1 paper
Minhao Cheng, Simranjit Singh, Patrick Chen +3
We study the most practical problem setup for evaluating adversarial robustness of a machine learning system with limited access: the hard-label black-box attack setting for genera…