Automating Interpretability: Discovering and Testing Visual Concepts Learned by Neural Networks

Open in new window