Convolutional Sequence to Sequence Learning

2017

Citations in the corpus

Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidates (2018, background): “However, note that subword regularization is not speciﬁc to this architecture and can be applicable to other NMT architectures without RNN, e.g., (Vaswani et al., 2017; Gehring et al., 2017).”
Beyond English-Centric Multilingual Machine Translation (2020, background): “Neural network models have been very successful for bilingual machine translation (Bahdanau et al., 2014; Gehring et al., 2017; Vaswani et al., 2017) and more recently, neural MMT models have shown promising results (Firat et al., 2016; Zhang et al., 2020).”
Why Self-Attention? A Targeted Evaluation of Neural Machine Translation Architectures (2018, methodology): “Since comparisons (Gehring et al., 2017; Vaswani et al., 2017; Hieber et al., 2017) are mainly carried out via BLEU (Papineni et al., 2002), it is inherently difﬁcult to attribute gains in BLEU to architectural properties.”
Importance of Search and Evaluation Strategies in Neural Dialogue Modeling (2018, background): “…by a neural network, and popular choices include recurrent neural networks (Mikolov et al., 2010; Sutskever et al., 2014; Cho et al., 2014; Bahdanau et al., 2014), convolutional networks (Dauphin et al., 2016; Gehring et al., 2017) and selfattention (Sukhbaatar et al., 2015; Vaswani et al., 2017).”
Sharing Attention Weights for Fast Transformer (2019, methodology): “More recently, multi-layer attention was successfully applied to convolutional neural MT systems [Gehring et al. , 2017] and Transformer systems [Vaswani et al. , 2017].”
The Missing Ingredient in Zero-Shot Neural Machine Translation (2019, methodology): “Our experiments use the standard WMT14 en→fr (39M) and en→de (4.5M) training datasets that are used to benchmark state-of-the-art NMT systems (Vaswani et al., 2017; Gehring et al., 2017; Chen et al., 2018).”
On the Relation between Position Information and Sentence Length in Neural Machine Translation (2019, methodology): “This is in contrast to the position embeddings (Gehring et al., 2017), a learned position vectors, which are not meant to attend to relative positions.”
Transcoding Compositionally: Using Attention to Find More Generalizable Solutions (2019, methodology): “In future work, we plan to install it into other popular seq2seq architectures such as convolutional seq2seq (Gehring et al., 2017) and Transformer models (Vaswani et al., 2017).”
Deep Learning on Traffic Prediction: Methods, Analysis, and Future Directions (2021, background): “CNN. [68] ﬁrst introduced the fully convolutional model for sequence to sequence learning.”
An Analysis of Attention Mechanisms: The Case of Word Sense Disambiguation in Neural Machine Translation (2018, background): “Therefore, Gehring et al. (2017) and Vaswani et al. (2017) feed a context vector into each decoder layer.”
Double Embeddings and CNN-based Sequence Labeling for Aspect Extraction (2018, methodology): “Although most existing models use LSTM (Hochre-iter and Schmidhuber, 1997) as the core building block to model sequences (Liu et al., 2015; Li and Lam, 2017), we noticed that CNN is also successful in many NLP tasks (Kim, 2014; Zhang et al., 2015; Gehring et al., 2017).”
Global Autoregressive Models for Data-Efficient Sequence Learning (2019, unknown): “These are currently the standard for neural seq2seq processing, with such representatives as RNN/LSTMs (Hochreiter and Schmidhuber, 1997; Sutskever et al., 2014), ConvS2S (Gehring et al., 2017), Transformer (Vaswani et al., 2017)).”
Controllable Neural Story Plot Generation via Reward Shaping (2018, unknown): “Neural language modeling-based story generation approaches in particular [Roemmele and Gordon, 2015; Khalifa et al. , 2017; Gehring et al. , 2017; Martin et al. , 2018 ] are prone to generating stories with little aim since each sentence, event, word, ∗ Denotes equal contribution. or letter is…”
Syntax-Aware Data Augmentation for Neural Machine Translation (2020, methodology): “A series of seq2seq NMT model, such as RNN model (Sutskever et al., 2014; Bah-danau et al., 2014), CNN (Gehring et al., 2017) model and the Transformer (Vaswani et al., 2017), have been proposed and received better and better performance.”
Beyond BLEU:Training Neural Machine Translation with Semantic Similarity (2019, methodology): “We use gated convolutional encoders and decoders (Gehring et al., 2017).”
Improving Domain Adaptation Translation with Domain Invariant and Specific Information (2019, background): “Neural machine translation (NMT) (Kalchbrenner and Blunsom, 2013; Cho et al., 2014; Sutskever et al., 2014; Bahdanau et al., 2014; Gehring et al., 2017) has made great progress and drawn much attention recently.”