Sparse Concept Bottleneck Models: Gumbel Tricks in Contrastive Learning

arXiv:2404.03323v1 Announce Type: new
Abstract: We propose a novel architecture and method of explainable classification with Concept Bottleneck Models (CBMs). While SOTA approaches to Image Classification task work as a black box, there is a growing demand for models that would provide interpreted results. Such a models often learn to predict the distribution over class labels using additional description of this target instances, called concepts. However, existing Bottleneck methods have a number of limitations: their accuracy is lower than that …

arxiv concept cs.ai cs.cv tricks type