Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation

Abstract

Neural Machine Translation (NMT) is an end-to-end learning approach for automated translation, with the potential to overcome many of the weaknesses of conventional phrase-based translation systems. Unfortunately, NMT systems are known to be computationally expensive both in training and in translation inference. Also, most NMT systems have difficulty with rare words. These issues have hindered NMT's use in practical deployments and services, where both accuracy and speed are essential. In this work, we present GNMT, Google's Neural Machine Translation system, which attempts to address many of these issues. Our model consists of a deep LSTM network with 8 encoder and 8 decoder layers using attention and residual connections. To improve parallelism and therefore decrease training time, our attention mechanism connects the bottom layer of the decoder to the top layer of the encoder. To accelerate the final translation speed, we employ low-precision arithmetic during inference computations. To improve handling of rare words, we divide words into a limited set of common sub-word units ("wordpieces") for both input and output. This method provides a good balance between the flexibility of "character"-delimited models and the efficiency of "word"-delimited models, naturally handles translation of rare words, and ultimately improves the overall accuracy of the system. Our beam search technique employs a length-normalization procedure and uses a coverage penalty, which encourages generation of an output sentence that is most likely to cover all the words in the source sentence. On the WMT'14 English-to-French and English-to-German benchmarks, GNMT achieves competitive results to state-of-the-art. Using a human side-by-side evaluation on a set of isolated simple sentences, it reduces translation errors by an average of 60% compared to Google's phrase-based production system.

Cited by in corpus (1337)

• BioBERT: a pre-trained biomedical language representation model for biomedical text mining
• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• Axiomatic Attribution for Deep Networks
• Convolutional Sequence to Sequence Learning
• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
• BERTScore: Evaluating Text Generation with BERT
• Photonics for artificial intelligence and neuromorphic computing
• ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks
• A Deep Reinforced Model for Abstractive Summarization
• Recent Trends in Deep Learning Based Natural Language Processing
• VisualBERT: A Simple and Performant Baseline for Vision and Language
• fastai: A Layered API for Deep Learning
• Mixed Precision Training
• PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization
• Unified Language Model Pre-training for Natural Language Understanding and Generation
• Deep Audio-Visual Speech Recognition
• Data Augmentation Generative Adversarial Networks
• CTRL: A Conditional Transformer Language Model for Controllable Generation
• VL-BERT: Pre-training of Generic Visual-Linguistic Representations
• ERNIE: Enhanced Representation through Knowledge Integration
• StarCraft II: A New Challenge for Reinforcement Learning
• Edge Learning for B5G Networks with Distributed Signal Processing: Semantic Communication, Edge Computing, and Wireless Sensing
• MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
• The ApolloScape Open Dataset for Autonomous Driving and its Application
• ET-BERT: A Contextualized Datagram Representation with Pre-training Transformers for Encrypted Traffic Classification
• Deep neural networks for the evaluation and design of photonic devices
• Deep Reinforcement Learning: An Overview
• Large-Scale Evolution of Image Classifiers
• Non-Autoregressive Neural Machine Translation
• Multimodal Intelligence: Representation Learning, Information Fusion, and Applications
• Graph Neural Networks with convolutional ARMA filters
• A Survey on Data Augmentation for Text Classification
• QANet: Combining Local Convolution with Global Self-Attention for Reading Comprehension
• Leveraging Pre-trained Checkpoints for Sequence Generation Tasks
• A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems
• Long short-term memory networks in memristor crossbars
• Train longer, generalize better: closing the generalization gap in large batch training of neural networks
• Synthetic and Natural Noise Both Break Neural Machine Translation
• Fine-Tuning Language Models from Human Preferences
• Aggregated Residual Transformations for Deep Neural Networks
• GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding
• Learning to Generate Reviews and Discovering Sentiment
• The Natural Language Decathlon: Multitask Learning as Question Answering
• Quasi-Recurrent Neural Networks
• Reconstructing quantum states with generative models
• Compute Trends Across Three Eras of Machine Learning
• A Simple Method for Commonsense Reasoning
• Neural Rating Regression with Abstractive Tips Generation for Recommendation
• Massively Multilingual Neural Machine Translation in the Wild: Findings and Challenges
• VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts
• Pixel-BERT: Aligning Image Pixels with Text by Deep Multi-Modal Transformers
• Improved training of end-to-end attention models for speech recognition
• Extremely Large Minibatch SGD: Training ResNet-50 on ImageNet in 15 Minutes
• What do Neural Machine Translation Models Learn about Morphology?
• A Transformer-based approach to Irony and Sarcasm detection
• Training Tips for the Transformer Model
• Adversarial Attacks on Deep Learning Models in Natural Language Processing: A Survey
• GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism
• CodeBERT: A Pre-Trained Model for Programming and Natural Languages
• Depthwise Separable Convolutions for Neural Machine Translation
• Learning to Remember Rare Events
• On the importance of single directions for generalization
• Benchmarking TPU, GPU, and CPU Platforms for Deep Learning
• Hybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation
• UniLMv2: Pseudo-Masked Language Models for Unified Language Model Pre-Training
• LXMERT: Learning Cross-Modality Encoder Representations from Transformers
• Device Placement Optimization with Reinforcement Learning
• An Actor-Critic Algorithm for Sequence Prediction
• Stand-Alone Self-Attention in Vision Models
• Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation
• Cognitive Science in the era of Artificial Intelligence: A roadmap for reverse-engineering the infant language-learner
• Vision-Language Models for Medical Report Generation and Visual Question Answering: A Review
• A Connection between Generative Adversarial Networks, Inverse Reinforcement Learning, and Energy-Based Models
• Multi-modal Understanding and Generation for Medical Images and Text via Vision-Language Pre-Training
• A Comprehensive Survey of Neural Architecture Search: Challenges and Solutions
• An Efficient Graph Convolutional Network Technique for the Travelling Salesman Problem
• A Deep Reinforcement Learning Chatbot
• Autoregressive Entity Retrieval
• Natural Language Processing Advancements By Deep Learning: A Survey
• Scalable Methods for 8-bit Training of Neural Networks
• A Survey of Domain Adaptation for Neural Machine Translation
• Unsupervised Machine Translation Using Monolingual Corpora Only
• ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT
• UNITER: UNiversal Image-TExt Representation Learning
• Portuguese Named Entity Recognition using BERT-CRF
• Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics
• L1-Norm Batch Normalization for Efficient Training of Deep Neural Networks
• Incorporating BERT into Neural Machine Translation
• MLPerf Training Benchmark
• High-Throughput In-Memory Computing for Binary Deep Neural Networks with Monolithically Integrated RRAM and 90nm CMOS
• Ten Years of Generative Adversarial Nets (GANs): A survey of the state-of-the-art
• End-to-end ASR: from Supervised to Semi-Supervised Learning with Modern Architectures
• Reservoir Computing with Random Skyrmion Textures
• Neural Transfer Learning for Repairing Security Vulnerabilities in C Code
• Recurrent Neural Networks (RNNs): A gentle Introduction and Overview
• Adversarial Ranking for Language Generation
• How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacy
• Named Entity Recognition and Classification on Historical Documents: A Survey
• ImageBERT: Cross-modal Pre-training with Large-scale Weak-supervised Image-Text Data
• Assessing The Factual Accuracy of Generated Text
• Survey on reinforcement learning for language processing
• Large scale distributed neural network training through online distillation
• Is Neural Machine Translation Ready for Deployment? A Case Study on 30 Translation Directions
• Text Summarization with Pretrained Encoders
• VideoBERT: A Joint Model for Video and Language Representation Learning
• Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation
• Tacotron: Towards End-to-End Speech Synthesis
• Fault Detection and Identification using Bayesian Recurrent Neural Networks
• Supporting Very Large Models using Automatic Dataflow Graph Partitioning
• Beyond Data and Model Parallelism for Deep Neural Networks
• How to Prove Your Model Belongs to You: A Blind-Watermark based Framework to Protect Intellectual Property of DNN
• Bayesian Recurrent Neural Networks
• Exploring Sparsity in Recurrent Neural Networks
• SignBERT+: Hand-model-aware Self-supervised Pre-training for Sign Language Understanding
• Compressing Large-Scale Transformer-Based Models: A Case Study on BERT
• Weighted Transformer Network for Machine Translation
• Sharp Minima Can Generalize For Deep Nets
• Learning the PE Header, Malware Detection with Minimal Domain Knowledge
• CANINE: Pre-training an Efficient Tokenization-Free Encoder for Language Representation
• ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning
• Investigating Backtranslation in Neural Machine Translation
• AutoML-Zero: Evolving Machine Learning Algorithms From Scratch
• Approximating Continuous Functions by ReLU Nets of Minimal Width
• Comparing BERT against traditional machine learning text classification
• Learning a bidirectional mapping between human whole-body motion and natural language using deep recurrent neural networks
• Emotion Recognition From Speech With Recurrent Neural Networks
• Learn What Not to Learn: Action Elimination with Deep Reinforcement Learning
• Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN
• Linking the Neural Machine Translation and the Prediction of Organic Chemistry Reactions
• On the Explainability of Natural Language Processing Deep Models
• Neural Machine Translation and Sequence-to-sequence Models: A Tutorial
• A Comprehensive Exploration on WikiSQL with Table-Aware Word Contextualization
• Identifying Solar Flare Precursors Using Time Series of SDO/HMI Images and SHARP Parameters
• Pretrained Language Models for Sequential Sentence Classification
• Unicoder-VL: A Universal Encoder for Vision and Language by Cross-modal Pre-training
• On the Dimensionality of Word Embedding
• Chemception: A Deep Neural Network with Minimal Chemistry Knowledge Matches the Performance of Expert-developed QSAR/QSPR Models
• Future Directions in Human Mobility Science
• Recent Advances in Deep Learning: An Overview
• TeaNet: universal neural network interatomic potential inspired by iterative electronic relaxations
• Measurement of Anomalous Diffusion Using Recurrent Neural Networks
• Compressing Recurrent Neural Network with Tensor Train
• Hardware Acceleration of Sparse and Irregular Tensor Computations of ML Models: A Survey and Insights
• Review: Deep Learning in Electron Microscopy
• Physics-informed linear regression is competitive with two Machine Learning methods in residential building MPC
• Sample Efficient Adaptive Text-to-Speech
• Terminal Brain Damage: Exposing the Graceless Degradation in Deep Neural Networks Under Hardware Fault Attacks
• A Teacher-Student Framework for Zero-Resource Neural Machine Translation
• TSViz: Demystification of Deep Learning Models for Time-Series Analysis
• Factorization tricks for LSTM networks
• StructBERT: Incorporating Language Structures into Pre-training for Deep Language Understanding
• Machine learning with neural networks
• MeanSum: A Neural Model for Unsupervised Multi-document Abstractive Summarization
• A Survey of Deep Learning Techniques for Neural Machine Translation
• Atomic Convolutional Networks for Predicting Protein-Ligand Binding Affinity
• Fluency-Guided Cross-Lingual Image Captioning
• Learning Deep Transformer Models for Machine Translation
• Block-Sparse Recurrent Neural Networks
• DEPTWEET: A Typology for Social Media Texts to Detect Depression Severities
• How to Fine-Tune BERT for Text Classification?
