1 paper
Subeen Park, Joowang Kim, Hakyung Lee +2
Deep learning models achieve strong performance across various domains but often rely on spurious correlations, making them vulnerable to distribution shifts. This issue is particu…