Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arXiv logo
Cornell University Logo

Computer Science > Machine Learning

arXiv:2305.18475 (cs)
[Submitted on 29 May 2023 (v1), last revised 24 Oct 2024 (this version, v3)]

Title:Approximation Rate of the Transformer Architecture for Sequence Modeling

Authors:Haotian Jiang, Qianxiao Li
View PDF
Abstract:The Transformer architecture is widely applied in sequence modeling applications, yet the theoretical understanding of its working principles remains limited. In this work, we investigate the approximation rate for single-layer Transformers with one head. We consider a class of non-linear relationships and identify a novel notion of complexity measures to establish an explicit Jackson-type approximation rate estimate for the Transformer. This rate reveals the structural properties of the Transformer and suggests the types of sequential relationships it is best suited for approximating. In particular, the results on approximation rates enable us to concretely analyze the differences between the Transformer and classical sequence modeling methods, such as recurrent neural networks.
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2305.18475 [cs.LG]
  (or arXiv:2305.18475v3 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2305.18475

Submission history

From: Haotian Jiang [view email]
[v1] Mon, 29 May 2023 10:56:36 UTC (2,075 KB)
[v2] Mon, 19 Feb 2024 03:38:19 UTC (4,529 KB)
[v3] Thu, 24 Oct 2024 08:13:01 UTC (4,515 KB)
Full-text links:

Access Paper:

  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs
< prev   |   next >
new | recent | 2023-05
Change to browse by:
cs.LG

References & Citations

export BibTeX citation

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
IArxiv Recommender (What is IArxiv?)

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)