Classifying visualized embeddings?

[daveramseymusic (Dave)]

I have about 500 cnn models trained for classification task.(in theory)

Half scored well on a test set. half scored poorly on the same test set.

If I visualized the final embeddings layer for each model.

Would it be reasonable to train a resnet classifier to predict “good model/bad model” just by looking at the final embeddings layer of each model?

I’m considering starting this project.
Does anyone have any experience, thoughts, or resources that might help?

Any help would make my day. Cheers!

[daveramseymusic (Dave)]

ps. here is where i’ve looked so far for how to visualize the embeddings:

Links from tonights discussion

Send them to Amazon Turk

Fastai v2 vision

Wiki topic to discuss fastai v2 vision module.

Visualizing intermediate layers a la Zeiler and Fergus

In lesson 1 Jeremy introduced the Zeiler and Fergus paper that visualizes intermediate layers to help us develop intuition for how the layers of CNNs progressively learn the building blocks needed for the classification task at hand. I was curious whether there’s a ready to use library for visualizing intermediate layers to help us beginners develop intuition for how CNNs work and when they fail to work with a set of images and classification task. There have been multiple questions in the foru…