Xiaomeng Hu, Ph.D. student Department of Computer Science and Engineering The Chinese University of Hong Kong Supervisor: Professor Tsung-Yi Ho

BIOGRAPHY

I am currently a third-year Ph.D. candidate at The Chinese University of Hong Kong under the supervision of Prof. Tsung-Yi Ho.

My primary research interests lie in (1) interpreting and controlling the behavior of large language models (LLMs) through mechanism understanding and intervention via decoding strategies, and (2) enhancing the complex reasoning capabilities of LLMs using advanced techniques such as reinforcement learning. I am open to potential collaborations or industry internships, especially those focusing on LLMs + RL, feel free to reach out to me via email.

NEWS

• Jul. 2025 Token Highlighter was integrated into the IBM ICX 360 toolkit .
• Dec. 2024 Token Highlighter was accepted by AAAI 2025 with an oral presentation (top 4.65%)
• Sep. 2024 Gradient Cuff was accepted by NeurIPS 2024.
• Apr. 2024 IBM Blog highlighted Gradient Cuff: What is red teaming for generative AI?
• Sep. 2023 RADAR was accepted by NeurIPS 2023.
• Jul. 2023 IBM Blog highlighted RADAR: What is AI forensics?
• Apr. 2022 P3 Ranker was accepted by SIGIR 2022.

CONTACT

• Email (Personal): greghxm [at] foxmail.com
  
• Email (Campus): greghxm [at] link.cuhk.edu.hk
  

RESEARCH FOCUS (RECENT)

• Large Language Models

• Reinforcement Learning

EDUCATION

• Ph.D. Computer Science and Engineering, The Chinese University of Hong Kong, Aug. 2023 -

• B.Eng. Artificial Intelligence, Northeastern University, Sep. 2019 - Jul. 2023

PUBLICATION

* denotes equal contribution.

Preprints

• [P1] Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho. Attention Slipping: A Mechanistic Understanding of Jailbreak Attacks and Defenses in LLMs. arxiv 2025.07 . (paper)

Conference papers

• [C4] Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho. Token Highlighter: Inspecting and Mitigating Jailbreak Prompts for Large Language Models. AAAI 2025 (Oral). (paper) (demo) (code) (IBM ICX 360 Coverage)

• [C3] Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho. Gradient Cuff: Detecting Jailbreak Attacks on Large Language Models by Exploring Refusal Loss Landscapes. NeurIPS 2024. (paper) (demo) (code) (IBM Blog)

• [C2] Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho. RADAR: Robust AI-Text Detection via Adversarial Learning. NeurIPS 2023. (paper) (demo) (code) (IBM Blog)

• [C1] Xiaomeng Hu*, Shi Yu*, Chenyan Xiong, Zhenghao Liu, Zhiyuan Liu, Ge Yu. P³ Ranker: Mitigating the Gaps between Pre-training and Ranking Fine-tuning with Prompt-based Learning and Pre-finetuning. SIGIR 2022. (paper) (code)

EXPERIENCE

• Alibaba Qwen Team, Apr. 2025 - Now
  Research Intern
  Topic: General Reinforcement Learning & LLM Safety Alignment
  

• IBM Research AI, Feb. 2023 - Sept. 2023
  Research Intern
  Topic: Adversarial Reinforcement Learning & AI-Text detection
  Project: RADAR (NeurIPS 2023)
  Product Demonstration: RADAR Demo
  

• THU-NLP Lab, Aug. 2021 - Jun. 2022
  Research Intern
  Topic: Efficient Model Finetuning & Information Retrieval
  Project: P3 Ranker (SIGIR 2022)

SERVICE

TEACHING ASSISTANT

• 2023 Fall: CSCI3130 Formal Languages and Automata Theory

• 2024 Spring: ENGG1110E Problem Solving By Programming (C language)

• 2024 Fall: CSCI3130 Formal Languages and Automata Theory

• 2025 Spring: CSCI3320 Fundamentals of Machine Learning

REVIEWER

• NeurIPS 2023, ICLR 2025, IJCAI 2025, ACL ARR 2025 February, ACL ARR 2025 May, NeurIPS 2025, AAAI 2026