Questions for "Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision"
-----------------------------------------------------------------------------------------------

Please send your answers to: schrodi@cs.uni-freiburg.de by 13:30 on 28.06.2023

    Question 1. Briefly describe their self-supervised learning approach. (~2-3 sentences)
    Question 2. How do they scale their dense model and `make it trainable`? (~3 sentences)
    Question 3. What types of fairness indicators do they consider? (~2-3 sentences)