1 paper
Sidney Bender, Benedikt Kunz, Ahmed Zeid +3
Foundation models remain vulnerable to spurious correlations and ``Clever Hans'' strategies. Explainable machine learning can find and remove such strategies for classifiers withou…