What Does a Machine Learn? Experiments on Concept Representations of Human Coders and Machine Learning Models

Our experiments show how XAI methods can trace model decisions and how these relate to disagreements among human coders

Abstract

We will soon post the results of our XAI experiments here.