Jason Wei
Research Scientist at Meta
Hanover, New Hampshire, United States
Summary
🤩
Rockstar🎓
Top SchoolJason Wei is a research scientist with nine years of experience building and investigating large language models and practical NLP techniques across top AI labs, currently at Meta after research roles at OpenAI and Google. His work spans instruction tuning, chain-of-thought prompting, and emergent abilities in LLMs, grounded by hands-on contributions to data augmentation and model engineering (including an EMNLP-presented NLP augmentation repo). Comfortable moving between deep research and production-minded implementation, he has designed CNN models and end-to-end training pipelines while shaping O(1)-level research projects. Based in Hanover, New Hampshire, he blends academic rigor from Dartmouth with elite technical training dating back to Thomas Jefferson High School for Science and Technology. Colleagues describe him as quietly productive—“some dude” on GitHub who delivers influential, reproducible code behind the scenes. Jason brings a mix of empirical curiosity and engineering discipline that accelerates both experimentation and deployment.
9 years of coding experience
4 years of employment as a software developer
Dartmouth CollegeThomas Jefferson High School for Science and Technology
Github Skills (15)
mask-rcnn10
faster-rcnn10
nlp10
data-augmentation10
python10
text-classification10
natural-language-processing10
machine-learning9
nltk9
keras8
tensorflow7
classification7
embedding6
n5
rnn-model5
Programming languages (3)
Github contributions (5)
jasonwei20/eda_nlp
Dec 2018 - May 2021
Data augmentation for NLP, presented at EMNLP 2019
Role in this project:
Data Scientist
Data ScientistContributions:55 commits, 1 PR, 48 pushes in 2 years 5 months
Contributions summary:Jason primarily worked on data augmentation techniques for Natural Language Processing, as indicated by the code changes and repository description. Their main contribution involves implementing and refining various data augmentation methods, including synonym replacement, random deletion, random swap, and random insertion, within the `nlp_aug.py` file. They also contributed to model training and evaluation in `2_train_eval.py` and designed and implemented a CNN model.
data-augmentationnatural-language-processingnlptext-classificationsynonyms
jasonwei20/text-classification-bert
Jul 2020 - Aug 2020
Contributions:21 commits, 20 pushes, 1 branch in 17 days
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.