h1

h2

h3

h4

h5
h6
http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png

Only Strict Saddles in the Energy Landscape of Predictive Coding Networks?

; ; ;

: arXiv

Umfang26 Seiten

Online
DOI: 10.48550/ARXIV.2408.11979

DOI: 10.18154/RWTH-2024-07912
URL: https://publications.rwth-aachen.de/record/991820/files/991820.pdf

Einrichtungen

  1. Lehrstuhl für Mathematik der Informationsverarbeitung (114510)
  2. Fachgruppe Mathematik (110000)

Projekte

  1. SFB 1481 A01 - Gradientenabstiegsmethoden für das Lernen tiefer neuronaler Netze (A01) (504263168) (504263168)
  2. DFG project 442047500 - SFB 1481: Sparsity und singuläre Strukturen (442047500) (442047500)

Inhaltliche Beschreibung (Schlagwörter)
Artificial Intelligence (cs.AI) (Genormte SW) ; FOS: Computer and information sciences (Genormte SW) ; I.2.6 (Genormte SW) ; Machine Learning (cs.LG) (Genormte SW) ; Machine Learning (stat.ML) (Genormte SW) ; Neural and Evolutionary Computing (cs.NE) (Genormte SW)

OpenAccess:
Download fulltext PDF
External link:
Download fulltextFulltext by arXiv.org

Dokumenttyp
Preprint

Format
online

Sprache
English

Externe Identnummern
arXiv: arXiv:2408.11979

Interne Identnummern
RWTH-2024-07912
Datensatz-ID: 991820

Beteiligte Länder
Germany, UK

 GO


Creative Commons Attribution CC BY 4.0 ; OpenAccess

QR Code for this record

The record appears in these collections:
Document types > Other document types > Preprints
Faculty of Mathematics and Natural Sciences (Fac.1) > Department of Mathematics
Publication server / Open Access
Public records
Publications database
114510
110000

 Record created 2024-08-25, last modified 2025-10-01


OpenAccess:
Download fulltext PDF
External link:
Download fulltextFulltext by arXiv.org
Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)