Scalable Interconnect Learning in Boolean Networks

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Kresse, Fabian, Yu, Emily, Lampert, Christoph H.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911160364695552
author Kresse, Fabian
Yu, Emily
Lampert, Christoph H.
author_facet Kresse, Fabian
Yu, Emily
Lampert, Christoph H.
contents Learned Differentiable Boolean Logic Networks (DBNs) already deliver efficient inference on resource-constrained hardware. We extend them with a trainable, differentiable interconnect whose parameter count remains constant as input width grows, allowing DBNs to scale to far wider layers than earlier learnable-interconnect designs while preserving their advantageous accuracy. To further reduce model size, we propose two complementary pruning stages: an SAT-based logic equivalence pass that removes redundant gates without affecting performance, and a similarity-based, data-driven pass that outperforms a magnitude-style greedy baseline and offers a superior compression-accuracy trade-off.
format Preprint
id arxiv_https___arxiv_org_abs_2507_02585
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Scalable Interconnect Learning in Boolean Networks
Kresse, Fabian
Yu, Emily
Lampert, Christoph H.
Machine Learning
Logic in Computer Science
Learned Differentiable Boolean Logic Networks (DBNs) already deliver efficient inference on resource-constrained hardware. We extend them with a trainable, differentiable interconnect whose parameter count remains constant as input width grows, allowing DBNs to scale to far wider layers than earlier learnable-interconnect designs while preserving their advantageous accuracy. To further reduce model size, we propose two complementary pruning stages: an SAT-based logic equivalence pass that removes redundant gates without affecting performance, and a similarity-based, data-driven pass that outperforms a magnitude-style greedy baseline and offers a superior compression-accuracy trade-off.
title Scalable Interconnect Learning in Boolean Networks
topic Machine Learning
Logic in Computer Science
url https://arxiv.org/abs/2507.02585