Year: 2017
Type: conference-paper
Abstract: While neural machine translation (NMT) has made remarkable progress in recent years, it is hard to interpret its internal workings due to the continuous representations and non-linearity of neural networks.In this work, we propose to use layer-wise relevance propagation (LRP) to compute the contribution of each contextual word to arbitrary hidden states in the attention-based encoderdecoder framework.We show that visualization with LRP helps to interpret the internal workings of NMT and analyze translation errors.
DOI: 10.18653/v1/p17-1106
Authors: Yanzhuo Ding, Yang Liu, Huanbo Luan, Maosong Sun
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2741040846
