Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces

arXiv:2212.14855v2 Announce Type: replace
Abstract: Explainable AI aims to overcome the black-box nature of complex ML models like neural networks by generating explanations for their predictions. Explanations often take the form of a heatmap identifying input features (e.g. pixels) that are relevant to the model's decision. These explanations, however, entangle the potentially multiple factors that enter into the overall complex decision strategy. We propose to disentangle explanations by extracting at some intermediate layer of a neural network, subspaces that capture …

abstract arxiv box cs.ai cs.cv cs.lg decision explainable ai features form heatmap however ml models nature network networks neural network neural networks pixels predictions type