• A Comprehensive Survey of Large Language Models and Multimodal Large Language Models in Medicine
• A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation
• A Surprisingly Robust Trick for Winograd Schema Challenge
• VIOLET : End-to-End Video-Language Transformers with Masked Visual-token Modeling
• From Image to Language: A Critical Analysis of Visual Question Answering (VQA) Approaches, Challenges, and Opportunities
• Assuring the Machine Learning Lifecycle: Desiderata, Methods, and Challenges
• NEZHA: Neural Contextualized Representation for Chinese Language Understanding
• DCN+: Mixed Objective and Deep Residual Coattention for Question Answering
• What the [MASK]? Making Sense of Language-Specific BERT Models
• Deep Learning for Steganalysis of Diverse Data Types: A review of methods, taxonomy, challenges and future directions
• Pre-Training BERT on Arabic Tweets: Practical Considerations
• Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustness
• Charformer: Fast Character Transformers via Gradient-based Subword Tokenization
• Self-Attentive Sequential Recommendation
• PhenoTagger: A Hybrid Method for Phenotype Concept Recognition using Human Phenotype Ontology
• Fonduer: Knowledge Base Construction from Richly Formatted Data
• OpenNMT: Neural Machine Translation Toolkit
• Story Generation from Sequence of Independent Short Descriptions
• Adv-BERT: BERT is not robust on misspellings! Generating nature adversarial samples on BERT
• Toward Subgraph-Guided Knowledge Graph Question Generation with Graph Neural Networks
• Zero-Shot Cross-lingual Classification Using Multilingual Neural Machine Translation
• Regularizing Deep Neural Networks by Noise: Its Interpretation and Optimization
• IndicBART: A Pre-trained Model for Indic Natural Language Generation
• On Accurate Evaluation of GANs for Language Generation
• Selfie: Self-supervised Pretraining for Image Embedding
• Dive into Deep Learning
• Globally Optimal Gradient Descent for a ConvNet with Gaussian Inputs
• Generating Wikipedia by Summarizing Long Sequences
• SYSTRAN's Pure Neural Machine Translation Systems
• Towards Safer Smart Contracts: A Sequence Learning Approach to Detecting Security Threats
• Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programming
• Efficient Neural Audio Synthesis
• Reversible Architectures for Arbitrarily Deep Residual Neural Networks
• Exploring Neural Transducers for End-to-End Speech Recognition
• Adversarial Neural Machine Translation
• On the Effectiveness of Mitigating Data Poisoning Attacks with Gradient Shaping
• Align, Mask and Select: A Simple Method for Incorporating Commonsense Knowledge into Language Representation Models
• Resurrecting the sigmoid in deep learning through dynamical isometry: theory and practice
• Massive Exploration of Neural Machine Translation Architectures
• Neural Abstractive Text Summarization with Sequence-to-Sequence Models
• Latent Space Data Assimilation by using Deep Learning
• Deep learning for presumed probability density function models
• The Sockeye 2 Neural Machine Translation Toolkit at AMTA 2020
• Modeling Global and Local Node Contexts for Text Generation from Knowledge Graphs
• Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation
• A Study of BFLOAT16 for Deep Learning Training
• Multi-level Residual Networks from Dynamical Systems View
• An Electro-Photonic System for Accelerating Deep Neural Networks
• All-Optical Spiking Neuron Based On Passive Micro-Resonator
• Physics-guided Deep Markov Models for Learning Nonlinear Dynamical Systems with Uncertainty
• Deep learning-based NLP Data Pipeline for EHR Scanned Document Information Extraction
• Learning from flowsheets: A generative transformer model for autocompletion of flowsheets
• Deep Learning for Environmentally Robust Speech Recognition: An Overview of Recent Developments
• SMILES2Vec: An Interpretable General-Purpose Deep Neural Network for Predicting Chemical Properties
• Semantic Tagging with Deep Residual Networks
• Fast Structured Decoding for Sequence Models
• Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topology
• Recent Advances in Natural Language Inference: A Survey of Benchmarks, Resources, and Approaches
• Empowering Things with Intelligence: A Survey of the Progress, Challenges, and Opportunities in Artificial Intelligence of Things
• A Convolutional Encoder Model for Neural Machine Translation
• LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding
• On Characterizing the Capacity of Neural Networks using Algebraic Topology
• Beto, Bentz, Becas: The Surprising Cross-Lingual Effectiveness of BERT
• Focus Group on Artificial Intelligence for Health
• Towards better decoding and language model integration in sequence to sequence models
• Combinatorial Testing for Deep Learning Systems
• A Comprehensive Survey on Hardware-Aware Neural Architecture Search
• BP-Transformer: Modelling Long-Range Context via Binary Partitioning
• Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifiers
• InterBERT: Vision-and-Language Interaction for Multi-modal Pretraining
• Gram-CTC: Automatic Unit Selection and Target Decomposition for Sequence Labelling
• Just ASK: Building an Architecture for Extensible Self-Service Spoken Language Understanding
• TBD: Benchmarking and Analyzing Deep Neural Network Training
• Challenges in Data-to-Document Generation
• TIPRDC: Task-Independent Privacy-Respecting Data Crowdsourcing Framework for Deep Learning with Anonymized Intermediate Representations
• Talk2Nav: Long-Range Vision-and-Language Navigation with Dual Attention and Spatial Memory
• Sequence-to-Sequence Models Can Directly Translate Foreign Speech
• GLAC Net: GLocal Attention Cascading Networks for Multi-image Cued Story Generation
• Training LSTM Networks with Resistive Cross-Point Devices
• Learning Algorithms for Active Learning
• Good Subnetworks Provably Exist: Pruning via Greedy Forward Selection
• Neural Network Distiller: A Python Package For DNN Compression Research
• CTRAN: CNN-Transformer-based Network for Natural Language Understanding
• HERO: Hierarchical Encoder for Video+Language Omni-representation Pre-training
• Text Compression-aided Transformer Encoding
• Local Monotonic Attention Mechanism for End-to-End Speech and Language Processing
• On the Binding Problem in Artificial Neural Networks
• Reinforcement Learning for Generative AI: State of the Art, Opportunities and Open Research Challenges
• Less is More: ClipBERT for Video-and-Language Learning via Sparse Sampling
• Tensorized Embedding Layers for Efficient Model Compression
• Multi-task Paired Masking with Alignment Modeling for Medical Vision-Language Pre-training
• Sticking to the Facts: Confident Decoding for Faithful Data-to-Text Generation
• DP-Forward: Fine-tuning and Inference on Language Models with Differential Privacy in Forward Pass
• M6: A Chinese Multimodal Pretrainer
• Identity Crisis: Memorization and Generalization under Extreme Overparameterization
• Reinforcement Learning Based Graph-to-Sequence Model for Natural Question Generation
• A Survey on State-of-the-art Deep Learning Applications and Challenges
• Differentiable Physics-informed Graph Networks
• Bridging Neural Machine Translation and Bilingual Dictionaries
• A Hybrid Convolutional Variational Autoencoder for Text Generation
• Universal Neural Machine Translation for Extremely Low Resource Languages
• Language Models with Transformers
• Multi-layer Representation Fusion for Neural Machine Translation
• Cross-Lingual Machine Reading Comprehension
• Faster Fuzzing: Reinitialization with Deep Neural Models
• An Unsupervised Autoregressive Model for Speech Representation Learning
• Reweighted Proximal Pruning for Large-Scale Language Representation
• TensorOpt: Exploring the Tradeoffs in Distributed DNN Training with Auto-Parallelism
• Attend and Diagnose: Clinical Time Series Analysis using Attention Models
• Generative Language Modeling for Automated Theorem Proving
• Massively Multilingual Neural Machine Translation
• Cache Telepathy: Leveraging Shared Resource Attacks to Learn DNN Architectures
• Neural Information Retrieval: A Literature Review
• Computational linguistics and Natural Language Processing
• Scalable Deep Learning on Distributed Infrastructures: Challenges, Techniques and Tools
• Natural Language Understanding with Privacy-Preserving BERT
• Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models
• Accurate Machine Learning Atmospheric Retrieval via a Neural Network Surrogate Model for Radiative Transfer
• Dynamics of Deep Neural Networks and Neural Tangent Hierarchy
• Natural Language Understanding for Argumentative Dialogue Systems in the Opinion Building Domain
• Attention U-Net as a surrogate model for groundwater prediction
• Minimizing the Bag-of-Ngrams Difference for Non-Autoregressive Neural Machine Translation
• Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision
• Mixed Precision Training With 8-bit Floating Point
• Deciphering the Language of Nature: A transformer-based language model for deleterious mutations in proteins
• A Selective Overview of Deep Learning
• Mixed-Precision Training for NLP and Speech Recognition with OpenSeq2Seq
• Calibration of Encoder Decoder Models for Neural Machine Translation
• Enabling Compute-Communication Overlap in Distributed Deep Learning Training Platforms
• Semantic Sentence Matching with Densely-connected Recurrent and Co-attentive Information
• On the Choice of Modeling Unit for Sequence-to-Sequence Speech Recognition
• On the Effectiveness of Transfer Learning for Code Search
• AMMUS : A Survey of Transformer-based Pretrained Models in Natural Language Processing
• Using Language Model to Bootstrap Human Activity Recognition Ambient Sensors Based in Smart Homes
• Deep Learning Based Chatbot Models
• GraphFormers: GNN-nested Transformers for Representation Learning on Textual Graph
• VALUE: A Multi-Task Benchmark for Video-and-Language Understanding Evaluation
• On the Geometry of Adversarial Examples
• Local minima in training of neural networks
• An Evaluation of Neural Machine Translation Models on Historical Spelling Normalization
• DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inference
• Two-Timescale Voltage Control in Distribution Grids Using Deep Reinforcement Learning
• Robust Neural Machine Translation with Doubly Adversarial Inputs
• Learning to summarize from human feedback
• Deep reinforcement learning-designed radiofrequency waveform in MRI
• End-to-End Speech Translation with Knowledge Distillation
• Learning to Generate Corrective Patches using Neural Machine Translation
• Capturing Evolution in Word Usage: Just Add More Clusters?
