Improving Language Modeling using Densely Connected Recurrent Neural Networks

Fréderic Godin, Joni Dambre, Wesley De Neve


Abstract
In this paper, we introduce the novel concept of densely connected layers into recurrent neural networks. We evaluate our proposed architecture on the Penn Treebank language modeling task. We show that we can obtain similar perplexity scores with six times fewer parameters compared to a standard stacked 2-layer LSTM model trained with dropout (Zaremba et al., 2014). In contrast with the current usage of skip connections, we show that densely connecting only a few stacked layers with skip connections already yields significant perplexity reductions.
Anthology ID:
W17-2622
Volume:
Proceedings of the 2nd Workshop on Representation Learning for NLP
Month:
August
Year:
2017
Address:
Vancouver, Canada
Editors:
Phil Blunsom, Antoine Bordes, Kyunghyun Cho, Shay Cohen, Chris Dyer, Edward Grefenstette, Karl Moritz Hermann, Laura Rimell, Jason Weston, Scott Yih
Venue:
RepL4NLP
SIG:
SIGREP
Publisher:
Association for Computational Linguistics
Note:
Pages:
186–190
Language:
URL:
https://aclanthology.org/W17-2622
DOI:
10.18653/v1/W17-2622
Bibkey:
Cite (ACL):
Fréderic Godin, Joni Dambre, and Wesley De Neve. 2017. Improving Language Modeling using Densely Connected Recurrent Neural Networks. In Proceedings of the 2nd Workshop on Representation Learning for NLP, pages 186–190, Vancouver, Canada. Association for Computational Linguistics.
Cite (Informal):
Improving Language Modeling using Densely Connected Recurrent Neural Networks (Godin et al., RepL4NLP 2017)
Copy Citation:
PDF:
https://aclanthology.org/W17-2622.pdf