Shortcut-Stacked Sentence Encoders for Multi-Domain Inference

Yixin Nie, Mohit Bansal


Abstract
We present a simple sequential sentence encoder for multi-domain natural language inference. Our encoder is based on stacked bidirectional LSTM-RNNs with shortcut connections and fine-tuning of word embeddings. The overall supervised model uses the above encoder to encode two input sentences into two vectors, and then uses a classifier over the vector combination to label the relationship between these two sentences as that of entailment, contradiction, or neural. Our Shortcut-Stacked sentence encoders achieve strong improvements over existing encoders on matched and mismatched multi-domain natural language inference (top single-model result in the EMNLP RepEval 2017 Shared Task (Nangia et al., 2017)). Moreover, they achieve the new state-of-the-art encoding result on the original SNLI dataset (Bowman et al., 2015).
Anthology ID:
W17-5308
Volume:
Proceedings of the 2nd Workshop on Evaluating Vector Space Representations for NLP
Month:
September
Year:
2017
Address:
Copenhagen, Denmark
Editors:
Samuel Bowman, Yoav Goldberg, Felix Hill, Angeliki Lazaridou, Omer Levy, Roi Reichart, Anders Søgaard
Venue:
RepEval
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
41–45
Language:
URL:
https://aclanthology.org/W17-5308
DOI:
10.18653/v1/W17-5308
Bibkey:
Cite (ACL):
Yixin Nie and Mohit Bansal. 2017. Shortcut-Stacked Sentence Encoders for Multi-Domain Inference. In Proceedings of the 2nd Workshop on Evaluating Vector Space Representations for NLP, pages 41–45, Copenhagen, Denmark. Association for Computational Linguistics.
Cite (Informal):
Shortcut-Stacked Sentence Encoders for Multi-Domain Inference (Nie & Bansal, RepEval 2017)
Copy Citation:
PDF:
https://aclanthology.org/W17-5308.pdf
Code
 easonnie/multiNLI_encoder +  additional community code
Data
MultiNLISNLI