I work on the theory and algorithms for machine learning. I am interested in
- deep learning theory
- optimization
- statistical learning
Find my papers on arXiv and Google Scholar.
I am a postdoc fellow at the Simons Institute at UC Berkeley hosted by Peter Bartlett and Bin Yu. I am a part of the NSF/Simons Collaboration on the Theoretical Foundations of Deep Learning. I received my Ph.D. in Computer Science at Johns Hopkins University, advised by Vladimir Braverman.
Contact me via Email.
Jingfeng Wu is a postdoctoral fellow at the Simons Institute for the Theory of Computing at UC Berkeley. His research focuses on deep learning theory, optimization, and statistical learning. He earned his Ph.D. in Computer Science from Johns Hopkins University in 2023. Prior to that, he received a B.S. in Mathematics (2016) and an M.S. in Applied Mathematics (2019), both from Peking University. In 2023, he was recognized as a Rising Star in Data Science by the University of Chicago and UC San Diego.
I work on the theory and algorithms for machine learning. I am interested in
Find my papers on arXiv and Google Scholar.
Large Stepsizes Accelerate Gradient
Descent for Regularized Logistic Regression
arXiv 2025
Minimax Optimal Convergence of Gradient
Descent in Logistic Regression via Large and Adaptive Stepsizes
ICML 2025 | poster
Large Stepsize Gradient Descent for
Logistic Loss: Non-Monotonicity of the Loss Improves Optimization Efficiency
COLT 2024 |
poster
Implicit Bias of Gradient Descent for
Logistic Regression at the Edge of Stability
NeurIPS 2023 (spotlight) |
poster
Benefits of Early Stopping in Gradient
Descent for Overparameterized Logistic Regression
ICML 2025 |
poster
Last Iterate Risk Bounds of SGD with
Decaying Stepsize for Overparameterized Linear Regression
ICML 2022 (long presentation) |
poster
Benign Overfitting of Constant-Stepsize SGD for Linear Regression
COLT 2021 (journal version in JMLR 2023)
Scaling Laws in Linear Regression:
Compute, Parameters, and Data
NeurIPS 2024
How Many Pretraining Tasks Are Needed
for
In-Context Learning of Linear Regression?
ICLR 2024 (spotlight) |
poster
The Power and Limitation of Pretraining-Finetuning for Linear Regression under
Covariate Shift
NeurIPS 2022 |
poster
Organizer
Conference Reviewer
ICML (2020 - 2025),
NeurIPS (2020 - 2025),
ICLR (2021 - 2025),
SODA (2026, subreviewer),
AISTATS (2021 - 2023),
UAI (2023),
AAAI (2021 - 2023, PC member reviewer)
Journal Reviewer
JMLR,
TPAMI,
TMLR,
SIMODS,
JAIR,
Applied Probability Journals,
IEEE Transactions on Information
Theory,