Are the hidden spaces of MLPs at different layers totally different? They can't be totally different because there's not that much individual transformation capacity...but then can we compare what differs?
If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:
@misc{holtzman-how-similar-are-2026,
author = {Holtzman, Ari},
title = {How similar are MLP hidden states to each other?},
year = {2026},
url = {https://hypogenic.ai/ideahub/idea/9UQ2hnE43Uunpez7bfF5}
}Please sign in to comment on this idea.
No comments yet. Be the first to share your thoughts!