Neuronale Netzwerkarchitektur

Details

Veröffentlichungs-Nr.

EP4604014

Aktenzeichen

EP24158204

Anmeldetag

16. Februar 2024

Veröffentlichung

20. August 2025

Rechtsraum

EP

IPC

G06N3/045G06N3/049G06N3/065G06N3/084

Abstract

Some examples of for training a neural network, wherein the neural network comprises a plurality of interconnected neurons, comprise receiving input data at a first neural network unit of the neural network and processing the input data using the first neural network unit to produce an output, a neural network unit comprising one or more of the neurons. Assign a first frequency to the first neural network unit and assign a second frequency associated to the second neural network unit. A difference value between the first frequency and the second frequency is determined and is used as an attention mechanism by, scaling the output of the first neural network unit in relation to the determined difference.

Anmelder

Firma

British Telecommunications public limited company

Land

🇬🇧 Großbritannien

Fachgebiete