Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > stat > arXiv:1810.03023

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Statistics > Machine Learning

arXiv:1810.03023 (stat)
[Submitted on 6 Oct 2018 (v1), last revised 9 Jan 2019 (this version, v2)]

Title:h-detach: Modifying the LSTM Gradient Towards Better Optimization

Authors:Devansh Arpit, Bhargav Kanuparthi, Giancarlo Kerg, Nan Rosemary Ke, Ioannis Mitliagkas, Yoshua Bengio
View a PDF of the paper titled h-detach: Modifying the LSTM Gradient Towards Better Optimization, by Devansh Arpit and 5 other authors
View PDF
Abstract:Recurrent neural networks are known for their notorious exploding and vanishing gradient problem (EVGP). This problem becomes more evident in tasks where the information needed to correctly solve them exist over long time scales, because EVGP prevents important gradient components from being back-propagated adequately over a large number of steps. We introduce a simple stochastic algorithm (\textit{h}-detach) that is specific to LSTM optimization and targeted towards addressing this problem. Specifically, we show that when the LSTM weights are large, the gradient components through the linear path (cell state) in the LSTM computational graph get suppressed. Based on the hypothesis that these components carry information about long term dependencies (which we show empirically), their suppression can prevent LSTMs from capturing them. Our algorithm\footnote{Our code is available at this https URL.} prevents gradients flowing through this path from getting suppressed, thus allowing the LSTM to capture such dependencies better. We show significant improvements over vanilla LSTM gradient based training in terms of convergence speed, robustness to seed and learning rate, and generalization using our modification of LSTM gradient on various benchmark datasets.
Comments: First two authors contributed equally. Published in ICLR 2019
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Cite as: arXiv:1810.03023 [stat.ML]
  (or arXiv:1810.03023v2 [stat.ML] for this version)
  https://doi.org/10.48550/arXiv.1810.03023
arXiv-issued DOI via DataCite

Submission history

From: Devansh Arpit [view email]
[v1] Sat, 6 Oct 2018 16:55:46 UTC (1,025 KB)
[v2] Wed, 9 Jan 2019 17:12:59 UTC (1,053 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled h-detach: Modifying the LSTM Gradient Towards Better Optimization, by Devansh Arpit and 5 other authors
  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
stat.ML
< prev   |   next >
new | recent | 2018-10
Change to browse by:
cs
cs.LG
stat

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
a export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

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?)
  • Author
  • Venue
  • Institution
  • Topic

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?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack