Computer Science > Neural and Evolutionary Computing
arXiv:1410.5401 (cs)
[Submitted on 20 Oct 2014 (v1), last revised 10 Dec 2014 (this version, v2)]
Neural Turing Machines
Alex Graves, Greg Wayne, Ivo Danihelka
Download PDF
We extend the capabilities of neural networks by coupling them to external memory resources, which they can interact with by attentional processes. The combined system is analogous to a Turing Machine or Von Neumann architecture but is differentiable end-to-end, allowing it to be efficiently trained with gradient descent. Preliminary results demonstrate that Neural Turing Machines can infer simple algorithms such as copying, sorting, and associative recall from input and output examples.
Subjects:Neural and Evolutionary Computing (cs.NE)
Cite as:arXiv:1410.5401 [cs.NE]
 (or arXiv:1410.5401v2 [cs.NE] for this version)
Submission history
From: Alex Graves [view email]
[v1] Mon, 20 Oct 2014 19:28:26 UTC (763 KB)
[v2] Wed, 10 Dec 2014 16:01:39 UTC (764 KB)
PDF Other formats
Current browse context: cs.NE
< prev
next >
new recent 1410
Change to browse by: cs
References & Citations
Google Scholar
Semantic Scholar
15 blog links (what is this?)
DBLP - CS Bibliography
listing | bibtex
Alex Graves
Greg Wayne
Ivo Danihelka
Export Bibtex Citation
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Related Papers
About arXivLabs
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
Privacy Policy
Web Accessibility Assistance
arXiv Operational Status
Get status notifications via email or slack
LoginHelp PagesAbout