Rosanne Liu
Research Scientist at ML Collective
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRosanne Liu is a research scientist at Google DeepMind and co-founder of ML Collective, with nine years of experience advancing machine learning through both industry research and nonprofit community-building. A PhD from Northwestern, she helped found Uber AI and has published at NeurIPS, ICLR, ICML and Science, while her work has been featured in WIRED and MIT Technology Review. Her research spans training dynamics, model capacity, and generative language models, and she contributed engineering work to influential open-source projects like PPLM for controllable text generation. Beyond papers, she organizes a long-running weekly reading group and runs mentorship and symposium programs to broaden participation in AI. She combines deep technical rigor with a persistent focus on democratizing research, diversity, and practical impact across scientific and industrial domains.
9 years of coding experience
12 years of employment as a software developer
Master of Science (M.S.) Machine Learning, Master of Science (M.S.) Machine Learning at University of MichiganBachelor Electrical Engineering, Bachelor Electrical Engineering at Fudan UniversityDoctor of Philosophy (PhD) Computer Science, Doctor of Philosophy (PhD) Computer Science at Northwestern University
English, Chinese
Github Skills (9)
transformers10
machine-learning10
deep-learning10
python10
natural-language-processing10
language-modeling10
nlp9
pytorch9
github7
Programming languages (2)
Github contributions (5)
uber-research/PPLM
Nov 2019 - Apr 2020
Plug and Play Language Model implementation. Allows to steer topic and attributes of GPT-2 models.
Role in this project:
ML Engineer
ML EngineerContributions:18 commits, 1 PR, 6 pushes in 4 months
Contributions summary:Rosanne appears to be contributing to the implementation and maintenance of a Plug and Play Language Model (PPLM). Their commits involve modifications to the `run_pplm.py` file, suggesting a focus on running and potentially experimenting with the model. The changes include updating the codebase to be in sync with the transformer codebase and the inclusion of human annotated data sheets. The user also added the separation of the classification head into its own file and added updated commands.
nlppytorchtransformerslanguage-modelnatural-language-generation
uber-research/CoordConv
Feb 2019 - Jan 2020
Contributions:34 commits, 8 pushes in 11 months
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.