Dimensionality and dynamics for next-generation artificial neural networks

Ge Wang; Feng-Lei Fan

doi:10.1016/j.patter.2025.101231

Dimensionality and dynamics for next-generation artificial neural networks

Ge Wang^*
, Feng-Lei Fan^*

^*Corresponding author for this work

Department of Data Science

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

4 Citations (Scopus)

13 Downloads (CityUHK Scholars)

Abstract

The recent awarding of the Nobel Prize in Physics to Geoffrey E. Hinton and John J. Hopfield highlights their profound impact on artificial neural networks. In this perspective, we explore how their foundational insights can drive the advancement of next-generation artificial intelligence (AI) models. We propose expanding beyond conventional architectures by introducing dimensionality through intra-layer links and dynamics via feedback loops. Network height and additional dimensions, alongside traditional width and depth, enhance learning capabilities, while entangled loops across scales induce emergent behaviors akin to phase transitions in physics. We discuss how these principles extend beyond transformers, fostering a new paradigm of intelligence inspired by physics-driven models and biological cognition mechanisms. © 2025 The Author(s).

Original language	English
Article number	101231
Number of pages	6
Journal	Patterns
Volume	6
Issue number	8
Online published	22 Apr 2025
DOIs	https://doi.org/10.1016/j.patter.2025.101231
Publication status	Published - 8 Aug 2025

Research Keywords

AI
Artificial intelligence
artificial neural network
deep learning
dimensionality expansion
feedback loop
Transformer

Publisher's Copyright Statement

This full text is made available under CC-BY 4.0. https://creativecommons.org/licenses/by/4.0/

Access Output

10.1016/j.patter.2025.101231

280886502.pdfFinal Published version, 1.87 MBLicence: CC BY

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{de0a5aca186049308018cf28d608c3e0,

title = "Dimensionality and dynamics for next-generation artificial neural networks",

abstract = "The recent awarding of the Nobel Prize in Physics to Geoffrey E. Hinton and John J. Hopfield highlights their profound impact on artificial neural networks. In this perspective, we explore how their foundational insights can drive the advancement of next-generation artificial intelligence (AI) models. We propose expanding beyond conventional architectures by introducing dimensionality through intra-layer links and dynamics via feedback loops. Network height and additional dimensions, alongside traditional width and depth, enhance learning capabilities, while entangled loops across scales induce emergent behaviors akin to phase transitions in physics. We discuss how these principles extend beyond transformers, fostering a new paradigm of intelligence inspired by physics-driven models and biological cognition mechanisms. {\textcopyright} 2025 The Author(s).",

keywords = "AI, Artificial intelligence, artificial neural network, deep learning, dimensionality expansion, feedback loop, Transformer",

author = "Ge Wang and Feng-Lei Fan",

year = "2025",

month = aug,

day = "8",

doi = "10.1016/j.patter.2025.101231",

language = "English",

volume = "6",

journal = "Patterns",

issn = "2666-3899",

publisher = "Cell Press",

number = "8",

}

TY - JOUR

T1 - Dimensionality and dynamics for next-generation artificial neural networks

AU - Wang, Ge

AU - Fan, Feng-Lei

PY - 2025/8/8

Y1 - 2025/8/8

N2 - The recent awarding of the Nobel Prize in Physics to Geoffrey E. Hinton and John J. Hopfield highlights their profound impact on artificial neural networks. In this perspective, we explore how their foundational insights can drive the advancement of next-generation artificial intelligence (AI) models. We propose expanding beyond conventional architectures by introducing dimensionality through intra-layer links and dynamics via feedback loops. Network height and additional dimensions, alongside traditional width and depth, enhance learning capabilities, while entangled loops across scales induce emergent behaviors akin to phase transitions in physics. We discuss how these principles extend beyond transformers, fostering a new paradigm of intelligence inspired by physics-driven models and biological cognition mechanisms. © 2025 The Author(s).

AB - The recent awarding of the Nobel Prize in Physics to Geoffrey E. Hinton and John J. Hopfield highlights their profound impact on artificial neural networks. In this perspective, we explore how their foundational insights can drive the advancement of next-generation artificial intelligence (AI) models. We propose expanding beyond conventional architectures by introducing dimensionality through intra-layer links and dynamics via feedback loops. Network height and additional dimensions, alongside traditional width and depth, enhance learning capabilities, while entangled loops across scales induce emergent behaviors akin to phase transitions in physics. We discuss how these principles extend beyond transformers, fostering a new paradigm of intelligence inspired by physics-driven models and biological cognition mechanisms. © 2025 The Author(s).

KW - AI

KW - Artificial intelligence

KW - artificial neural network

KW - deep learning

KW - dimensionality expansion

KW - feedback loop

KW - Transformer

UR - https://www.scopus.com/pages/publications/105003175086

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-105003175086&origin=recordpage

U2 - 10.1016/j.patter.2025.101231

DO - 10.1016/j.patter.2025.101231

M3 - RGC 21 - Publication in refereed journal

C2 - 40843340

SN - 2666-3899

VL - 6

JO - Patterns

JF - Patterns

IS - 8

M1 - 101231

ER -

Dimensionality and dynamics for next-generation artificial neural networks

Abstract

Research Keywords

Publisher's Copyright Statement

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this