← Search

Roland Simon Zimmermann

2 accepted papers

2021

Exemplary Natural Images Explain CNN Activations Better than State-of-the-Art Feature Visualization

ICLR 2021poster

Feature visualizations such as synthetic maximally activating images are a widely used explanation method to better understand the information processing of convolutional neural networks (CNNs). At the same time, there are concerns that these visualizations might not accurately represent CNNs' inner…

2021

How Well do Feature Visualizations Support Causal Understanding of CNN Activations?

NeurIPS 2021spotlight

A precise understanding of why units in an artificial network respond to certain stimuli would constitute a big step towards explainable artificial intelligence. One widely used approach towards this goal is to visualize unit responses via activation maximization. These feature visualizations are pu…