• Hierarchical Multitask Learning for CTC-based Speech Recognition
• XPipe: Efficient Pipeline Model Parallelism for Multi-GPU DNN Training
• Scale MLPerf-0.6 models on Google TPU-v3 Pods
• CG-BERT: Conditional Text Generation with BERT for Generalized Few-shot Intent Detection
• Pre-trained Language Models in Biomedical Domain: A Systematic Survey
• Deep Independently Recurrent Neural Network (IndRNN)
• Provable Benefit of Orthogonal Initialization in Optimizing Deep Linear Networks
• Bridging the Gap between Spatial and Spectral Domains: A Survey on Graph Neural Networks
• Hierarchical Transformers for Multi-Document Summarization
• Search Engine Guided Non-Parametric Neural Machine Translation
• Passport-aware Normalization for Deep Model Protection
• From Zero to Hero: On the Limitations of Zero-Shot Cross-Lingual Transfer with Multilingual Transformers
• Controlling the Output Length of Neural Machine Translation
• Generating Sentences by Editing Prototypes
• Zero-shot Knowledge Transfer via Adversarial Belief Matching
• Neural Text Generation: A Practical Guide
• Accuracy of neural networks for the simulation of chaotic dynamics: precision of training data vs precision of the algorithm
• Aspect Term Extraction with History Attention and Selective Transformation
• Rethinking Cross-modal Interaction from a Top-down Perspective for Referring Video Object Segmentation
• Interpreting Black Box Models via Hypothesis Testing
• A Survey of Orthographic Information in Machine Translation
• HeBERT & HebEMO: a Hebrew BERT Model and a Tool for Polarity Analysis and Emotion Recognition
• A Survey on Neural Network Interpretability
• Using BERT for Word Sense Disambiguation
• SAMU-XLSR: Semantically-Aligned Multimodal Utterance-level Cross-Lingual Speech Representation
• Characterization of anomalous diffusion through convolutional transformers
• DRAGNN: A Transition-based Framework for Dynamically Connected Neural Networks
• A Call for Prudent Choice of Subword Merge Operations in Neural Machine Translation
• A Distributional Approach to Controlled Text Generation
• A Survey on Contextualised Semantic Shift Detection
• Transfer Fine-Tuning: A BERT Case Study
• Information Aware Max-Norm Dirichlet Networks for Predictive Uncertainty Estimation
• Joint Training for Neural Machine Translation Models with Monolingual Data
• Diagnosing BERT with Retrieval Heuristics
• VD-BERT: A Unified Vision and Dialog Transformer with BERT
• Progress Notes Classification and Keyword Extraction using Attention-based Deep Learning Models with BERT
• Placeto: Learning Generalizable Device Placement Algorithms for Distributed Machine Learning
• VAIN: Attentional Multi-agent Predictive Modeling
• DAPPLE: A Pipelined Data Parallel Approach for Training Large Models
• Machine Reading Comprehension: a Literature Review
• TwinBERT: Distilling Knowledge to Twin-Structured BERT Models for Efficient Retrieval
• A Dual-Attention Learning Network with Word and Sentence Embedding for Medical Visual Question Answering
• SpeechBERT: An Audio-and-text Jointly Learned Language Model for End-to-end Spoken Question Answering
• Stochastic Mirror Descent on Overparameterized Nonlinear Models: Convergence, Implicit Regularization, and Generalization
• ST-HOI: A Spatial-Temporal Baseline for Human-Object Interaction Detection in Videos
• Accelerating Transformer Inference for Translation via Parallel Decoding
• Visual Relationship Detection with Visual-Linguistic Knowledge from Multimodal Representations
• Geometry of energy landscapes and the optimizability of deep neural networks
• CharacterBERT: Reconciling ELMo and BERT for Word-Level Open-Vocabulary Representations From Characters
• Localising In-Domain Adaptation of Transformer-Based Biomedical Language Models
• The Impact of Preprocessing on Arabic-English Statistical and Neural Machine Translation
• They, Them, Theirs: Rewriting with Gender-Neutral English
• Automated Architecture Design for Deep Neural Networks
• Combating Adversarial Misspellings with Robust Word Recognition
• scb-mt-en-th-2020: A Large English-Thai Parallel Corpus
• Towards Transfer Learning for End-to-End Speech Synthesis from Deep Pre-Trained Language Models
• Neural Program Synthesis with Priority Queue Training
• Discovering the Compositional Structure of Vector Representations with Role Learning Networks
• Unsupervised State Representation Learning in Atari
• Explaining Software Bugs Leveraging Code Structures in Neural Machine Translation
• TRANS-BLSTM: Transformer with Bidirectional LSTM for Language Understanding
• EVIL: Exploiting Software via Natural Language
• Online Learning for Neural Machine Translation Post-editing
• GDP: Generalized Device Placement for Dataflow Graphs
• Learning to Segment Inputs for NMT Favors Character-Level Processing
• Seeing Out of tHe bOx: End-to-End Pre-training for Vision-Language Representation Learning
• Platform for Situated Intelligence
• Unsupervised Cross-lingual Adaptation for Sequence Tagging and Beyond
• IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages
• Sequence tagging for biomedical extractive question answering
• Exploiting Deep Representations for Neural Machine Translation
• Dual-mode ASR: Unify and Improve Streaming ASR with Full-context Modeling
• Boosting Neural Networks to Decompile Optimized Binaries
• Machine-Translation History and Evolution: Survey for Arabic-English Translations
• CharacterBERT and Self-Teaching for Improving the Robustness of Dense Retrievers on Queries with Typos
• Query-Based Abstractive Summarization Using Neural Networks
• Concept Learning through Deep Reinforcement Learning with Memory-Augmented Neural Networks
• EmotionX-KU: BERT-Max based Contextual Emotion Classifier
• CheMixNet: Mixed DNN Architectures for Predicting Chemical Properties using Multiple Molecular Representations
• KR-BERT: A Small-Scale Korean-Specific Language Model
• Can We Generate Shellcodes via Natural Language? An Empirical Study
• Context in Neural Machine Translation: A Review of Models and Evaluations
• Modeling Source Syntax for Neural Machine Translation
• Improving Sequence-to-Sequence Learning via Optimal Transport
• Neural Language Generation: Formulation, Methods, and Evaluation
• Mask and Cloze: Automatic Open Cloze Question Generation using a Masked Language Model
• AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference
• A Comparative Study of Transformer-Based Language Models on Extractive Question Answering
• Direct speech-to-speech translation with a sequence-to-sequence model
• Bridging Text and Video: A Universal Multimodal Transformer for Video-Audio Scene-Aware Dialog
• A Targeted Attack on Black-Box Neural Machine Translation with Parallel Data Poisoning
• Low Dimensional Convolutional Neural Network For Solar Flares GOES Time Series Classification
• Cold-Start Reinforcement Learning with Softmax Policy Gradient
• Themis: Fair and Efficient GPU Cluster Scheduling
• Multi-Head Attention with Disagreement Regularization
• Deep Learning for Scene Classification: A Survey
• Enabling Multi-Source Neural Machine Translation By Concatenating Source Sentences In Multiple Languages
• An Exploration of Word Embedding Initialization in Deep-Learning Tasks
• Analyzing and Exploiting NARX Recurrent Neural Networks for Long-Term Dependencies
• Phenotyping of Clinical Notes with Improved Document Classification Models Using Contextualized Neural Language Models
• BioNetExplorer: Architecture-Space Exploration of Bio-Signal Processing Deep Neural Networks for Wearables
• Hierarchically Structured Reinforcement Learning for Topically Coherent Visual Story Generation
• Recommending Themes for Ad Creative Design via Visual-Linguistic Representations
• Paraphrase Generation with Deep Reinforcement Learning
• Neutron: An Implementation of the Transformer Translation Model and its Variants
• BIGPATENT: A Large-Scale Dataset for Abstractive and Coherent Summarization
• Learning to Parse and Translate Improves Neural Machine Translation
• Learning to Evaluate Translation Beyond English: BLEURT Submissions to the WMT Metrics 2020 Shared Task
• Confidence through Attention
• Characterizing the Deep Neural Networks Inference Performance of Mobile Applications
• Foundations of Vector Retrieval
• Mitigating Edge Machine Learning Inference Bottlenecks: An Empirical Study on Accelerating Google Edge Models
• Distilling Knowledge Learned in BERT for Text Generation
• Inclusion of Domain-Knowledge into GNNs using Mode-Directed Inverse Entailment
• A Hierarchical Multi-Modal Encoder for Moment Localization in Video Corpus
• Machine learning with observers predicts complex spatiotemporal behavior
• Noisin: Unbiased Regularization for Recurrent Neural Networks
• Scaling Laws for Neural Machine Translation
• Wat zei je? Detecting Out-of-Distribution Translations with Variational Transformers
• Learning Translations via Matrix Completion
• MKD: a Multi-Task Knowledge Distillation Approach for Pretrained Language Models
• Towards Robust Neural Machine Translation
• Marvel: A Data-centric Compiler for DNN Operators on Spatial Accelerators
• Rate Distortion For Model Compression: From Theory To Practice
• Improving Readability for Automatic Speech Recognition Transcription
• Detecting Hallucinated Content in Conditional Neural Sequence Generation
• Bertrand-DR: Improving Text-to-SQL using a Discriminative Re-ranker
• Implicit Dimension Identification in User-Generated Text with LSTM Networks
• Measuring Annotator Agreement Generally across Complex Structured, Multi-object, and Free-text Annotation Tasks
• Automatic Cross-Replica Sharding of Weight Update in Data-Parallel Training
• Machine Translation of Low-Resource Spoken Dialects: Strategies for Normalizing Swiss German
• Does injecting linguistic structure into language models lead to better alignment with brain recordings?
• Federated Learning of N-gram Language Models
• Efficient Visual Recognition with Deep Neural Networks: A Survey on Recent Advances and New Directions
• Memory-augmented Neural Machine Translation
• Margin-based Parallel Corpus Mining with Multilingual Sentence Embeddings
• Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for Sampling Sequences Without Replacement
• Zero-Shot Dual Machine Translation
• Unsupervised Neural Machine Translation with SMT as Posterior Regularization
• Contrastive Visual-Linguistic Pretraining
• Learning to Focus: Cascaded Feature Matching Network for Few-shot Image Recognition
• SDM: Sequential Deep Matching Model for Online Large-scale Recommender System
• Multi-Channel Auto-Calibration for the Atmospheric Imaging Assembly using Machine Learning
• Ancient-Modern Chinese Translation with a Large Training Dataset
• Utilizing Language Relatedness to improve Machine Translation: A Case Study on Languages of the Indian Subcontinent
• Learning by Abstraction: The Neural State Machine
• How to Measure Gender Bias in Machine Translation: Optimal Translators, Multiple Reference Points
• A Study of Reinforcement Learning for Neural Machine Translation
• Estimate and Replace: A Novel Approach to Integrating Deep Neural Networks with Existing Applications
• Specializing Multilingual Language Models: An Empirical Study
• Multilingual Constituency Parsing with Self-Attention and Pre-Training
• Learning Convolutional Text Representations for Visual Question Answering
• Exploring Benefits of Transfer Learning in Neural Machine Translation
• Non-Autoregressive Neural Machine Translation with Enhanced Decoder Input
• Competence-based Curriculum Learning for Neural Machine Translation
• A Survey on Document-level Neural Machine Translation: Methods and Evaluation
• Training Deeper Neural Machine Translation Models with Transparent Attention
• Time-aware Large Kernel Convolutions
• Adversarial Training with Voronoi Constraints
• VIVO: Visual Vocabulary Pre-Training for Novel Object Captioning
• Pretrained Language Models for Document-Level Neural Machine Translation
• Bit-Flip Attack: Crushing Neural Network with Progressive Bit Search
• Context-Aware Learning for Neural Machine Translation
• DICT-MLM: Improved Multilingual Pre-Training using Bilingual Dictionaries
• Do Neural Nets Learn Statistical Laws behind Natural Language?
• Extracting UMLS Concepts from Medical Text Using General and Domain-Specific Deep Learning Models
• When and Why are Pre-trained Word Embeddings Useful for Neural Machine Translation?
• A Random Matrix Perspective on Mixtures of Nonlinearities for Deep Learning
• Multi-branch Attentive Transformer
• Dynamic Layer Aggregation for Neural Machine Translation with Routing-by-Agreement
• Auditing Data Provenance in Text-Generation Models
• Neural Machine Translating from Natural Language to SPARQL
• Analogue Quantum Simulation: A New Instrument for Scientific Understanding
• A Universal Parent Model for Low-Resource Neural Machine Translation Transfer
• Triangular Architecture for Rare Language Translation
• Multilingual Named Entity Recognition Using Pretrained Embeddings, Attention Mechanism and NCRF
• Pre-Trained Models: Past, Present and Future
• Cross-Lingual Semantic Role Labeling with High-Quality Translated Training Corpus
• Anomaly Detection in Multivariate Non-stationary Time Series for Automatic DBMS Diagnosis
• Efficient Algorithms for Device Placement of DNN Graph Operators
• Multilingual Name Entity Recognition and Intent Classification Employing Deep Learning Architectures
• Understanding Reuse, Performance, and Hardware Cost of DNN Dataflows: A Data-Centric Approach Using MAESTRO
• ATOM: Commit Message Generation Based on Abstract Syntax Tree and Hybrid Ranking
• PAWS: Paraphrase Adversaries from Word Scrambling
• Automatic Generation of Text Descriptive Comments for Code Blocks
• Leveraging Weakly Supervised Data to Improve End-to-End Speech-to-Text Translation
• A Deep Reinforcement Learning Chatbot (Short Version)
• Unsupervised Multi-modal Neural Machine Translation
• Match-Tensor: a Deep Relevance Model for Search
• Breaking the Beam Search Curse: A Study of (Re-)Scoring Methods and Stopping Criteria for Neural Machine Translation
• Eigenvalue distribution of nonlinear models of random matrices
• Supersonic: Learning to Generate Source Code Optimizations in C/C++
• Slip Detection: Analysis and Calibration of Univariate Tactile Signals
• Local Contextual Attention with Hierarchical Structure for Dialogue Act Recognition
• Accelerating Sparse Deep Neural Networks
• Liputan6: A Large-scale Indonesian Dataset for Text Summarization
• Multimodal Joint Attribute Prediction and Value Extraction for E-commerce Product
• Efficient (Soft) Q-Learning for Text Generation with Limited Good Data
• Ad Headline Generation using Self-Critical Masked Language Model
• SS-Auto: A Single-Shot, Automatic Structured Weight Pruning Framework of DNNs with Ultra-High Efficiency
• Deep Neural Machine Translation with Linear Associative Unit
• Trainable Greedy Decoding for Neural Machine Translation
• Tensor2Tensor for Neural Machine Translation
• Unsupervised Neural Machine Translation with Weight Sharing
• LIMIT-BERT : Linguistic Informed Multi-Task BERT
• From Audio to Semantics: Approaches to end-to-end spoken language understanding
• ExplainaBoard: An Explainable Leaderboard for NLP
• When Does Pretraining Help? Assessing Self-Supervised Learning for Law and the CaseHOLD Dataset
• Espresso: A Fast End-to-end Neural Speech Recognition Toolkit
• Deep Recurrent Neural Network for Protein Function Prediction from Sequence
• Token-Level Ensemble Distillation for Grapheme-to-Phoneme Conversion
• Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection
• A Variational Dirichlet Framework for Out-of-Distribution Detection
• Topic Modeling with Contextualized Word Representation Clusters
• SparseBERT: Rethinking the Importance Analysis in Self-attention
• Why Do Masked Neural Language Models Still Need Common Sense Knowledge?
