Zi Liang (Research Page)
Table of Contents
Now Zi Liang is a PhD student in Astaple Group, the Hong Kong Polytechnic University (PolyU).
1. Education
- 2016.09-2020.06: Bachelor Degree, in Northeastern University, in Shenyang, on cybernetics (Control Theory);
- 2020.09-2023.06: Master Degree, in iMiss Group of Xi'an Jiaotong University, in Xi'an, on software engineer and research for task-oriented dialogue systems;
- 2023.10-now: The Hong Kong Polytechnic University in Hong Kong. Research of interests: dialogue system, safety problem of AI, and interpretability of LLMs.
2. He was interested in:
- Dialogue systems (2020.09~2022.11) –> generalized AI (2022.11~now)
- PLM-based NLG (2019.11-2023.04) –> LLMs-based NLG (2023.04~now)
- AI safety (2021.03~now)
- Interpretability of current AI models (2023.07~now)
3. Paper Published.
3.1. Preprint
- Liang, Zi, et al. "Healing Unsafe Dialogue Responses with Weak Supervision Signals." arXiv preprint arXiv:2305.15757 (2023). [Paper][Code]
- Wang Yanyun, et al. "TSFool: Crafting Highly-Imperceptible Adversarial Time Series through Multi-Objective Gray-Box Attack". arXiv preprint (2024). [Paper] [Code]
3.3. Corporations
- Zhang, Shuo, et al. "Multi-Action Dialog Policy Learning from Logged User Feedback." arXiv preprint arXiv:2302.13505 (2023). [Paper]
4. Contact Me
- GitHub: https://github.com/liangzid
- MAIL: frostliangzi@qq.com
- Google Scholar: HERE