, potential causal factors, or stepwise reasoning behind the AI’s conclusion.
Textual explanations can be employed on top of other explainability methods, for example, for post-hoc explanations like LIME to add another explanation modality.