1 paper
Jishnu Mukhoti, Yarin Gal, Philip H. S. Torr +1
Pre-trained foundation models, due to their enormous capacity and exposure to vast amounts of data during pre-training, are known to have learned plenty of real-world concepts. An…