The authors of Not All Adversarial Examples Require a Complex Defense: Identifying Over-optimized Adversarial Examples with IQR-based Logit Thresholding have not publicly listed the code yet.

Request code directly from the authors: Ask Authors for Code

Get an expert to implement this paper: Request Implementation

(OR if you have code to share with the community, please add it here 😊🙏 )