Lower bound on the information-loss incurred by compressing a depth-2 feed-forward layer of transformer into a single fully connected layer.
Saved in:
| Title: | Lower bound on the information-loss incurred by compressing a depth-2 feed-forward layer of transformer into a single fully connected layer. |
|---|---|
| Authors: | Matzinger, Heinrich1, matzi@math.gatech.edu, Komatuzaki, Aran1 |
| Source: | Journal of Algorithms & Computational Technology; 6/9/2026, Vol. 20, p1-18, 18p |
| Database: | Applied Science & Technology Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
Be the first to leave a comment!
