Vanishing Gradient Problem

The Vanishing Gradient Problem is a phenomenon in deep neural networks where the gradients of the loss function with respect to the weights become increasingly small during backpropagation, leading to negligible weight updates in earlier layers and hindering the training of deep architectures.

419

Publications

93.7K

Citations

1.4K

Authors

478

Institutions

Publications per year

2017–2026

365

Authors

1.4K

  1. YB
    Yoshua Bengio

    Université de Montréal, Canada

    8 publications · 4.7K citations

    8

    h-index

  2. GL
    Guoqi Li

    Tsinghua University, China

    5 publications · 593 citations

    5

    h-index

  3. LD
    Lei Deng

    Tsinghua University, China

    5 publications · 593 citations

    5

    h-index

  4. TM
    Tomáš Mikolov

    Meta (Israel), Israel

    5 publications · 4.6K citations

    5

    h-index

  5. KQ
    Kilian Q. Weinberger

    Cornell University, United States

    4 publications · 47.4K citations

    4

    h-index

1–5 of 1.4K

Venues

  1. arXiv (Cornell University)

    Repository · 92 publications · 16.1K citations

    38

    h-index

  2. IEEE Access

    25 publications · 1.6K citations

    20

    h-index

  3. 12

    h-index

  4. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 11 publications · 926 citations

    10

    h-index

  5. Applied Sciences

    7 publications · 330 citations

    7

    h-index

1–5