New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

explainability tools for representation #157

Open

mattersoflight opened this issue Sep 9, 2024 · 1 comment

Assignees

Labels

enhancement representation

Milestone

Member

mattersoflight commented Sep 9, 2024 •

edited

Loading

We should extend the explainability tools developed for robust virtual staining to representation learning. I suggest the following:

Export a movie of features (input, PC of features) for a chosen track.
Effective receptive field of a chosen embedding dimension.

mattersoflight assigned ziw-liu

mattersoflight added this to the v0.3.0 milestone

ziw-liu added enhancement representation labels

Collaborator

ziw-liu commented Sep 12, 2024

Effective receptive field of a chosen embedding dimension.

The embedding is the global average pooling of the deepest feature map, so by definition its ERF is the entire input image.

ziw-liu modified the milestones: v0.3.0, v0.4.0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment