Han Wang

I am Han Wang, a second-year Ph.D. candidate at UIUC, advised by Prof. Huan Zhang. I also work closely with Prof. Gang Wang at UIUC.

Prior to my Ph.D., I was fortunate to work with Prof. Fei Wu (ZJU), Prof. Tat-Seng Chua (NUS), and Prof. Sharon Li (UW-Madison).

I am interning at IBM Research in New York this summer. Happy to chat if you are interested in my research.

Email | Google Scholar | Github | Linkedin

Research Interests

My current research is mainly around the LLM safety and evaluation.

LLM Safety

CoT monitoring, deceptive reasoning, code security, and realistic risks in autonomous agents.

CoT Monitoring Agent Safety Code Security Deception

LLM Evaluation

Benchmarks and evaluation protocols for reasoning models and agents.

Agent Evaluation Benchmark Contamination

/ / Projects (* denotes the equal contribution)

MonitorBench: A Comprehensive Benchmark for Chain-of-Thought Monitorability in Large Language Models

Han Wang*, Yifan Sun*, Brian Ko*, Mann Talati, Jiawen Gong, Zimeng Li, Naicheng Yu, Xucheng Yu, Wei Shen, Vedant Jolly, Huan Zhang

Preprint

[Paper] [Code]
CoT-Guard: Small Models for Strong Monitoring

Nirav Diwan*, Han Wang*, Berkcan Kapusuzoglu, Ramin Moradi, Supriyo Chakraborty, Giri Iyengar, Sambit Sahu, Huan Zhang, Gang Wang

Preprint

[Paper] [Code]
FORTIS: Benchmarking Over-Privilege in Agent Skills

Li Li, Chenxiao Yu, Han Wang, Wei Yang, Ryan A. Rossi, Franck Dernoncourt, Xiyang Hu, Philip S. Yu, Chaowei Xiao, Huan Zhang, Yue Zhao

Preprint

[Paper] [Code]
DecepChain: Inducing Deceptive Reasoning from Large Language Model

Wei Shen*, Han Wang*, Haoyu Li*, Huan Zhang

ICML 2026

[Paper] [Code] [Project]
On The Fragility of Benchmark Contamination Detection in Reasoning Models

Han Wang*, Haoyu Li*, Brian Ko*, Huan Zhang

ICLR 2026

[Paper] [Code] [Openreview]
How do Visual Attributes Influence Web Agents? A Comprehensive Evaluation of User Interface Design Factors

Kuai Yu*, Naicheng Yu*, Han Wang, Rui Yang, Huan Zhang

ACL 2026 Findings

[Paper] [Code]
AlphaOne: Reasoning Models Thinking Slow and Fast at Test Time

Junyu Zhang*, Runpei Dong*, Han Wang, Xuying Ning, Haoran Geng, Peihao Li, Xialin He, Yutong Bai, Jitendra Malik, Saurabh Gupta, Huan Zhang

EMNLP 2025 Main && MATH-AI Workshop @ NeurIPS 2025

[Paper] [Code] [Project] [Press]
The Emperor's New Clothes in Benchmarking? A Rigorous Examination of Mitigation Strategies for LLM Benchmark Data Contamination

Yifan Sun*, Han Wang*, Dongbai Li*, Gang Wang, Huan Zhang

ICML 2025 && Data Problems Workshop @ ICLR 2025

[Paper] [Code]
Steering Away from Harm: An Adaptive Approach to Defending Vision Language Model Against Jailbreaks

Han Wang, Gang Wang, Huan Zhang

CVPR 2025

[Paper] [Code]
ALI-Agent: Assessing LLMs' Alignment with Human Values via Agent-based Evaluation

Jingnan Zheng*, Han Wang*, An Zhang, Tai D. Nguyen, Jun Sun, Tat-Seng Chua

NeurIPS 2024

[Paper] [Code]

Industrial Experience

IBM Research, NY
Research Scientist Intern • May 2026 to Present

Education

University of Illinois Urbana-Champaign, IL
Ph.D. • Aug. 2024 to Present

Zhejiang University, China
B.Eng. • Aug. 2020 to June 2024

Services

Conference Reviewer: NeurIPS 2025 - 2026, ICLR 2026, ICML 2026, COLM 2026, ACL ARR 2025 - 2026, Trustworthy AI Workshop @ ICLR 2026, AISec Workshop @ ACM CCS 2025, MATH-AI Workshop @ NeurIPS 2025

Journal Reviewer: IEEE TNNLS 2025, Neural Computation 2026

Teaching Assistant: ECE 484 (Principles of Safe Autonomy), UIUC, Fall 2025