Relevant Irrelevance: Generating Alterfactual Explanations for Image Classifiers

arXiv:2405.05295v1 Announce Type: cross
Abstract: In this paper, we demonstrate the feasibility of alterfactual explanations for black box image classifiers. Traditional explanation mechanisms from the field of Counterfactual Thinking are a widely-used paradigm for Explainable Artificial Intelligence (XAI), as they follow a natural way of reasoning that humans are familiar with. However, most common approaches from this field are based on communicating information about features or characteristics that are especially important for an AI's decision. However, to fully understand a …

abstract artificial artificial intelligence arxiv black box box classifiers counterfactual cs.ai cs.cv cs.lg explainable artificial intelligence however humans image intelligence natural paper paradigm reasoning thinking type xai