This could be an interesting new paper by Mikhail Belkin and his team. This is a more theoretical work.
From the abstract:
"Understanding how neural networks learn and organize features is central to understanding their behavior. Much existing theory of feature learning has focused on the emergence of a global low-dimensional predictive geometry.
We show that this picture is incomplete.
In regression problems with clustered data, we demonstrate that multilayer perceptrons (MLPs) naturally develop monosemantic specialized neurons: individual neurons become strongly aligned with a specific predictive feature relevant to a particular region of the input space.
Rather than learning a single global low-dimensional representation, MLPs learn a collection of local low-dimensional representations that can collectively span a high-dimensional space.
This specialization provably gives MLPs a data-efficiency advantage over feature-learning methods based on a global low-dimensional representation."
"... Our main contributions are as follows.
We show that in trained MLPs, a substantial fraction of individual neurons
become monosemantic, specializing by aligning predominantly with a single
cluster-specific predictive direction.
This specialization allows the MLP to learn both the relevant local low-dimensional features and an implicit clustering that determines where each feature is useful.
The resulting behavior resembles mixture-of-experts routing, but emerges within a standard MLP without an explicit routing module or expert decomposition. ..."
No comments:
Post a Comment