• Towards Understanding the Faults of JavaScript-Based Deep Learning Systems
• Bayesian Sparsification of Recurrent Neural Networks
• Kaleido-BERT: Vision-Language Pre-training on Fashion Domain
• OmniNet: A unified architecture for multi-modal multi-task learning
• Adversarial Attacks and Defense on Texts: A Survey
• VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
• BoostingBERT:Integrating Multi-Class Boosting into BERT for NLP Tasks
• Natural Question Generation with Reinforcement Learning Based Graph-to-Sequence Model
• Orchestrating the Development Lifecycle of Machine Learning-Based IoT Applications: A Taxonomy and Survey
• Information Aggregation for Multi-Head Attention with Routing-by-Agreement
• SEAL: Segment-wise Extractive-Abstractive Long-form Text Summarization
• PharmMT: A Neural Machine Translation Approach to Simplify Prescription Directions
• Synthetic Test Data Generation Using Recurrent Neural Networks: A Position Paper
• Multimodal Image Captioning for Marketing Analysis
• Simple, Scalable Adaptation for Neural Machine Translation
• Single-/Multi-Source Cross-Lingual NER via Teacher-Student Learning on Unlabeled Data in Target Language
• Improving Neural Machine Translation with Conditional Sequence Generative Adversarial Nets
• Steering Output Style and Topic in Neural Response Generation
• One-shot and few-shot learning of word embeddings
• A Correspondence Between Random Neural Networks and Statistical Field Theory
• Cross-Lingual BERT Transformation for Zero-Shot Dependency Parsing
• Domain-specific Communication Optimization for Distributed DNN Training
• BERT Meets Chinese Word Segmentation
• Simplify-then-Translate: Automatic Preprocessing for Black-Box Machine Translation
• Tiny Transformers for Environmental Sound Classification at the Edge
• Connecting the Dots Between MLE and RL for Sequence Prediction
• Contextual Parameter Generation for Universal Neural Machine Translation
• Reinforcement Learning Upside Down: Don't Predict Rewards -- Just Map Them to Actions
• QAScore -- An Unsupervised Unreferenced Metric for the Question Generation Evaluation
• Learning to Write Notes in Electronic Health Records
• Knowledge Enhanced Contextual Word Representations
• Towards Preemptive Detection of Depression and Anxiety in Twitter
• Hybrid Self-Attention Network for Machine Translation
• Fast Cross-domain Data Augmentation through Neural Sentence Editing
• A Research Agenda: Dynamic Models to Defend Against Correlated Attacks
• Latent-Variable Non-Autoregressive Neural Machine Translation with Deterministic Inference Using a Delta Posterior
• Noisy Self-Knowledge Distillation for Text Summarization
• Neural network gradient-based learning of black-box function interfaces
• AMBERT: A Pre-trained Language Model with Multi-Grained Tokenization
• Beam Search with Bidirectional Strategies for Neural Response Generation
• SMT vs NMT: A Comparison over Hindi & Bengali Simple Sentences
• Information-Theoretic Text Hallucination Reduction for Video-grounded Dialogue
• Personalized Re-ranking for Recommendation
• Deep Neural Machine Translation with Weakly-Recurrent Units
• AutoDDL: Automatic Distributed Deep Learning with Near-Optimal Bandwidth Cost
• Low-Precision Batch-Normalized Activations
• Understanding and Improving Encoder Layer Fusion in Sequence-to-Sequence Learning
• Semi-Autoregressive Neural Machine Translation
• Tree Transformer: Integrating Tree Structures into Self-Attention
• An Encoder-Decoder Framework Translating Natural Language to Database Queries
• Large-Scale Discrete Fourier Transform on TPUs
• On the Interpolation of Contextualized Term-based Ranking with BM25 for Query-by-Example Retrieval
• CREDIT: Coarse-to-Fine Sequence Generation for Dialogue State Tracking
• BERTSurv: BERT-Based Survival Models for Predicting Outcomes of Trauma Patients
• AI Matrix: A Deep Learning Benchmark for Alibaba Data Centers
• Creative GANs for generating poems, lyrics, and metaphors
• Efficient Attentions for Long Document Summarization
• PaSE: Parallelization Strategies for Efficient DNN Training
• Neural networks are a priori biased towards Boolean functions with low entropy
• Morphosyntactic probing of multilingual BERT models
• Incremental Adaptation of NMT for Professional Post-editors: A User Study
• Retrieving Sequential Information for Non-Autoregressive Neural Machine Translation
• An Empirical Study of Large-Batch Stochastic Gradient Descent with Structured Covariance Noise
• Hard-Coded Gaussian Attention for Neural Machine Translation
• What does Attention in Neural Machine Translation Pay Attention to?
• SimCLS: A Simple Framework for Contrastive Learning of Abstractive Summarization
• Toward a Period-Specific Optimized Neural Network for OCR Error Correction of Historical Hebrew Texts
• Translating Phrases in Neural Machine Translation
• BERT-DST: Scalable End-to-End Dialogue State Tracking with Bidirectional Encoder Representations from Transformer
• Improving Sign Language Translation with Monolingual Data by Sign Back-Translation
• Code-switched inspired losses for generic spoken dialog representations
• POINTER: Constrained Progressive Text Generation via Insertion-based Generative Pre-training
• Neural Machine Translation: Challenges, Progress and Future
• Analysis of Predictive Coding Models for Phonemic Representation Learning in Small Datasets
• Can You Tell Me How to Get Past Sesame Street? Sentence-Level Pretraining Beyond Language Modeling
• Low-Resource Language Modelling of South African Languages
• ConfuciuX: Autonomous Hardware Resource Assignment for DNN Accelerators using Reinforcement Learning
• Non-Autoregressive Text Generation with Pre-trained Language Models
• Phrase-Based Attentions
• Compositional Generalization for Primitive Substitutions
• Named Entity Recognition in the Legal Domain using a Pointer Generator Network
• Keep and Learn: Continual Learning by Constraining the Latent Space for Knowledge Preservation in Neural Networks
• Exploring the limits of Concurrency in ML Training on Google TPUs
• Infusing Disease Knowledge into BERT for Health Question Answering, Medical Inference and Disease Name Recognition
• Data Augmentation for Low-Resource Named Entity Recognition Using Backtranslation
• Modelling Latent Translations for Cross-Lingual Transfer
• Defending Against Backdoor Attacks in Natural Language Generation
• KAISA: An Adaptive Second-Order Optimizer Framework for Deep Neural Networks
• AMR Parsing as Sequence-to-Graph Transduction
• Self-Supervised Learning to Prove Equivalence Between Straight-Line Programs via Rewrite Rules
• Lacking the embedding of a word? Look it up into a traditional dictionary
• Building Chinese Biomedical Language Models via Multi-Level Text Discrimination
• LAnoBERT: System Log Anomaly Detection based on BERT Masked Language Model
• Exploring Architectures, Data and Units For Streaming End-to-End Speech Recognition with RNN-Transducer
• CS1QA: A Dataset for Assisting Code-based Question Answering in an Introductory Programming Course
• Adversarial Learning for Chinese NER from Crowd Annotations
• Privacy-Preserving Models for Legal Natural Language Processing
• Plumber: Diagnosing and Removing Performance Bottlenecks in Machine Learning Data Pipelines
• XtremeDistilTransformers: Task Transfer for Task-agnostic Distillation
• Do Transformer Attention Heads Provide Transparency in Abstractive Summarization?
• Scalable Transformers for Neural Machine Translation
• Multi-Modal Detection of Alzheimer's Disease from Speech and Text
• Evaluating Transfer Learning for Simplifying GitHub READMEs
• Introducing Aspects of Creativity in Automatic Poetry Generation
• Large-Scale Deep Learning Optimizations: A Comprehensive Survey
• Give your Text Representation Models some Love: the Case for Basque
• Bag-of-Words as Target for Neural Machine Translation
• Navigating with Graph Representations for Fast and Scalable Decoding of Neural Language Models
• Dynamic Past and Future for Neural Machine Translation
• Implementing Neural Turing Machines
• CBAG: Conditional Biomedical Abstract Generation
• Improving Back-Translation with Uncertainty-based Confidence Estimation
• Atlas: A Dataset and Benchmark for E-commerce Clothing Product Categorization
• Discriminative Adversarial Search for Abstractive Summarization
• Literature Retrieval for Precision Medicine with Neural Matching and Faceted Summarization
• End-to-end Speech Recognition with Word-based RNN Language Models
• Speech-language Pre-training for End-to-end Spoken Language Understanding
• tf.data: A Machine Learning Data Processing Framework
• Transfer Learning in Multilingual Neural Machine Translation with Dynamic Vocabulary
• Distilling BERT into Simple Neural Networks with Unlabeled Transfer Data
• HanoiT: Enhancing Context-aware Translation via Selective Context
• Efficient and Generic 1D Dilated Convolution Layer for Deep Learning
• Improving Biomedical Pretrained Language Models with Knowledge
• Literature review on vulnerability detection using NLP technology
• Enhancing Machine Translation with Dependency-Aware Self-Attention
• Augmenting semantic lexicons using word embeddings and transfer learning
• StructuralLM: Structural Pre-training for Form Understanding
• Language as a matrix product state
• Synchronous Bidirectional Inference for Neural Sequence Generation
• A Computational Model of Commonsense Moral Decision Making
• Multi-object Tracking via End-to-end Tracklet Searching and Ranking
• A Study of Multilingual Neural Machine Translation
• Cross-lingual Pre-training Based Transfer for Zero-shot Neural Machine Translation
• Character-based NMT with Transformer
• An Empirical Study of Efficient ASR Rescoring with Transformers
• Large-scale Pretraining for Neural Machine Translation with Tens of Billions of Sentence Pairs
• A Latent Morphology Model for Open-Vocabulary Neural Machine Translation
• ERNIE at SemEval-2020 Task 10: Learning Word Emphasis Selection by Pre-trained Language Model
• Norm-Based Curriculum Learning for Neural Machine Translation
• SAFER: A Structure-free Approach for Certified Robustness to Adversarial Word Substitutions
• Biomedical Entity Representations with Synonym Marginalization
• Tell Me Why You Feel That Way: Processing Compositional Dependency for Tree-LSTM Aspect Sentiment Triplet Extraction (TASTE)
• BaPipe: Exploration of Balanced Pipeline Parallelism for DNN Training
• Neural Machine Translation: A Review of Methods, Resources, and Tools
• E2E-VLP: End-to-End Vision-Language Pre-training Enhanced by Visual Learning
• Automated Essay Scoring Using Transformer Models
• ChineseBERT: Chinese Pretraining Enhanced by Glyph and Pinyin Information
• Czert -- Czech BERT-like Model for Language Representation
• Hard but Robust, Easy but Sensitive: How Encoder and Decoder Perform in Neural Machine Translation
• Robust Deep Sensing Through Transfer Learning in Cognitive Radio
• Learning to Coordinate Multiple Reinforcement Learning Agents for Diverse Query Reformulation
• Regularizing Neural Machine Translation by Target-bidirectional Agreement
• Deep Probabilistic Programming Languages: A Qualitative Study
• Question Type Guided Attention in Visual Question Answering
• Feedforward and Recurrent Neural Networks Backward Propagation and Hessian in Matrix Form
• Chunk-Based Bi-Scale Decoder for Neural Machine Translation
• Later-stage Minimum Bayes-Risk Decoding for Neural Machine Translation
• Sharing Attention Weights for Fast Transformer
• Adversarial Learning of Privacy-Preserving and Task-Oriented Representations
• Importance-Aware Learning for Neural Headline Editing
• Why not be Versatile? Applications of the SGNMT Decoder for Machine Translation
• Neural Chinese Word Segmentation as Sequence to Sequence Translation
• Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision (Short Version)
• Semantic Explanations of Predictions
• BERT as a Teacher: Contextual Embeddings for Sequence-Level Reward
• Filtering and Mining Parallel Data in a Joint Multilingual Space
• Laplacian Networks: Bounding Indicator Function Smoothness for Neural Network Robustness
• Shifts: A Dataset of Real Distributional Shift Across Multiple Large-Scale Tasks
• DziriBERT: a Pre-trained Language Model for the Algerian Dialect
• Recurrent Graph Syntax Encoder for Neural Machine Translation
• A Comparison of Approaches to Document-level Machine Translation
• Discourse-Aware Neural Rewards for Coherent Text Generation
• Fast Sequence Generation with Multi-Agent Reinforcement Learning
• CNN Is All You Need
• Controllable Text Simplification with Explicit Paraphrasing
• How Robust Are Character-Based Word Embeddings in Tagging and MT Against Wrod Scramlbing or Randdm Nouse?
