Thang Luong
Principal Scientist & Director Of Research at Google
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolThang Luong is a Principal Scientist and Director of Research at Google DeepMind with 16 years of multidisciplinary AI experience spanning search, vision and language. A Stanford NLP PhD (advisor Chris Manning), he co-led high-profile projects such as Meena/LaMDA and helped advance methods like Luong Attention, ELECTRA and NoisyStudent. Thang combines deep theoretical work with hands-on engineering—contributing to flagship open-source efforts like TensorFlow NMT, CoreNLP and tensor2tensor—reflecting a rare blend of production-grade code and foundational research. Based in Palo Alto, he also co-founded VietAI to grow AI capacity in Vietnam, and currently focuses on multimodal learning at scale and questions around superintelligence.
15 years of coding experience
9 years of employment as a software developer
Bachelor of Computing Computer Science, Bachelor of Computing Computer Science at National University of SingaporeDiploma Mathematics, Diploma Mathematics at VNU-HCM High School for the GiftedDoctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Stanford University
Github Skills (28)
machine-translation10
python10
model-driven10
machine-learning10
java10
ml10
mle10
model-building10
transformer-models10
javas10
deep-learning10
tensorflow10
natural-language-processing10
neural-network10
nlp10
Programming languages (5)
Github contributions (5)
tensorflow/nmt
Jun 2017 - Feb 2019
TensorFlow Neural Machine Translation Tutorial
Role in this project:
Back-end Developer
Back-end DeveloperContributions:60 commits, 2 PRs, 2 pushes in 1 year 8 months
Contributions summary:Thang primarily focused on improving the TensorFlow Neural Machine Translation (NMT) tutorial by cleaning and improving code. The commits involved modifications to training scripts (`nmt/train.py`) with enhancements to internal/external evaluation and output logging. Further contributions included enhancements to inference scripts and test files (`nmt/inference.py`, `nmt/inference_test.py`) with flags for specifying input and output files. The user also refactored model creation code.
seq2seqmachine-translationtranslationnmttensorflow
stanfordnlp/CoreNLP
Sep 2013 - Nov 2014
CoreNLP: A Java suite of core NLP tools for tokenization, sentence segmentation, NER, parsing, coreference, sentiment analysis, etc.
Role in this project:Back-end Developer
Back-end DeveloperContributions:267 commits in 1 year 2 months
Contributions summary:Thang's commits focused on implementing the Arabic Feature Factory, indicating a strong involvement in the back-end development of the CoreNLP project. The commits show code changes in the ArabicFeatureFactory.java file, demonstrating work on tokenization, sentence segmentation, named entity recognition, and sentiment analysis. This includes the integration of Arabic-specific word lists and processing of morphological features within the context of Arabic language processing.
word-embeddingssentiment-analysisnatural-language-processingstanford-nlpsuite
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.