No 1. 《Learning-to-Rank with BERT in TF-Ranking》
    No 2. 【让人人都变得“彬彬有礼”：礼貌迁移任务——在保留意义的同时将非礼貌语句转换为礼貌语句，提供包含1.39M + 实例的数据集】
    No 3. 《TLDR: Extreme Summarization of Scientific Documents》
    No 4. 《Lite Transformer with Long-Short Range Attention》
    No 5. 《Attention Module is Not Only a Weight: Analyzing Transformers with Vector Norms》
    No 6. 《ToTTo: A Controlled Table-To-Text Generation Dataset》
    No 7. 《Light-Weighted CNN for Text Classification》
    No 8. 《Named Entity Recognition without Labelled Data: A Weak Supervision Approach》
    No 9. 《Pyramid Attention Networks for Image Restoration》
    No 10. 《CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualization》
    No 11. 【ByteSing：效果逼真的中文歌唱合成系统】
    No 12. 《Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels》
    No 13. 《NAIST COVID: Multilingual COVID-19 Twitter and Weibo Dataset》
    No 14. 《CoronaVis: A Real-time COVID-19 Tweets Analyzer》
    No 15. 《VGGSound: A Large-scale Audio-Visual Dataset》
    No 16. 《Reinforcement Learning with Augmented Data》
    No 17. 《Neural Additive Models: Interpretable Machine Learning with Neural Nets》
    No 18. 《Knowledge Graph Embeddings and Explainable AI》
    No 19. 《Why should we add early exits to neural networks?》
    No 20. 《Decoupling Global and Local Representations from/for Image Generation》
    No 21. 《Improved Residual Networks for Image and Video Recognition》
    No 22. 《Deep Learning for Screening COVID-19 using Chest X-Ray Images》
    No 23. 《One-Shot Identity-Preserving Portrait Reenactment》
    No 24. 《The Creation and Detection of Deepfakes: A Survey》
    No 25. 《DGL-KE: Training Knowledge Graph Embeddings at Scale》
    No 26. 《Explainable Deep Learning: A Field Guide for the Uninitiated》
    No 27. 《SIGN: Scalable Inception Graph Neural Networks》
    No 28. 《DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference》
    No 29. 《On the Synergies between Machine Learning and Stereo: a Survey》
    No 30. 《Unsupervised Real Image Super-Resolution via Generative Variational AutoEncoder》