• GNN-XML: Graph Neural Networks for Extreme Multi-label Text Classification
• Problems with automating translation of movie/TV show subtitles
• On the quantization of recurrent neural networks
• Modernizing Historical Documents: a User Study
• Generating Representative Headlines for News Stories
• Consistency of a Recurrent Language Model With Respect to Incomplete Decoding
• LAMPRET: Layout-Aware Multimodal PreTraining for Document Understanding
• Product Insights: Analyzing Product Intents in Web Search
• Multiscale Collaborative Deep Models for Neural Machine Translation
• Deep neural network initialization with decision trees
• Pre-training Text Representations as Meta Learning
• iCapsNets: Towards Interpretable Capsule Networks for Text Classification
• High-Performance Deep Learning via a Single Building Block
• SkillSpan: Hard and Soft Skill Extraction from English Job Postings
• Intelligent Video Editing: Incorporating Modern Talking Face Generation Algorithms in a Video Editor
• A Target-Agnostic Attack on Deep Models: Exploiting Security Vulnerabilities of Transfer Learning
• Neural Class Expression Synthesis
• Detecting Overfitting via Adversarial Examples
• Enhanced Meta-Learning for Cross-lingual Named Entity Recognition with Minimal Resources
• UM-IU@LING at SemEval-2019 Task 6: Identifying Offensive Tweets Using BERT and SVMs
• ReWE: Regressing Word Embeddings for Regularization of Neural Machine Translation Systems
• Jointly Optimizing Diversity and Relevance in Neural Response Generation
• Word Shape Matters: Robust Machine Translation with Visual Embedding
• Mono vs Multilingual Transformer-based Models: a Comparison across Several Language Tasks
• HateProof: Are Hateful Meme Detection Systems really Robust?
• Modelling Sequential Music Track Skips using a Multi-RNN Approach
• Multi-node Bert-pretraining: Cost-efficient Approach
• Class LM and word mapping for contextual biasing in End-to-End ASR
• Neural Machine Translation with Adequacy-Oriented Learning
• End-to-End Speech Recognition: A review for the French Language
• Text Analysis in Adversarial Settings: Does Deception Leave a Stylistic Trace?
• Horizontally Fused Training Array: An Effective Hardware Utilization Squeezer for Training Novel Deep Learning Models
• Variable Name Recovery in Decompiled Binary Code using Constrained Masked Language Modeling
• Contextual Text Denoising with Masked Language Models
• CAGFuzz: Coverage-Guided Adversarial Generative Fuzzing Testing of Deep Learning Systems
• KVL-BERT: Knowledge Enhanced Visual-and-Linguistic BERT for Visual Commonsense Reasoning
• Analysis of the Evolution of Advanced Transformer-Based Language Models: Experiments on Opinion Mining
• Grid Search Hyperparameter Benchmarking of BERT, ALBERT, and LongFormer on DuoRC
• SuperChat: Dialogue Generation by Transfer Learning from Vision to Language using Two-dimensional Word Embedding and Pretrained ImageNet CNN Models
• Paying Attention to Multi-Word Expressions in Neural Machine Translation
• Multi-Head Multi-Layer Attention to Deep Language Representations for Grammatical Error Detection
• Psycholinguistic Tripartite Graph Network for Personality Detection
• EmbRace: Accelerating Sparse Communication for Distributed Training of NLP Neural Networks
• Superbloom: Bloom filter meets Transformer
• Echo: Compiler-based GPU Memory Footprint Reduction for LSTM RNN Training
• Improving the Performance of Online Neural Transducer Models
• DeLS-3D: Deep Localization and Segmentation with a 3D Semantic Map
• StrucTexT: Structured Text Understanding with Multi-Modal Transformers
• Parallax: Sparsity-aware Data Parallel Training of Deep Neural Networks
• Towards General Purpose Vision Systems
• Towards Neural Machine Translation with Partially Aligned Corpora
• On Accelerating Distributed Convex Optimizations
• Simplified End-to-End MMI Training and Voting for ASR
• Utilizing Character and Word Embeddings for Text Normalization with Sequence-to-Sequence Models
• Greedy Search with Probabilistic N-gram Matching for Neural Machine Translation
• Long Short-Term Attention
• KnowDis: Knowledge Enhanced Data Augmentation for Event Causality Detection via Distant Supervision
• Uniform-in-Time Weak Error Analysis for Stochastic Gradient Descent Algorithms via Diffusion Approximation
• Adaptive Dense-to-Sparse Paradigm for Pruning Online Recommendation System with Non-Stationary Data
• AdvAug: Robust Adversarial Augmentation for Neural Machine Translation
• Cross-Lingual Relation Extraction with Transformers
• Named entity recognition in chemical patents using ensemble of contextual language models
• Students Need More Attention: BERT-based AttentionModel for Small Data with Application to AutomaticPatient Message Triage
• Generating Logical Forms from Graph Representations of Text and Entities
• FashionBERT: Text and Image Matching with Adaptive Loss for Cross-modal Retrieval
• Dual Supervised Learning for Natural Language Understanding and Generation
• Optimizing Multi-GPU Parallelization Strategies for Deep Learning Training
• Prediction with a Short Memory
• Streaming Object Detection for 3-D Point Clouds
• Demystifying the MLPerf Benchmark Suite
• Understanding How Encoder-Decoder Architectures Attend
• Robust Beam Search for Encoder-Decoder Attention Based Speech Recognition without Length Bias
• Conditional Text Generation for Harmonious Human-Machine Interaction
• Question Generation by Transformers
• Reward Optimization for Neural Machine Translation with Learned Metrics
• Multi-Modal Data Augmentation for End-to-End ASR
• How can AI Automate End-to-End Data Science?
• Automatic Post-Editing for Machine Translation
• QuTiBench: Benchmarking Neural Networks on Heterogeneous Hardware
• Meta-path Analysis on Spatio-Temporal Graphs for Pedestrian Trajectory Prediction
• Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots
• Multi-Task Learning for Conversational Question Answering over a Large-Scale Knowledge Base
• A Deep Learning and Gamification Approach to Energy Conservation at Nanyang Technological University
• Optimal Completion Distillation for Sequence Learning
• exploRNN: Understanding Recurrent Neural Networks through Visual Exploration
• Composed Variational Natural Language Generation for Few-shot Intents
• Democratizing Production-Scale Distributed Deep Learning
• Recent Trends in the Use of Deep Learning Models for Grammar Error Handling
• Guiding High-Performance SAT Solvers with Unsat-Core Predictions
• A systematic comparison of grapheme-based vs. phoneme-based label units for encoder-decoder-attention models
• Cross-Modulation Networks for Few-Shot Learning
• AI-Powered Social Bots
• One Big Net For Everything
• Multimodal Dialogue State Tracking By QA Approach with Data Augmentation
• Going Wider: Recurrent Neural Network With Parallel Cells
• UniTrans: Unifying Model Transfer and Data Transfer for Cross-Lingual Named Entity Recognition with Unlabeled Data
• On the Blindspots of Convolutional Networks
• Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
• Hybrid Data-Model Parallel Training for Sequence-to-Sequence Recurrent Neural Network Machine Translation
• Pre-trained Language Model for Biomedical Question Answering
• Deep Ensembles on a Fixed Memory Budget: One Wide Network or Several Thinner Ones?
• Why Neural Machine Translation Prefers Empty Outputs
• Multimodal Learning for Hateful Memes Detection
• Unsupervised Multimodal Neural Machine Translation with Pseudo Visual Pivoting
• Reference Language based Unsupervised Neural Machine Translation
• Deep Neural Network for Semantic-based Text Recognition in Images
• MARMOT: A Deep Learning Framework for Constructing Multimodal Representations for Vision-and-Language Tasks
• Towards Unsupervised Language Understanding and Generation by Joint Dual Learning
• Assigning function to protein-protein interactions: a weakly supervised BioBERT based approach using PubMed abstracts
• Interactive-Predictive Neural Machine Translation through Reinforcement and Imitation
• Denoising based Sequence-to-Sequence Pre-training for Text Generation
• S2ORC: The Semantic Scholar Open Research Corpus
• Teaching Machines to Converse
• Exploring Hyper-Parameter Optimization for Neural Machine Translation on GPU Architectures
• Attentional Speech Recognition Models Misbehave on Out-of-domain Utterances
• Improving Deep Transformer with Depth-Scaled Initialization and Merged Attention
• How Can BERT Help Lexical Semantics Tasks?
