Questions for the seminar Paper "Learning and Leveraging World Models in Visual Representation Learning"
-----------------------------------------------------------------------------------------------
Please send your answers to: jesslen@cs.uni-freiburg.de

1) Why is the presented method considered a world model? (~2 sentences)
2) Are Image World Models equivariant or invariant? Why? (~2 sentences)
3) What are the requirements to learn a good representation using world models ? (~3-4 sentences)