Testing the Generalization Power of Neural Network Models Across NLI Benchmarks

Show full item record



Permalink

http://hdl.handle.net/10138/304485

Citation

Talman , A J & Chatzikyriakidis , S 2019 , Testing the Generalization Power of Neural Network Models Across NLI Benchmarks . in T Linzen , G Chrupała , Y Belinkov & D Hupkes (eds) , The Workshop BlackboxNLP on Analyzing and Interpreting Neural Networks for NLP at ACL 2019 : Proceedings of the Second Workshop . The Association for Computational Linguistics , Stroudsburg , pp. 85-94 , 2019 ACL Workshop BlackboxNLP , Florence , Italy , 01/08/2019 .

Title: Testing the Generalization Power of Neural Network Models Across NLI Benchmarks
Author: Talman, Aarne Johannes; Chatzikyriakidis, Stergios
Other contributor: University of Helsinki, Department of Digital Humanities
Linzen, Tal
Chrupała, Grzegorz
Belinkov, Yonatan
Hupkes, Dieuwke

Publisher: The Association for Computational Linguistics
Date: 2019-08-01
Language: eng
Number of pages: 10
Belongs to series: The Workshop BlackboxNLP on Analyzing and Interpreting Neural Networks for NLP at ACL 2019 Proceedings of the Second Workshop
ISBN: 978-1-950737-30-7
URI: http://hdl.handle.net/10138/304485
Abstract: Neural network models have been very successful in natural language inference, with the best models reaching 90% accuracy in some benchmarks. However, the success of these models turns out to be largely benchmark specific. We show that models trained on a natural language inference dataset drawn from one benchmark fail to perform well in others, even if the notion of inference assumed in these benchmarks is the same or similar. We train six high performing neural network models on different datasets and show that each one of these has problems of generalizing when we replace the original test set with a test set taken from another corpus designed for the same task. In light of these results, we argue that most of the current neural network models are not able to generalize well in the task of natural language inference. We find that using large pre-trained language models helps with transfer learning when the datasets are similar enough. Our results also highlight that the current NLI datasets do not cover the different nuances of inference extensively enough.
Subject: 113 Computer and information sciences
6121 Languages
Rights:


Files in this item

Total number of downloads: Loading...

Files Size Format View
W19_4810.pdf 335.9Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record