• SHAPED: Shared-Private Encoder-Decoder for Text Style Adaptation
• Automated proof synthesis for propositional logic with deep neural networks
• Embodied BERT: A Transformer Model for Embodied, Language-guided Visual Task Completion
• Reinforced Iterative Knowledge Distillation for Cross-Lingual Named Entity Recognition
• Cutting-off Redundant Repeating Generations for Neural Abstractive Summarization
• Sequence to Multi-Sequence Learning via Conditional Chain Mapping for Mixture Signals
• Text Generation with Exemplar-based Adaptive Decoding
• Open Domain Web Keyphrase Extraction Beyond Language Modeling
• Fast Interleaved Bidirectional Sequence Generation
• Rethinking Dialogue State Tracking with Reasoning
• $A^{4}NT$: Author Attribute Anonymity by Adversarial Training of Neural Machine Translation
• Translating Translationese: A Two-Step Approach to Unsupervised Machine Translation
• A Brief Survey of Multilingual Neural Machine Translation
• Learning from Learning Machines: Optimisation, Rules, and Social Norms
• TauRieL: Targeting Traveling Salesman Problem with a deep reinforcement learning inspired architecture
• SGD-QA: Fast Schema-Guided Dialogue State Tracking for Unseen Services
• Hierarchical Transformer Network for Utterance-level Emotion Recognition
• Word-based Domain Adaptation for Neural Machine Translation
• Boosting Active Learning for Speech Recognition with Noisy Pseudo-labeled Samples
• Skin disease diagnosis with deep learning: a review
• Positional Attention-based Frame Identification with BERT: A Deep Learning Approach to Target Disambiguation and Semantic Frame Selection
• LOCAL: Low-Complex Mapping Algorithm for Spatial DNN Accelerators
• Regressive Ensemble for Machine Translation Quality Evaluation
• Data-Free Knowledge Distillation with Soft Targeted Transfer Set Synthesis
• A General Model for Aggregating Annotations Across Simple, Complex, and Multi-Object Annotation Tasks
• Multilingual Neural Machine Translation for Zero-Resource Languages
• A Small Claims Court for the NLP: Judging Legal Text Classification Strategies With Small Datasets
• Anchor & Transform: Learning Sparse Embeddings for Large Vocabularies
• Learning Efficient Lexically-Constrained Neural Machine Translation with External Memory
• Attaining the Unattainable? Reassessing Claims of Human Parity in Neural Machine Translation
• Learning a Multi-Domain Curriculum for Neural Machine Translation
• Capsule-Transformer for Neural Machine Translation
• Upcycle Your OCR: Reusing OCRs for Post-OCR Text Correction in Romanised Sanskrit
• Learning from Imperfect Annotations
• Multi-stage Pretraining for Abstractive Summarization
• Latent Topic Conversational Models
• Has Machine Translation Achieved Human Parity? A Case for Document-level Evaluation
• A Runtime-Based Computational Performance Predictor for Deep Neural Network Training
• Log-based Anomaly Detection Without Log Parsing
• Neural network approach to classifying alarming student responses to online assessment
• Attention Forcing for Sequence-to-sequence Model Training
• Enriching Non-Autoregressive Transformer with Syntactic and SemanticStructures for Neural Machine Translation
• PiSLTRc: Position-informed Sign Language Transformer with Content-aware Convolution
• Multi-Domain Neural Machine Translation with Word-Level Adaptive Layer-wise Domain Mixing
• Cseq2seq: Cyclic Sequence-to-Sequence Learning
• Dynamic Sparse Graph for Efficient Deep Learning
• Improving Pre-Trained Multilingual Models with Vocabulary Expansion
• Can NMT Understand Me? Towards Perturbation-based Evaluation of NMT Models for Code Generation
• Neural CRF Model for Sentence Alignment in Text Simplification
• Extractive Summary as Discrete Latent Variables
• A Practical Guide to Studying Emergent Communication through Grounded Language Games
• Low-Shot Classification: A Comparison of Classical and Deep Transfer Machine Learning Approaches
• Seq2Seq RNN based Gait Anomaly Detection from Smartphone Acquired Multimodal Motion Data
• Robust Reading Comprehension with Linguistic Constraints via Posterior Regularization
• HufuNet: Embedding the Left Piece as Watermark and Keeping the Right Piece for Ownership Verification in Deep Neural Networks
• Testing Untestable Neural Machine Translation: An Industrial Case
• On-the-Fly Syntax Highlighting using Neural Networks
• Exploring the importance of context and embeddings in neural NER models for task-oriented dialogue systems
• Single-Read Reconstruction for DNA Data Storage Using Transformers
• Hybrid CTC-Attention based End-to-End Speech Recognition using Subword Units
• IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization
• AR: Auto-Repair the Synthetic Data for Neural Machine Translation
• Teaching Temporal Logics to Neural Networks
• Neural Open Information Extraction
• A Convolutional Neural Network for Language-Agnostic Source Code Summarization
• Dense Information Flow for Neural Machine Translation
• Learning Language Specific Sub-network for Multilingual Machine Translation
• AdvPicker: Effectively Leveraging Unlabeled Data via Adversarial Discriminator for Cross-Lingual NER
• Advancing CTC-CRF Based End-to-End Speech Recognition with Wordpieces and Conformers
• Testing pre-trained Transformer models for Lithuanian news clustering
• CapWAP: Captioning with a Purpose
• Vacaspati: A Diverse Corpus of Bangla Literature
• Is In-hospital Meta-information Useful for Abstractive Discharge Summary Generation?
• Contextual embedding and model weighting by fusing domain knowledge on Biomedical Question Answering
• Conversational Question Answering: A Survey
• Bridging Subword Gaps in Pretrain-Finetune Paradigm for Natural Language Generation
• On the Privacy Risks of Deploying Recurrent Neural Networks in Machine Learning Models
• Sub-word Level Lip Reading With Visual Attention
• Classifying Long Clinical Documents with Pre-trained Transformers
• Backretrieval: An Image-Pivoted Evaluation Metric for Cross-Lingual Text Representations Without Parallel Corpora
• Tensor Processing Primitives: A Programming Abstraction for Efficiency and Portability in Deep Learning & HPC Workloads
• UzBERT: pretraining a BERT model for Uzbek
• Compressing LSTM Networks by Matrix Product Operators
• Document Graph for Neural Machine Translation
• A little goes a long way: Improving toxic language classification despite data scarcity
• Token-level Adaptive Training for Neural Machine Translation
• Learning to Represent Image and Text with Denotation Graph
• ST-BERT: Cross-modal Language Model Pre-training For End-to-end Spoken Language Understanding
• Noisy Text Data: Achilles' Heel of popular transformer based NLP models
• Speeding up Deep Model Training by Sharing Weights and Then Unsharing
• Text Data Augmentation: Towards better detection of spear-phishing emails
• PolyDL: Polyhedral Optimizations for Creation of High Performance DL primitives
• Greedy Optimization Provably Wins the Lottery: Logarithmic Number of Winning Tickets is Enough
• Adding A Filter Based on The Discriminator to Improve Unconditional Text Generation
• The Mathematical Foundations of Manifold Learning
• Evaluating Amharic Machine Translation
• TTPP: Temporal Transformer with Progressive Prediction for Efficient Action Anticipation
• Neural Text Generation with Artificial Negative Examples
• Explaining Natural Language Processing Classifiers with Occlusion and Language Modeling
• Contextualized Rewriting for Text Summarization
• What's in a Name? -- Gender Classification of Names with Character Based Machine Learning Models
• Revisiting Language Encoding in Learning Multilingual Representations
• Search to Distill: Pearls are Everywhere but not the Eyes
• Transfer Learning Robustness in Multi-Class Categorization by Fine-Tuning Pre-Trained Contextualized Language Models
• Memory-Augmented Neural Networks for Machine Translation
• Event Detection as Question Answering with Entity Information
• AdsGNN: Behavior-Graph Augmented Relevance Modeling in Sponsored Search
• K-TanH: Efficient TanH For Deep Learning
• An Empirical Accuracy Law for Sequential Machine Translation: the Case of Google Translate
• Efficient Bidirectional Neural Machine Translation
• "The Boating Store Had Its Best Sail Ever": Pronunciation-attentive Contextualized Pun Recognition
• The TALP-UPC System for the WMT Similar Language Task: Statistical vs Neural Machine Translation
• The relational processing limits of classic and contemporary neural network models of language processing
• Daydream: Accurately Estimating the Efficacy of Optimizations for DNN Training
• PERL: Pivot-based Domain Adaptation for Pre-trained Deep Contextualized Embedding Models
• Transferring Monolingual Model to Low-Resource Language: The Case of Tigrinya
• Alleviating the Burden of Labeling: Sentence Generation by Attention Branch Encoder-Decoder Network
• Semi-Supervised Learning with Data Augmentation for End-to-End ASR
• Accelerating Large Scale Knowledge Distillation via Dynamic Importance Sampling
• Assessing Demographic Bias in Named Entity Recognition
• Advancing Acoustic-to-Word CTC Model with Attention and Mixed-Units
• DTMT: A Novel Deep Transition Architecture for Neural Machine Translation
• Recurrent Neural Network-based Model for Accelerated Trajectory Analysis in AIMD Simulations
• A Survey of Techniques All Classifiers Can Learn from Deep Networks: Models, Optimizations, and Regularization
• Unified recurrent neural network for many feature types
• Zero-shot Dependency Parsing with Pre-trained Multilingual Sentence Representations
• Fast and Simple Mixture of Softmaxes with BPE and Hybrid-LightRNN for Language Generation
• BanditRank: Learning to Rank Using Contextual Bandits
• Time to Take Emoji Seriously: They Vastly Improve Casual Conversational Models
• Improving Bidirectional Decoding with Dynamic Target Semantics in Neural Machine Translation
• SubCharacter Chinese-English Neural Machine Translation with Wubi encoding
• Make (Nearly) Every Neural Network Better: Generating Neural Network Ensembles by Weight Parameter Resampling
• Artificial Quantum Neural Network: quantum neurons, logical elements and tests of convolutional nets
• "Bilingual Expert" Can Find Translation Errors
• Stage-based Hyper-parameter Optimization for Deep Learning
• Explicit Sentence Compression for Neural Machine Translation
• Style Obfuscation by Invariance
• Simplifying Sentences with Sequence to Sequence Models
• Dynamic Sentence Sampling for Efficient Training of Neural Machine Translation
• MANet: Multimodal Attention Network based Point- View fusion for 3D Shape Recognition
• Efficient Contextual Representation Learning Without Softmax Layer
• Multi-Instance Learning for End-to-End Knowledge Base Question Answering
• Improving Human Text Comprehension through Semi-Markov CRF-based Neural Section Title Generation
• EPNAS: Efficient Progressive Neural Architecture Search
• Toward Computation and Memory Efficient Neural Network Acoustic Models with Binary Weights and Activations
• Automated Description Generation for Software Patches
• Predictive Authoring for Brazilian Portuguese Augmentative and Alternative Communication
• Conceptualizing Suicidal Behavior: Utilizing Explanations of Predicted Outcomes to Analyze Longitudinal Social Media Data
• Sinhala-English Parallel Word Dictionary Dataset
• End-to-End Entity Detection with Proposer and Regressor
• Multilingual Word Embeddings using Multigraphs
• Residual Error: a New Performance Measure for Adversarial Robustness
• Neural Machine Translation with Explicit Phrase Alignment
• Iterative Domain-Repaired Back-Translation
• Finding Better Subword Segmentation for Neural Machine Translation
• Attention-based sequence-to-sequence model for speech recognition: development of state-of-the-art system on LibriSpeech and its application to non-native English
• Learning Light-Weight Translation Models from Deep Transformer
• Using Linguistic Typology to Enrich Multilingual Lexicons: the Case of Lexical Gaps in Kinship
• Reasoning over Vision and Language: Exploring the Benefits of Supplemental Knowledge
• Structure-Invariant Testing for Machine Translation
• Packaging and Sharing Machine Learning Models via the Acumos AI Open Platform
• Patient Cohort Retrieval using Transformer Language Models
• Neural Machine Translation: A Review and Survey
• Traditional IR rivals neural models on the MS MARCO Document Ranking Leaderboard
• Addressing the Vulnerability of NMT in Input Perturbations
• Attention-Based End-to-End Speech Recognition on Voice Search
• Margin-Based Regularization and Selective Sampling in Deep Neural Networks
• Generating Classical Chinese Poems from Vernacular Chinese
• Semantic Parsing with Dual Learning
• Entity Projection via Machine Translation for Cross-Lingual NER
• On the Importance of Subword Information for Morphological Tasks in Truly Low-Resource Languages
• A Unified Framework of Deep Neural Networks by Capsules
• Architecture for a multilingual Wikipedia
• Language Tags Matter for Zero-Shot Neural Machine Translation
• Boosting Span-based Joint Entity and Relation Extraction via Squence Tagging Mechanism
• CancerBERT: a BERT model for Extracting Breast Cancer Phenotypes from Electronic Health Records
• High-performance stock index trading: making effective use of a deep LSTM neural network
• A spelling correction model for end-to-end speech recognition
• A Feasible Framework for Arbitrary-Shaped Scene Text Recognition
• On Learning the Transformer Kernel
• HEAR: Hearing Enhanced Audio Response for Video-grounded Dialogue
• Modeling Fine-Grained Entity Types with Box Embeddings
• Shatter: An Efficient Transformer Encoder with Single-Headed Self-Attention and Relative Sequence Partitioning
• Understanding Learning Dynamics for Neural Machine Translation
• Lessons Learned from Applying off-the-shelf BERT: There is no Silver Bullet
• Exploring Fluent Query Reformulations with Text-to-Text Transformers and Reinforcement Learning
• Sequence Model with Self-Adaptive Sliding Window for Efficient Spoken Document Segmentation
• ModiPick: SLA-aware Accuracy Optimization For Mobile Deep Inference
• A Hybrid Word-Character Approach to Abstractive Summarization
• Video Caption Dataset for Describing Human Actions in Japanese
• Collaborative Training of GANs in Continuous and Discrete Spaces for Text Generation
• Evaluating Neural Word Embeddings for Sanskrit
• Enhanced Neural Machine Translation by Learning from Draft
• Attention Forcing for Machine Translation
• Adversarial Joint Training with Self-Attention Mechanism for Robust End-to-End Speech Recognition
• Sequence-Level Training for Non-Autoregressive Neural Machine Translation
• Towards one-shot learning for rare-word translation with external experts
• Being-ahead: Benchmarking and Exploring Accelerators for Hardware-Efficient AI Deployment
• Improving Neural Machine Translation by Bidirectional Training
• SignBERT: Pre-Training of Hand-Model-Aware Representation for Sign Language Recognition
• Vulnerability Under Adversarial Machine Learning: Bias or Variance?
