On Multiplicative Integration with Recurrent Neural Networks
Abstract
We introduce a general simple structural design called “Multiplicative Integration” (MI) to improve recurrent neural networks (RNNs). MI changes the way of how the information flow gets integrated in the computational building block of an RNN, while introducing almost no extra parameters. The new structure can be easily embedded into many popular RNN models, including LSTMs and GRUs. We empirically analyze its learning behaviour and conduct evaluations on several tasks using different RNN models. Our experimental results demonstrate that Multiplicative Integration can provide a substantial performance boost over many of the existing RNN models.
Cite
Text
Wu et al. "On Multiplicative Integration with Recurrent Neural Networks." Neural Information Processing Systems, 2016.Markdown
[Wu et al. "On Multiplicative Integration with Recurrent Neural Networks." Neural Information Processing Systems, 2016.](https://mlanthology.org/neurips/2016/wu2016neurips-multiplicative/)BibTeX
@inproceedings{wu2016neurips-multiplicative,
title = {{On Multiplicative Integration with Recurrent Neural Networks}},
author = {Wu, Yuhuai and Zhang, Saizheng and Zhang, Ying and Bengio, Yoshua and Salakhutdinov, Ruslan},
booktitle = {Neural Information Processing Systems},
year = {2016},
pages = {2856-2864},
url = {https://mlanthology.org/neurips/2016/wu2016neurips-multiplicative/}
}