CoChat

Learning to Forget: Continual Prediction with LSTM

Gers, Felix A., Schmidhuber, Jürgen, Cummins, Fred · 2000 · Neural Computation
Cited by 4,823

The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.

Abstract

Long short-term memory (LSTM; Hochreiter & Schmidhuber, 1997) can solve numerous tasks not solvable by previous learning algorithms for recurrent neural networks (RNNs). We identify a weakness of LSTM networks processing continual input streams that are not a priori segmented into subsequences with explicitly marked ends at which the network's internal state could be reset. Without resets, the state may grow indefinitely and eventually cause the network to break down. Our remedy is a novel, adaptive “forget gate” that enables an LSTM cell to learn to reset itself at appropriate times, thus releasing internal resources. We review illustrative benchmark problems on which standard LSTM outperforms other RNN algorithms. All algorithms (including LSTM) fail to solve continual versions of these problems. LSTM with forget gates, however, easily solves them, and in an elegant way.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1162/089976600300015015

Cites

2 of 2 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

  • Long Short-Term Memory

    Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 22×

    “…to Forget: Continual Prediction with LSTM Felix A. Gers Jürgen Schmidhuber Fred Cummins IDSIA, 6900 Lugano, Switzerland Long short-term memory (LSTM; Hochreiter & Schmidhuber, 1997) can solve numerous tasks not solvable by previous learning algorithms for recurrent neural networks (RNNs).”
    InfluentialSelf-citation
  • Learning long-term dependencies with gradient descent is difficult

    Bengio, Y., Simard, P., & Frasconi, P. (1994) · IEEE Transactions on Neural Networks · doi:10.1109/72.279181 · Cited by 8,542 · cited 1×

    “This implies that the backpropagated error quickly either vanishes or blows up (Hochreiter and Schmidhuber, 1997; Bengio et al., 1994).”
    Background

Research works printed but not linked: 12

  1. 10.1162/neco.1989.1.3.372 · DOI · Paper page
  2. 10.1016/0893-6080(89)90022-1 · DOI
  3. 10.1109/72.548162 · DOI · Paper page
  4. Mozer M. C., 1989, Complex Systems, 3, 349
  5. 10.1109/72.410363 · DOI · Paper page
  6. 10.1080/09540098908915650 · DOI · Paper page
  7. 10.1162/neco.1992.4.2.243 · DOI · Paper page
  8. 10.1162/089976699300016467 · DOI · Paper page
  9. 10.1142/S0129065789000037 · DOI
  10. 10.1162/neco.1989.1.1.39 · DOI · Paper page
  11. 10.1016/0893-6080(88)90007-X · DOI · Paper page
  12. 10.1162/neco.1990.2.4.490 · DOI · Paper page

Cited by

See all 4,823

Metadata

Authors
Gers, F. A., Schmidhuber, J., & Cummins, F.
Affiliations
IDSIA, 6900 Lugano, Switzerland
Venue
Neural Computation
Publisher
MIT Press
Published
2000-10-01
Volume
12
Issue
10
Pages
2451-2471
ISSN
  • 0899-7667
  • 1530-888X
Cited by
4,823
References
2
Language
en

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.