Openings for PhD Students:  I am looking for self-motivated students in the areas of (1) AI safety and alignment and (2) reinforcement learning. If you are interested, please email me your CV, detailed transcripts, and GRE/TOEFL scores (if available). Students with experience in AI, machine learning, or security are preferred.

News

09/2026:	Our papers on reward shaping for robust RL and data poisoning against world models have been accepted to NeurIPS 2026. Congratulations, Zixuan and Yibin!
08/2026:	Our paper on risk-conditioned RLHF has been accepted to the main conference of EMNLP 2026. Well done, Zixuan!
06/2026:	Bangyan Ju joined Dr. Zheng's group as a new PhD Student. Welcome on board, Bangyan!
01/2026:	Our paper on robust optimization for mitigating reward hacking in reinforcement learning has been accepted to ICRL 2026. Congratulations, Zixuan!
11/2025:	Dr. Zheng receieved two grants from Open Philanthropy's Technical AI Safety Research program. Thank you, Open Philanthropy!
11/2025:	Our paper on fair resource allocation with probing in multi-armed bandits has been accepted to AAAI 2026. Congratulations Tianyi!
09/2025:	Our paper on diffuion-guided semantic attacks against RL agents has been accepted to NeurIPS 2025. Congratulations Xiaolin!
08/2025:	Yibin Hu joined Dr. Zheng's group as a new PhD Student. Welcome on board, Yibin!
05/2025:	Dr. Zheng received a one-year grant from the Louisiana Board of Regents to study LLM safety (sole PI).
01/2025:	Dr. Zheng is offering a new graduate elective on AI Safety and Security.

Short Bio

Zizhan Zheng is an Associate Professor of Computer Science at Tulane University. His recent work has centered on (1) developing automated red-teaming, proactive defense, alignment, and interpretability techniques to enhance the security and safety of both single-agent and cooperative AI systems, and (2) integrating reinforcement learning and generative AI for efficient and trustworthy decision-making, along with their applications in various science and engineering domains.

Previously, he also worked on the optimization and economics of networked systems, including computer networks, cloud computing systems, and cyber-physical systems, and he contributed to the science of security using tools from machine learning and game theory.

[CV][Google Scholar] [DBLP] [ResearchGate]

Research Interests

• AI safety and security
• Reinforcement learning
• LLMs and Agentic AI
• Networks