Neural Backdoor Detection: Activation Clustering Defense

This simulator renders every sample a poisoned classifier assigns to one output class as a point in a real 3D projection of the network's last hidden-layer activation space. A hidden trigger pushes a poisoned minority of that class into a tight, separated cluster, exactly as it does in a real backdoored network. Adjust the poison rate, trigger separation and clean-activation spread, then run the real k-means Activation Clustering defense (Chen et al., 2018) to flag the suspected cluster, or reveal the planted ground truth to check how the defense performed — with live precision, recall and a silhouette-style separation score.