Foundational Challenges in Assuring Alignment and Safety of Large Language Models
Zhaowei Zhang (张钊为)
Ph.D. Student
School of Intelligence Science and Technology
Peking University
Email: zwzhang [at] stu (dot) pku (dot) edu (dot) cn
[Google Scholar] [Github] [Twitter] [LinkedIn]
Zhaowei is pronounced as "Ju" (as in judge) + "ou" (as in out) + "Way"; Zhang, or Cheung in Hong Kong, is "Ju" (as in judge) + "on" | audio ([International Phonetic Alphabet, IPA]) here: [tʂɑuwei][tʂɑŋ].
Research Topics
- AI Alignment and Interaction
- Multi-Agent System
- Scalable Oversight
- LLMs and Foundation Models
- Socio-Technical Theory
- Game Theory and Mechanism Design
I am currently a Ph.D. Student at School of Intelligence Science and Technology, Peking University as well as a research intern at BIGAI. I am in the group of PAIR-Lab led by Prof. Yaodong Yang. The long-term goal of my research is to find a way to build a trustworthy and social AI system. To this end, my research focuses on AI Alignment and Multi-Agent System. In particular, I am currently interested in exploring the potential of the value learning and reward gaming problem for creating an intelligent agent free from utilitarianism (i.e. being only motivated by reward modeling), especially in complex scenarios such as social interaction.
Before entering PKU, I was a Bachelor Student at Wuhan University majoring in Computer Science and Technology. I received the B.E. degree from School of Computer Science at Wuhan University in 2023.