Liang Zhang

About

I am a second-year Ph.D. student in computer science at the Max Planck ETH Center for Learning Systems, jointly advised by Niao He from ETH Zurich and Michael Muehlebach from the Max Planck Institute for Intelligent Systems.

My research focus lies in the intersection between optimization and machine learning. I have a keen interest in developing optimization algorithms for trustworthy and reliable machine learning, particularly in areas such as differential privacy, fairness, and reproducibility.

Prior to this, I received my master's degree in computer science from ETH Zurich in 2021 and bachelor's degree in physics from Peking University in 2019.

Publications

Google Scholar

• DPZero: Private Fine-Tuning of Language Models without Backpropagation.
  Liang Zhang, Bingcong Li, Kiran Koshy Thekumparampil, Sewoong Oh, Niao He.
  Working Paper [arXiv] [Workshop]
• Generalization Bounds of Nonconvex-(Strongly)-Concave Stochastic Minimax Optimization.
  Siqi Zhang*, Yifan Hu*, Liang Zhang, Niao He.
  AISTATS 2024 [arXiv]
• Optimal Guarantees for Algorithmic Reproducibility and Gradient Complexity in Convex Optimization.
  Liang Zhang*, Junchi Yang*, Amin Karbasi, Niao He.
  NeurIPS 2023 (Spotlight) [NeurIPS] [arXiv]
• Bring Your Own Algorithm for Optimal Differentially Private Stochastic Minimax Optimization.
  Liang Zhang, Kiran Koshy Thekumparampil, Sewoong Oh, Niao He.
  NeurIPS 2022 [NeurIPS] [arXiv]

Talks

• SIAM Conference on Optimization (OP23), on differentially private stochastic minimax optimization.

Teaching

• Optimization for Data Science (Spring 2024), ETH Zurich.
• Foundations of Reinforcement Learning (Autumn 2021, Spring 2023), ETH Zurich.

Service

• Reviewer for NeurIPS, ICLR, and ICML.