Whether to only highlight pixels with positive contributions,
negative or both. Defaults to POSITIVE.
Values:
POLARITY_UNSPECIFIED (0):
Default value. This is the same as POSITIVE.
POSITIVE (1):
Highlights the pixels/outlines that were most
influential to the model's prediction.
NEGATIVE (2):
Setting polarity to negative highlights areas
that does not lead to the models's current
prediction.
BOTH (3):
Shows both positive and negative
attributions.