Jinxing Li
Research Scientist · Machine Learning | Space Physics
Research Scientist
Persona Identities
San Francisco, CA, USA
I’m a Research Scientist at Persona Identities, where I build machine-learning systems to fight identity fraud. Before joining industry, I spent a decade as a researcher at UCLA, studying the physics of Earth’s radiation belts and magnetosphere using spacecraft data and large-scale data-driven models.
My research sits at the intersection of machine learning, plasma physics, and AI for real-world systems. I have led NASA- and NSF-funded projects developing neural-network models of the near-Earth space environment, and I have published in journals such as Nature Communications, Geophysical Research Letters, and JGR: Space Physics. More recently, I’ve been working on layout-aware representation learning for open-set fraud detection (CVPR/AI4RWC 2026).
I earned my Ph.D. in Space Physics from Peking University in 2015, with a visiting year at UCLA under Prof. Richard Thorne, and a B.S. in Applied Physics from Xi’an Jiaotong University.
Outside of research, I serve on the boards of several Chinese-American scientific and alumni organizations, chairing scholarship programs that support students in STEM.
news
| Feb 01, 2026 | New paper accepted at the CVPR / AI4RWC 2026 workshop: Layout-Aware Representation Learning for Open-Set ID Fraud Discovery. |
|---|---|
| Nov 01, 2025 | Joined Persona Identities as a Research Scientist, working on machine learning for identity fraud detection. |
| Sep 01, 2025 | I gave a TEDx talk on Sound of Space and Predicting Space Environment with AI |
| Jul 15, 2025 | New paper out in Frontiers in Astronomy and Space Sciences: a comparative study of MLP, CNN, LSTM and Transformer networks for modeling magnetospheric ion distributions. |
| Jun 01, 2024 | Two new papers on chorus spectral gaps and the fine structure of magnetosonic waves published in JGR: Space Physics. |