Saved in:
Bibliographic Details
Main Author: Bouhsine, Taha
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.12276
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911509439840256
author Bouhsine, Taha
author_facet Bouhsine, Taha
contents We introduce the yat-product, a kernel operator combining quadratic alignment with inverse-square proximity. We prove it is a Mercer kernel, analytic, Lipschitz on bounded domains, and self-regularizing, admitting a unique RKHS embedding. Neural Matter Networks (NMNs) use yat-product as the sole non-linearity, replacing conventional linear-activation-normalization blocks with a single geometrically-grounded operation. This architectural simplification preserves universal approximation while shifting normalization into the kernel itself via the denominator, rather than relying on separate normalization layers. Empirically, NMN-based classifiers match linear baselines on MNIST while exhibiting bounded prototype evolution and superposition robustness. In language modeling, Aether-GPT2 achieves lower validation loss than GPT-2 with a comparable parameter budget while using yat-based attention and MLP blocks. Our framework unifies kernel learning, gradient stability, and information geometry, establishing NMNs as a principled alternative to conventional neural architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2603_12276
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle No More DeLuLu: Physics-Inspired Kernel Networks for Geometrically-Grounded Neural Computation
Bouhsine, Taha
Machine Learning
We introduce the yat-product, a kernel operator combining quadratic alignment with inverse-square proximity. We prove it is a Mercer kernel, analytic, Lipschitz on bounded domains, and self-regularizing, admitting a unique RKHS embedding. Neural Matter Networks (NMNs) use yat-product as the sole non-linearity, replacing conventional linear-activation-normalization blocks with a single geometrically-grounded operation. This architectural simplification preserves universal approximation while shifting normalization into the kernel itself via the denominator, rather than relying on separate normalization layers. Empirically, NMN-based classifiers match linear baselines on MNIST while exhibiting bounded prototype evolution and superposition robustness. In language modeling, Aether-GPT2 achieves lower validation loss than GPT-2 with a comparable parameter budget while using yat-based attention and MLP blocks. Our framework unifies kernel learning, gradient stability, and information geometry, establishing NMNs as a principled alternative to conventional neural architectures.
title No More DeLuLu: Physics-Inspired Kernel Networks for Geometrically-Grounded Neural Computation
topic Machine Learning
url https://arxiv.org/abs/2603.12276