Huan Zhang
Scholar

Huan Zhang

Google Scholar ID: LTa3GzEAAAAJ
Assistant Professor at University of Illinois Urbana-Champaign
Machine learningFormal VerificationTrustworthy AIMachine Learning SecurityOptimization
Citations & Impact
All-time
Citations
17,082
 
H-index
48
 
i10-index
76
 
Publications
20
 
Co-authors
23
list available
Resume (English only)
Academic Achievements
  • - Recipient of Schmidt Futures AI2050 Early Career Fellowship
  • - Recipient of MathWorks research award
  • - Published multiple papers on topics such as learning stabilizing neural controllers, certified rendering, improving data efficiency for LLM reinforcement fine-tuning, and accelerating neural network verification
Research Experience
  • - Assistant Professor at the Department of Electrical and Computer Engineering, University of Illinois Urbana-Champaign (UIUC)
  • - Research Projects: Including compact and certified models (e.g., neural network controllers), large frontier models (e.g., LLMs, VLMs), and agentic AI models (e.g., embodied agents)
Education
  • - Ph.D. in Computer Science from UCLA in 2020, advisor: Prof. Cho-Jui Hsieh
  • - Bachelor's degree from Zhejiang University (ZJU) in 2012
  • - Postdoctoral researcher at Carnegie Mellon University (CMU) with Prof. Zico Kolter from 2021-2023
Background
  • - Research Interests: Building trustworthy AI systems that can be safely and reliably used in mission-critical tasks
  • - Professional Field: Formal verification, deep neural networks, AI safety, robustness, and efficiency
  • - Brief Introduction: Developed a novel formal verification framework for deep neural networks that scales to millions of neurons. The team's open-source neural network verifier, α,β-CROWN, has been used in various applications such as robotics, control, and power systems.
Miscellany
  • - Openings: Looking for students with strong technical backgrounds in machine learning, artificial intelligence, and their applications
  • - Relevant experience: Trustworthy machine learning, formal verification/certification, AI safety/security, or agentic AI is preferred but not required