The reviewed record of science sign in
Pith

arxiv: 1909.00102 · v1 · pith:XNFWBHZH · submitted 2019-08-31 · cs.CL · cs.CR· cs.LG· stat.ML

Knowledge Enhanced Attention for Robust Natural Language Inference

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:XNFWBHZHrecord.jsonopen to challenge →

classification cs.CL cs.CRcs.LGstat.ML
keywords modelsattentionadversarialinferenceknowledgelanguagemethodnatural
0
0 comments X
read the original abstract

Neural network models have been very successful at achieving high accuracy on natural language inference (NLI) tasks. However, as demonstrated in recent literature, when tested on some simple adversarial examples, most of the models suffer a significant drop in performance. This raises the concern about the robustness of NLI models. In this paper, we propose to make NLI models robust by incorporating external knowledge to the attention mechanism using a simple transformation. We apply the new attention to two popular types of NLI models: one is Transformer encoder, and the other is a decomposable model, and show that our method can significantly improve their robustness. Moreover, when combined with BERT pretraining, our method achieves the human-level performance on the adversarial SNLI data set.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.