Thang Luong is a staff research scientist at Google Brain. His main interest these days is to learn from the vast amount of unlabeled data. His research is multi-disciplinary from search (DeepRank), to vision (UDA, NoisyStudent, MPL) and language (QANet, ELECTRA). He is a co-founder the Meena project (now known as LaMDA) and an inventor of LuongAttention for neural machine translation. Sample works: conversation, semi-supervised learning, and language pretraining. Thang Luong was formerly a PhD candidate from Stanford NLP group. All of his publications can be found at Google Scholar.