- Save and put your features in a Tensor [idx_layer, idx_tokens (t), feature] first.
- Run calculation_MI.py and plot_MI.ipynb respectively then.
In CV tasks, Mid Layer is recommended due to its balance between low-level and high-level pattern.
-
Notifications
You must be signed in to change notification settings - Fork 0
Calculate the mutual information between inputs and outputs of SSM and Attention, using MINE.
License
BDeMo/MINE_SSM_Attention
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
Calculate the mutual information between inputs and outputs of SSM and Attention, using MINE.
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published