0 个仓库
Evaluating the stability of model explanations under adversarial perturbations of high-importance pixels.
Distinct from Adversarial Robustness Testing: Specifically tests the stability of the explanation/heatmap, whereas the parent focuses on the model's output vulnerability.
No awesome GitHub repositories for security & cryptography · Explanation Robustness Testing yet. Submit a GitHub URL or browse the filters below.
暂无仓库列表 —— 成为第一个提交的人吧。