search

keisen / tf-keras-vis

Neural network visualization toolkit for tf.keras
https://keisen.github.io/tf-keras-vis-docs/
MIT License
313 stars 45 forks source link

How to output the visualization intermediate layer #104

Open Jie-Lii opened 4 months ago

Jie-Lii commented 4 months ago 
    score = BinaryScore(label[i])
    replace2linear = ReplaceToLinear()
    saliency = Saliency(model_m, model_modifier=replace2linear, clone=True)
    saliency_map = saliency(score, mat, normalize_map=True)

I‘m using the Vanilla Saliency method in tf keras vis to visualize my model, but I can only obtain the visualization results of the input data. Is there a method in this package that can visualize intermediate outputs? Looking forward to your reply.