• Faster, Simpler and More Accurate Hybrid ASR Systems Using Wordpieces
• Investigation of BERT Model on Biomedical Relation Extraction Based on Revised Fine-tuning Mechanism
• Sentiment-based Candidate Selection for NMT
• Understanding Transformers for Bot Detection in Twitter
• AI Powered Compiler Techniques for DL Code Optimization
• Modeling Past and Future for Neural Machine Translation
• GroupBERT: Enhanced Transformer Architecture with Efficient Grouped Structures
• Shallow-to-Deep Training for Neural Machine Translation
• Towards a resource for multilingual lexicons: an MT assisted and human-in-the-loop multilingual parallel corpus with multi-word expression annotation
• Masked ELMo: An evolution of ELMo towards fully contextual RNN language models
• PAIR: Planning and Iterative Refinement in Pre-trained Transformers for Long Text Generation
• How Do Source-side Monolingual Word Embeddings Impact Neural Machine Translation?
• GraphCode2Vec: Generic Code Embedding via Lexical and Program Dependence Analyses
• Domain Agnostic Real-Valued Specificity Prediction
• Sentence Bag Graph Formulation for Biomedical Distant Supervision Relation Extraction
• Classifying the Cosmic-Ray Proton and Light Groups on the LHAASO-KM2A Experiment with the Graph Neural Network
• Syntax-Enhanced Neural Machine Translation with Syntax-Aware Word Representations
• Synchronous Bidirectional Neural Machine Translation
• On the Application of Data-Driven Deep Neural Networks in Linear and Nonlinear Structural Dynamics
• Reinforced Generative Adversarial Network for Abstractive Text Summarization
• Simplifying Neural Machine Translation with Addition-Subtraction Twin-Gated Recurrent Networks
• Model Interpretability and Rationale Extraction by Input Mask Optimization
• Constrained Language Models Yield Few-Shot Semantic Parsers
• Human-centric Metric for Accelerating Pathology Reports Annotation
• Hippo: Taming Hyper-parameter Optimization of Deep Learning with Stage Trees
• Improving Text Generation with Student-Forcing Optimal Transport
• Sequence-to-Sequence ASR Optimization via Reinforcement Learning
• Adversarial Subword Regularization for Robust Neural Machine Translation
• UIBert: Learning Generic Multimodal Representations for UI Understanding
• Paraphrases as Foreign Languages in Multilingual Neural Machine Translation
• Using Interlinear Glosses as Pivot in Low-Resource Multilingual Machine Translation
• Adversarial Sub-sequence for Text Generation
• Efficient Estimation of Influence of a Training Instance
• ActBERT: Learning Global-Local Video-Text Representations
• Code Generation from Natural Language with Less Prior and More Monolingual Data
• Enhancing Pre-trained Chinese Character Representation with Word-aligned Attention
• Aligned Image-Word Representations Improve Inductive Transfer Across Vision-Language Tasks
• Language Transfer for Early Warning of Epidemics from Social Media
• Smoothing and Shrinking the Sparse Seq2Seq Search Space
• Variational Knowledge Distillation for Disease Classification in Chest X-Rays
• RLTIR: Activity-based Interactive Person Identification based on Reinforcement Learning Tree
• Goal-driven text descriptions for images
• BERT-JAM: Boosting BERT-Enhanced Neural Machine Translation with Joint Attention
• In-training Matrix Factorization for Parameter-frugal Neural Machine Translation
• Understanding Natural Language Instructions for Fetching Daily Objects Using GAN-Based Multimodal Target-Source Classification
• A Sketch-Based Neural Model for Generating Commit Messages from Diffs
• BERT-based Chinese Text Classification for Emergency Domain with a Novel Loss Function
• Investigating Label Bias in Beam Search for Open-ended Text Generation
• Depth Growing for Neural Machine Translation
• Learning a Reinforced Agent for Flexible Exposure Bracketing Selection
• Deceptive Deletions for Protecting Withdrawn Posts on Social Platforms
• Uncertainty-Aware Semantic Augmentation for Neural Machine Translation
• SLAM-Inspired Simultaneous Contextualization and Interpreting for Incremental Conversation Sentences
• Adversarial Self-Supervised Data-Free Distillation for Text Classification
• Modeling Coreference Relations in Visual Dialog
• Accelerated Reinforcement Learning for Sentence Generation by Vocabulary Prediction
• Recurrent Neural Network from Adder's Perspective: Carry-lookahead RNN
• GWLAN: General Word-Level AutocompletioN for Computer-Aided Translation
• Learning Architectures from an Extended Search Space for Language Modeling
• Vanishing Nodes: Another Phenomenon That Makes Training Deep Neural Networks Difficult
• An Intrinsic Nearest Neighbor Analysis of Neural Machine Translation Architectures
• Understand customer reviews with less data and in short time: pretrained language representation and active learning
• Towards Accurate Quantization and Pruning via Data-free Knowledge Transfer
• Seq-SG2SL: Inferring Semantic Layout from Scene Graph Through Sequence to Sequence Learning
• Pretrained Language Models for Dialogue Generation with Multiple Input Sources
• Correction of Faulty Background Knowledge based on Condition Aware and Revise Transformer for Question Answering
• Dual Inference for Improving Language Understanding and Generation
• Distributional Discrepancy: A Metric for Unconditional Text Generation
• ML Based Lineage in Databases
• Sensor Transformation Attention Networks
• Non-asymptotic Excess Risk Bounds for Classification with Deep Convolutional Neural Networks
• Depth-Adaptive Neural Networks from the Optimal Control viewpoint
• Integration of TensorFlow based Acoustic Model with Kaldi WFST Decoder
• A First Look at Mobile Intelligence: Architecture, Experimentation and Challenges
• SUPP.AI: Finding Evidence for Supplement-Drug Interactions
• Explainable Semantic Space by Grounding Language to Vision with Cross-Modal Contrastive Learning
• Bayesian Attention Modules
• Exploring Pre-training with Alignments for RNN Transducer based End-to-End Speech Recognition
• Indirectly Supervised English Sentence Break Prediction Using Paragraph Break Probability Estimates
• Stay On-Topic: Generating Context-specific Fake Restaurant Reviews
• Stronger Transformers for Neural Multi-Hop Question Generation
• Convolutional Attention-based Seq2Seq Neural Network for End-to-End ASR
• Adma: A Flexible Loss Function for Neural Networks
• Search Spaces for Neural Model Training
• An Optimized and Energy-Efficient Parallel Implementation of Non-Iteratively Trained Recurrent Neural Networks
• Automatic Classification of Human Translation and Machine Translation: A Study from the Perspective of Lexical Diversity
• Weak Supervision helps Emergence of Word-Object Alignment and improves Vision-Language Tasks
• DeepTitle -- Leveraging BERT to generate Search Engine Optimized Headlines
• Jointly Encoding Word Confusion Network and Dialogue Context with BERT for Spoken Language Understanding
• Amharic-Arabic Neural Machine Translation
• On the Downstream Performance of Compressed Word Embeddings
• On the Copying Behaviors of Pre-Training for Neural Machine Translation
• Word, Subword or Character? An Empirical Study of Granularity in Chinese-English NMT
• MOOCRep: A Unified Pre-trained Embedding of MOOC Entities
• FastSeq: Make Sequence Generation Faster
• Global-aware Beam Search for Neural Abstractive Summarization
• A Comprehensive Study on Temporal Modeling for Online Action Detection
• Learning Neural Models for Natural Language Processing in the Face of Distributional Shift
• SQWA: Stochastic Quantized Weight Averaging for Improving the Generalization Capability of Low-Precision Deep Neural Networks
• Aligning the Pretraining and Finetuning Objectives of Language Models
• Improved Speech Representations with Multi-Target Autoregressive Predictive Coding
• Knowledge Efficient Deep Learning for Natural Language Processing
• Deep Learning: Our Miraculous Year 1990-1991
• Efficiently Reusing Old Models Across Languages via Transfer Learning
• DCoM: A Deep Column Mapper for Semantic Data Type Detection
• Off-the-Shelf Unsupervised NMT
• Sketching Transformed Matrices with Applications to Natural Language Processing
• DynamicEmbedding: Extending TensorFlow for Colossal-Scale Applications
• Largest Eigenvalues of the Conjugate Kernel of Single-Layered Neural Networks
• Bayesian Sparsification of Gated Recurrent Neural Networks
• Massively Parallel Cross-Lingual Learning in Low-Resource Target Language Translation
• Sampling-free Uncertainty Estimation in Gated Recurrent Units with Exponential Families
• Taylor's law for Human Linguistic Sequences
• Optimizing Sponsored Search Ranking Strategy by Deep Reinforcement Learning
• Memory Slices: A Modular Building Block for Scalable, Intelligent Memory Systems
• PairConnect: A Compute-Efficient MLP Alternative to Attention
• Interpretable Self-supervised Multi-task Learning for COVID-19 Information Retrieval and Extraction
• Synergy: Resource Sensitive DNN Scheduling in Multi-Tenant Clusters
• Artificial Intelligence : from Research to Application ; the Upper-Rhine Artificial Intelligence Symposium (UR-AI 2019)
• Guiding Teacher Forcing with Seer Forcing for Neural Machine Translation
• CytonMT: an Efficient Neural Machine Translation Open-source Toolkit Implemented in C++
• Lattice Rescoring Strategies for Long Short Term Memory Language Models in Speech Recognition
• Bootstrapped Q-learning with Context Relevant Observation Pruning to Generalize in Text-based Games
• Breaking the Memory Wall for AI Chip with a New Dimension
• To Beam Or Not To Beam: That is a Question of Cooperation for Language GANs
• A Multilingual Modeling Method for Span-Extraction Reading Comprehension
• Learning by Fixing: Solving Math Word Problems with Weak Supervision
• MergeDistill: Merging Pre-trained Language Models using Distillation
• Analyzing Architectures for Neural Machine Translation Using Low Computational Resources
• Language-Independent Representor for Neural Machine Translation
• Machine Translation between Vietnamese and English: an Empirical Study
• A Lightweight Recurrent Network for Sequence Modeling
• Dynamically Composing Domain-Data Selection with Clean-Data Selection by "Co-Curricular Learning" for Neural Machine Translation
• A Semi-Supervised Approach for Low-Resourced Text Generation
• Defending Democracy: Using Deep Learning to Identify and Prevent Misinformation
• Towards Fully Automated Manga Translation
• Phases of learning dynamics in artificial neural networks: with or without mislabeled data
• Multilingual Neural RST Discourse Parsing
• BERT_SE: A Pre-trained Language Representation Model for Software Engineering
• Applying SoftTriple Loss for Supervised Language Model Fine Tuning
• GLOW : Global Weighted Self-Attention Network for Web Search
• BERT might be Overkill: A Tiny but Effective Biomedical Entity Linker based on Residual Convolutional Neural Networks
• Building Compact and Robust Deep Neural Networks with Toeplitz Matrices
• Crowdsourcing Parallel Corpus for English-Oromo Neural Machine Translation using Community Engagement Platform
• Fast and Accurate Deep Bidirectional Language Representations for Unsupervised Learning
• Detection and Prediction of Nutrient Deficiency Stress using Longitudinal Aerial Imagery
• Automatic Dialogic Instruction Detection for K-12 Online One-on-one Classes
• Experiments with Rich Regime Training for Deep Learning
• AGenT Zero: Zero-shot Automatic Multiple-Choice Question Generation for Skill Assessments
• UPB at SemEval-2020 Task 12: Multilingual Offensive Language Detection on Social Media by Fine-tuning a Variety of BERT-based Models
• Long Short-Term Memory Neuron Equalizer
• Self-Attentional Models Application in Task-Oriented Dialogue Generation Systems
• Testing Machine Translation via Referential Transparency
• Human Languages in Source Code: Auto-Translation for Localized Instruction
• How Furiously Can Colourless Green Ideas Sleep? Sentence Acceptability in Context
• Phantom: A High-Performance Computational Core for Sparse Convolutional Neural Networks
• An In-depth Walkthrough on Evolution of Neural Machine Translation
• Subword Language Model for Query Auto-Completion
