search

PAIR-code / saliency

Framework-agnostic implementation for state-of-the-art saliency methods (XRAI, BlurIG, SmoothGrad, and more).
https://pair-code.github.io/saliency/
Apache License 2.0
953 stars 191 forks source link

In Examples_pytorch.ipynb file unable to find XRAI mask for other classes by replacing call_model_args #90

Open arunsubk opened 1 year ago

arunsubk commented 1 year ago

Hi,

Firstly thanks for this great repo.

In the Examples_pytorch.ipynb file, If i change call_model_args = {class_idx_str: prediction_class} to call_model_args = {class_idx_str: <ANY_OTHER_CLASS>} XRAI mask still remains that of Doberman (i.e. class 236). What do I do to see the visualization of any other class in the Doberman image? I had for instance set ANY_OTHER_CLASS to 562 (fountain) and was hoping to see difference gradient image of the class. But I still see the Doberman.

Could you please help? Thanks,