Interpretable Deep Learning under Fire

Xinyang Zhang, Ningfei Wang, Hua Shen +3

Providing explanations for deep neural network (DNN) models is crucial for their use in security-sensitive domains. A plethora of interpretation models have been proposed to help u…