_{^{Image credit: John D. Cook}}

Interpretability and Deep Learning

- ErUM train the trainers workshop 2022 -

The slides of the lecture can be found here: https://indico.scc.kit.edu/event/2645/contributions/9862/

Exercise I: Activation Maximization

Part II: Prediction Analysis

Disclaimer

The examples are inspired from Deep Learning for Physics Research. Find the full exercise page at: http://deeplearningphysics.org

Solutions for the tasks can be found in the respective "solution" branch

Software requirements:

Tested for tf.keras 2.8.0