Questions for "How Well do Feature Visualizations Support Causal Understanding of CNN Activations?"
-----------------------------------------------------------------------------------------------
Please send your answers to: schroepp@cs.uni-freiburg.de by 13:15 on 17.06.2022

1. How are the feature visualizations in Figure 1 A generated? (2-3 sentences, explain in your own words)

2. Which question does the paper try to answer? Explain the experimental setup. (3 sentences)

3. Explain the main findings of the paper based on Figure 3. (2-3 sentences)