Pantypes: Diverse Representatives for Self-Explainable Models

arXiv:2403.09383v1 Announce Type: cross
Abstract: Prototypical self-explainable classifiers have emerged to meet the growing demand for interpretable AI systems. These classifiers are designed to incorporate high transparency in their decisions by basing inference on similarity with learned prototypical objects. While these models are designed with diversity in mind, the learned prototypes often do not sufficiently represent all aspects of the input distribution, particularly those in low density regions. Such lack of sufficient data representation, known as representation bias, has been …

abstract ai systems arxiv classifiers cs.lg decisions demand diverse diversity inference interpretable ai mind objects stat.ml systems transparency type