Techniques

Tensor Programs IV: Feature Learning in Infinite-Width Neural Networks ICML 2021 TenTrans Large-Scale Multilingual Machine Translation System for WMT21 EMNLP 2021 Terra: Imperative-Symbolic Co-Execution of Imperative Deep Learning Programs NeurIPS 2021 The Effect of Pretraining on Extractive Summarization for Scientific Documents NAACL 2021 The Heads Hypothesis: A Unifying Statistical Approach Towards Understanding Multi-Headed Attention in BERT AAAI 2021 The Heterogeneity Hypothesis: Finding Layer-Wise Differentiated Network Architectures CVPR 2021 The Intrinsic Dimension of Images and Its Impact on Learning ICLR 2021 The Laughing Machine: Predicting Humor in Video WACV 2021 The Lottery Tickets Hypothesis for Supervised and Self-Supervised Pre-Training in Computer Vision Models CVPR 2021 The Neural Tangent Link Between CNN Denoisers and Non-Local Filters CVPR 2021 The Power of Scale for Parameter-Efficient Prompt Tuning EMNLP 2021 The Spectrum of Fisher Information of Deep Networks Achieving Dynamical Isometry AISTATS 2021 THOR, Trace-based Hardware-driven Layer-Oriented Natural Gradient Descent Computation AAAI 2021 Three-Module Modeling For End-to-End Spoken Language Understanding Using Pre-Trained DNN-HMM-Based Acoustic-Phonetic Model Interspeech 2021 Three Steps to Multimodal Trajectory Prediction: Modality Clustering, Classification and Synthesis ICCV 2021 Tighter Risk Certificates for Neural Networks JMLR 2021 Token-Aware Virtual Adversarial Training in Natural Language Understanding AAAI 2021 Topic-Aware Multi-turn Dialogue Modeling AAAI 2021 Topic Scene Graph Generation by Attention Distillation From Caption ICCV 2021 Topic Transferable Table Question Answering EMNLP 2021 Topological Attention for Time Series Forecasting NeurIPS 2021 TöRF: Time-of-Flight Radiance Fields for Dynamic Scene View Synthesis NeurIPS 2021 TorontoCL at CMCL 2021 Shared Task: RoBERTa with Multi-Stage Fine-Tuning for Eye-Tracking Prediction NAACL 2021 Towards Better Robust Generalization with Shift Consistency Regularization ICML 2021 Towards Certifying L-infinity Robustness using Neural Networks with L-inf-dist Neurons ICML 2021