Honghao Liu

AI Ph.D. Student

Honghao Liu

I study trustworthy AI, agentic AI, and computational neuroscience.

CV GitHub Google Scholar Email

Research

Cognitive systems, secure data and models, and swarm for future prediction.

Trustworthy AI

Safety and robustness for large reasoning models, including conflict-driven attack analysis and agentic workflow generation benchmarks.

Privacy-Preserving LLMs

Entity-based encrypted synthetic data for continual pretraining while protecting personally identifiable information.

Neural Representation

EEG and fMRI workflows for multi-talker speech analysis from brain activity.

Selected Publications

View all

Recent work

  1. 2026

    Conflicts Make Large Reasoning Models Vulnerable to Attacks. ACL Findings.

  2. 2026

    Continual Pretraining on Encrypted Synthetic Data for Privacy-Preserving LLMs. EACL findings.

  3. 2026

    A Survey on LLM-as-a-Judge. The Innovation.

Portrait of Honghao Liu

Current

Ph.D. student at HKUST(GuangZhou), Artificial Intelligence.

I am supervised by Prof. Lionel Ni and currently work as a research intern at the International Digital Economy Academy in Shenzhen. Previously, I completed an MPhil in Data Science and Analytics at HKUST supervised by Prof. Qiong Luo and a BEng in Computer Science at Wuhan University of Technology.