• Modeling Coverage for Non-Autoregressive Neural Machine Translation
• Reducing Data Motion to Accelerate the Training of Deep Neural Networks
• Interpretable Entity Representations through Large-Scale Typing
• Explaining Documents' Relevance to Search Queries
• Conversational Response Re-ranking Based on Event Causality and Role Factored Tensor Event Embedding
• PhoMT: A High-Quality and Large-Scale Benchmark Dataset for Vietnamese-English Machine Translation
• HARE: a Flexible Highlighting Annotator for Ranking and Exploration
• An Empirical Investigation of Multi-bridge Multilingual NMT models
• Reference Network for Neural Machine Translation
• ${\bar{G}}_{mst}$:An Unbiased Stratified Statistic and a Fast Gradient Optimization Algorithm Based on It
• Integrating Categorical Features in End-to-End ASR
• Leveraging Personal Navigation Assistant Systems Using Automated Social Media Traffic Reporting
• The University of Sydney's Machine Translation System for WMT19
• Improving Zero-shot Multilingual Neural Machine Translation for Low-Resource Languages
• Effective Use of Graph Convolution Network and Contextual Sub-Tree forCommodity News Event Extraction
• Learning to Reformulate the Queries on the WEB
• Dynamic Programming Encoding for Subword Segmentation in Neural Machine Translation
• Enhancing Clinical Information Extraction with Transferred Contextual Embeddings
• Parsimonious Morpheme Segmentation with an Application to Enriching Word Embeddings
• Improving Truthfulness of Headline Generation
• Task-adaptive Pre-training of Language Models with Word Embedding Regularization
• Deep Multi-Kernel Convolutional LSTM Networks and an Attention-Based Mechanism for Videos
• TAG : Type Auxiliary Guiding for Code Comment Generation
• Positioning yourself in the maze of Neural Text Generation: A Task-Agnostic Survey
• Learning to Detect Unacceptable Machine Translations for Downstream Tasks
• Neural Mention Detection
• A Massively Multilingual Analysis of Cross-linguality in Shared Embedding Space
• STN4DST: A Scalable Dialogue State Tracking based on Slot Tagging Navigation
• Cross-Modality Relevance for Reasoning on Language and Vision
• Deep Hierarchical Classification for Category Prediction in E-commerce System
• Vector-Vector-Matrix Architecture: A Novel Hardware-Aware Framework for Low-Latency Inference in NLP Applications
• E3: Entailment-driven Extracting and Editing for Conversational Machine Reading
• Application of Video-to-Video Translation Networks to Computational Fluid Dynamics
• Lexically Cohesive Neural Machine Translation with Copy Mechanism
• Improving Scientific Article Visibility by Neural Title Simplification
• Neural Machine Translation for Multilingual Grapheme-to-Phoneme Conversion
• MIRA: Leveraging Multi-Intention Co-click Information in Web-scale Document Retrieval using Deep Neural Networks
• Detecting ESG topics using domain-specific language models and data augmentation approaches
• Sequential Modelling with Applications to Music Recommendation, Fact-Checking, and Speed Reading
• An Emergency Medical Services Clinical Audit System driven by Named Entity Recognition from Deep Learning
• NTT's Machine Translation Systems for WMT19 Robustness Task
• EAT: a simple and versatile semantic representation format for multi-purpose NLP
• ProphetNet-Ads: A Looking Ahead Strategy for Generative Retrieval Models in Sponsored Search Engine
• Towards Fully Bilingual Deep Language Modeling
• Multi-modal Transformer for Video Retrieval
• Morphological Skip-Gram: Using morphological knowledge to improve word representation
• Measuring prominence of scientific work in online news as a proxy for impact
• Style Pooling: Automatic Text Style Obfuscation for Improved Classification Fairness
• An Evaluation Dataset and Strategy for Building Robust Multi-turn Response Selection Model
• Optimizing AD Pruning of Sponsored Search with Reinforcement Learning
• Unsupervised Learning Layers for Video Analysis
• On Multilingual Training of Neural Dependency Parsers
• LfEdNet: A Task-based Day-ahead Load Forecasting Model for Stochastic Economic Dispatch
• Learning to Expand: Reinforced Pseudo-relevance Feedback Selection for Information-seeking Conversations
• Langevin Cooling for Domain Translation
• ARMAN: Pre-training with Semantically Selecting and Reordering of Sentences for Persian Abstractive Summarization
• Duluth at SemEval-2020 Task 7: Using Surprise as a Key to Unlock Humorous Headlines
• A Closer Look at Few-Shot Crosslingual Transfer: The Choice of Shots Matters
• Cross-Modal Alignment with Mixture Experts Neural Network for Intral-City Retail Recommendation
• Deep Collective Learning: Learning Optimal Inputs and Weights Jointly in Deep Neural Networks
• How LSTM Encodes Syntax: Exploring Context Vectors and Semi-Quantization on Natural Text
• Improving End-to-end Speech Recognition with Pronunciation-assisted Sub-word Modeling
• DeepZensols: Deep Natural Language Processing Framework
• An Empirical Study of Mini-Batch Creation Strategies for Neural Machine Translation
• Translation, Sentiment and Voices: A Computational Model to Translate and Analyse Voices from Real-Time Video Calling
• Learning Energy-Based Approximate Inference Networks for Structured Applications in NLP
• Speeding Up Neural Machine Translation Decoding by Cube Pruning
• Logographic Subword Model for Neural Machine Translation
• Evaluating Syntactic Properties of Seq2seq Output with a Broad Coverage HPSG: A Case Study on Machine Translation
• Can the Transformer Be Used as a Drop-in Replacement for RNNs in Text-Generating GANs?
• Recurrent multiple shared layers in Depth for Neural Machine Translation
• Neural Sequence Model Training via $\alpha$-divergence Minimization
• Approximate Distribution Matching for Sequence-to-Sequence Learning
• Future-Prediction-Based Model for Neural Machine Translation
• Predicting protein secondary structure with Neural Machine Translation
• Deploying a BERT-based Query-Title Relevance Classifier in a Production System: a View from the Trenches
• BIG MOOD: Relating Transformers to Explicit Commonsense Knowledge
• Identifying and Exploiting Structures for Reliable Deep Learning
• An Empirical Study of Adequate Vision Span for Attention-Based Neural Machine Translation
• An Empirical Comparison of Syllabuses for Curriculum Learning
• Attentive Sequence to Sequence Translation for Localizing Clips of Interest by Natural Language Descriptions
• Neural Task Planning with And-Or Graph Representations
• Searching Personal Collections
• Automatic Post-Editing for Vietnamese
• Single-Queue Decoding for Neural Machine Translation
• Machine Translation in Pronunciation Space
• Learning to Represent Bilingual Dictionaries
• Dialogue Summarization with Supporting Utterance Flow Modeling and Fact Regularization
• R-grams: Unsupervised Learning of Semantic Units in Natural Language
• Training Neural Machine Translation (NMT) Models using Tensor Train Decomposition on TensorFlow (T3F)
• Learning Representations by Stochastic Meta-Gradient Descent in Neural Networks
• The Helsinki Neural Machine Translation System
• Learning to Copy for Automatic Post-Editing
• Stylized Text Generation Using Wasserstein Autoencoders with a Mixture of Gaussian Prior
• Subspace Approximation for Approximate Nearest Neighbor Search in NLP
• Generating Titles for Web Tables
• A crossover code for high-dimensional composition
• Robust, Deep, and Reinforcement Learning for Management of Communication and Power Networks
• The Cross-Lingual Arabic Information REtrieval (CLAIRE) System
• Full-Sentence Models Perform Better in Simultaneous Translation Using the Information Enhanced Decoding Strategy
• A cost-benefit analysis of cross-lingual transfer methods
• Using CollGram to Compare Formulaic Language in Human and Neural Machine Translation
• The Detection of Distributional Discrepancy for Text Generation
• Morphological and Language-Agnostic Word Segmentation for NMT
• Pseudo Siamese Network for Few-shot Intent Generation
• Iterative Dual Domain Adaptation for Neural Machine Translation
• Graph-based Filtering of Out-of-Vocabulary Words for Encoder-Decoder Models
• Shareable Representations for Search Query Understanding
• Residual Tree Aggregation of Layers for Neural Machine Translation
• Size Matters: Foundation Model for Czech HTML documents
• Bi-Decoder Augmented Network for Neural Machine Translation
• A Topic Guided Pointer-Generator Model for Generating Natural Language Code Summaries
• On the Significance of Question Encoder Sequence Model in the Out-of-Distribution Performance in Visual Question Answering
• Pretrained Transformers for Simple Question Answering over Knowledge Graphs
• Power Law Graph Transformer for Machine Translation and Representation Learning
• Artificial stochastic neural network on the base of double quantum wells
• PolyScientist: Automatic Loop Transformations Combined with Microkernels for Optimization of Deep Learning Primitives
• XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages
• Modeling Diagnostic Label Correlation for Automatic ICD Coding
• Learning to Generate Multiple Style Transfer Outputs for an Input Sentence
• Residual Continual Learning
• All You Can Embed: Natural Language based Vehicle Retrieval with Spatio-Temporal Transformers
• Image to Language Understanding: Captioning approach
• Guider l'attention dans les modeles de sequence a sequence pour la prediction des actes de dialogue
• SYSTRAN Purely Neural MT Engines for WMT2017
• Sentence-wise Smooth Regularization for Sequence to Sequence Learning
• Multi-Task Learning and Adapted Knowledge Models for Emotion-Cause Extraction
• Open-endedness in AI systems, cellular evolution and intellectual discussions
• Automatic Documentation of ICD Codes with Far-Field Speech Recognition
• Algorithm to Compilation Co-design: An Integrated View of Neural Network Sparsity
• Generating Symbolic Reasoning Problems with Transformer GANs
• Large Language Models -- the Future of Fundamental Physics?
• Generating Watermarked Adversarial Texts
• Using Structured Input and Modularity for Improved Learning
• An End-to-End Approach to Automatic Speech Assessment for Cantonese-speaking People with Aphasia
• Learning to Decipher Hate Symbols
• Calculating Question Similarity is Enough: A New Method for KBQA Tasks
• Corpora Generation for Grammatical Error Correction
• IIITT@Dravidian-CodeMix-FIRE2021: Transliterate or translate? Sentiment analysis of code-mixed text in Dravidian languages
• Searching for TrioNet: Combining Convolution with Local and Global Self-Attention
• Pegasus@Dravidian-CodeMix-HASOC2021: Analyzing Social Media Content for Detection of Offensive Text
• The Agnostic Structure of Data Science Methods
• A Corpus of Controlled Opinionated and Knowledgeable Movie Discussions for Training Neural Conversation Models
• Parallel Blockwise Knowledge Distillation for Deep Neural Network Compression
• Dynamic Curriculum Learning for Low-Resource Neural Machine Translation
• Customizing Contextualized Language Models forLegal Document Reviews
• Resurrecting Submodularity for Neural Text Generation
• Accelerating Distributed SGD for Linear Regression using Iterative Pre-Conditioning
• Empirical Evaluation of Deep Learning Model Compression Techniques on the WaveNet Vocoder
• DPRK-BERT: The Supreme Language Model
• Understanding and Enhancing the Use of Context for Machine Translation
• DORB: Dynamically Optimizing Multiple Rewards with Bandits
• Keeping Notes: Conditional Natural Language Generation with a Scratchpad Mechanism
• Adapting MARBERT for Improved Arabic Dialect Identification: Submission to the NADI 2021 Shared Task
• Multilingual AMR-to-Text Generation
• Data Augmentation for Abstractive Query-Focused Multi-Document Summarization
• An empirical analysis of phrase-based and neural machine translation
• NameRec*: Highly Accurate and Fine-grained Person Name Recognition
• Layer-Wise Multi-View Learning for Neural Machine Translation
• Improving OOV Detection and Resolution with External Language Models in Acoustic-to-Word ASR
• Efficient transfer learning for NLP with ELECTRA