Yang Liu

Yang Liu is an Assistant Professor in the Department of Electrical and Computer Engineering at the Baskin School of Engineering, University of California, Santa Cruz. Previously, they were affiliated with Harvard University and earned their PhD in 2015 from the Department of EECS at the University of Michigan, Ann Arbor. Their research lies at the intersection of machine learning, fairness, and trustworthy AI, with a strong focus on large language models, federated learning, and causal reasoning.

Their research interests include:

• Machine Learning and Fairness
• Federated and Privacy-Preserving Learning
• Large Language Model Safety and Unlearning
• Causal Inference and Counterfactual Reasoning
• Anomaly Detection and Robust Forecasting
• Human-AI Interaction and Ethical AI

Recent publications (2024–2025) demonstrate a strong trend in developing methods for machine unlearning, fairness in LLMs, and robustness under label noise and distribution shifts. Their work frequently appears in top-tier venues such as NeurIPS, ICLR, ICML, AAAI, and KDD, often in collaboration with researchers like Zhaowei Zhu, Mingyan Liu, Jiaheng Wei, and Kun Zhang. Themes include algorithmic fairness, model accountability, and human-aligned AI systems.

Scientific contributions include:

• Frameworks for LLM unlearning and model editing
• Methods for fair classification and recourse
• Robust time series forecasting under anomalies
• Test-time adaptation in multimodal models
• Causal approaches to debiasing and policy learning

While no formal advising list is provided, the depth and volume of collaborative work suggest active mentorship of graduate students and postdocs. Their research program is highly active, with numerous ongoing projects in trustworthy and socially responsible AI.