The original paper does not define a set threshold for the binary mask creation over neuron activation vectors, however, in our implementation is implicitly set using average_precision_score; we should clarify this in the docs and/or allow the user to set the threshold:
|
scores.append(average_precision_score(y_train, X_train[:, i])) |
The original paper does not define a set threshold for the binary mask creation over neuron activation vectors, however, in our implementation is implicitly set using
average_precision_score; we should clarify this in the docs and/or allow the user to set the threshold:NeuroX/neurox/interpretation/iou_probe.py
Line 37 in f2314cc