7 papers · 1 filter
Bernstein-Schur Kernels: Random Features by Sketched Modulation and Radial Randomization
Taha Bouhsine
Bernstein--Schur kernels are products of a finite-feature kernel and a completely monotone shift-invariant kernel: nonstationary kernels falling between the shift-invariant and dot…
A Universal Reproducing Kernel Hilbert Space from Polynomial Alignment and IMQ Distance
Taha Bouhsine
Inverse-multiquadric (IMQ) kernel sections decay at infinity. Can polynomial alignment add persistent directional responses while preserving universal approximation? The \yat{} (Ya…
No More DeLuLu: Physics-Inspired Kernel Networks for Geometrically-Grounded Neural Computation
Taha Bouhsine
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,…
SLAY: Geometry-Aware Spherical Linearized Attention with Yat-Kernel
Jose Miguel Luna, Taha Bouhsine, Krzysztof Choromanski
We propose a new class of linear-time attention mechanisms based on a relaxed and computationally efficient formulation of the recently introduced E-Product, often referred to as t…
In Defense of Cosine Similarity: Normalization Eliminates the Gauge Freedom
Taha Bouhsine
Steck, Ekanadham, and Kallus [arXiv:2403.05440] demonstrate that cosine similarity of learned embeddings from matrix factorization models can be rendered arbitrary by a diagonal ``…
Deep Learning 2.0: Artificial Neurons That Matter -- Reject Correlation, Embrace Orthogonality
Taha Bouhsine
We introduce a yat-product-powered neural network, the Neural Matter Network (NMN), a breakthrough in deep learning that achieves non-linear pattern recognition without activation…
SimO Loss: Anchor-Free Contrastive Loss for Fine-Grained Supervised Contrastive Learning
Taha Bouhsine, Imad El Aaroussi, Atik Faysal +1
We introduce a novel anchor-free contrastive learning (AFCL) method leveraging our proposed Similarity-Orthogonality (SimO) loss. Our approach minimizes a semi-metric discriminativ…