Jinzhe Zeng
Assistant Professor at University of Science and Technology of China
Suzhou City, Jiangsu, China
Summary
🤩
Rockstar🎓
Top SchoolJinzhe Zeng is an assistant professor and computational scientist with 11 years of experience at the intersection of chemistry, machine learning, and scientific software engineering. Trained as a PhD researcher in Chemistry at Rutgers, he has driven production-grade contributions to high-impact open-source projects like DeepMD-kit and DP-GEN, improving core training pipelines, descriptor modules, and reproducibility features for deep potential models. He blends back-end C++/Python engineering, ML model design, and DevOps automation—having fixed memory and compression bugs, integrated Gaussian support, and tuned CI/build workflows for conda-forge. Now launching a tenure-track role in AI and Data Science at USTC in Suzhou, he brings hands-on domain knowledge of molecular simulation to bridge academic research and robust tooling. A detail that often goes unnoticed: he has repeatedly focused on reproducibility and model compression, enabling scalable deployment of many-body potential models beyond lab prototypes.
11 years of coding experience
Bachelor of Science, Chemistry, Bachelor of Science, Chemistry at East China Normal UniversityHigh School Diploma, High School Diploma at Nanjing Jinling High SchoolDoctor of Philosophy - PhD, Chemistry and Chemical Biology, Doctor of Philosophy - PhD, Chemistry and Chemical Biology at Rutgers University
Chinese, English
Stackoverflow
63reputation
123kreached
3answers
0questions
Github Skills (31)
conda-forge10
github-ci10
c-language10
python10
machine-learning10
bash10
automation10
deep-learning10
tensorflow10
build-automation10
automations10
cprogramming-language10
githubaction-workflow10
unit-testing9
horovod9
Programming languages (33)
Github contributions (5)
deepmodeling/deepmd-kit
Feb 2019 - Jan 2023
A deep learning package for many-body potential energy representation and molecular dynamics
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:15 releases, 1853 reviews, 560 commits in 4 years
Contributions summary:Jinzhe's contributions primarily focused on implementing changes to the core training components and descriptor modules, which includes the integration of a new option `enable_atom_ener_coeff`. Their work involved modifying the loss functions and integrating type embeddings, indicating a focus on improving model accuracy and flexibility. Furthermore, the user addressed bugs related to model compression and memory allocation in the C++ API and the tests.
lammpspythonipideepmdtensorflow
deepmodeling/dpgen
Jun 2019 - Jan 2023
The deep potential generator to generate a deep-learning based model of interatomic potential energy and force field
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:4 releases, 325 reviews, 251 commits in 3 years 7 months
Contributions summary:Jinzhe significantly contributed to the project by adding support for Gaussian calculations, including implementing functions for input generation and output processing. They also expanded the code to include unit tests for the Gaussian features and implemented the functionality to take clusters for the deep learning models, which improved the model’s performance. Furthermore, the user integrated a mechanism to ensure a random seed for models, ensuring reproducibility, and implemented the ability to compress the model.
model-baseddeep-learningfieldforce-fieldforce
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.