I am a PhD student at UNSW Sydney, under the supervision of Senior Lecturer Yuekang Li, Associate Professor Huadong Mo, Lecturer Nan Sun, and Scientia Professor Jingling Xue. My research centers on human–AI distinguishability, particularly in web environments, and I also work on energy management systems. I work closely with Wei Song and Yi Liu.
My work IllusionCAPTCHA has been reported by 20+ news outlets and adopted as a case study in a university course study guide.
News
- 2026: Paper on tool-calling in agentic LLMs accepted at NeurIPS 2026.
- 2026: Paper accepted at IEEE S&P 2026 (PUFFERDOS).
- 2026: Papers accepted at EMNLP 2026 and ACL Findings 2026.
- 2026: NgCaptcha published at WWW 2026 Companion.
- 2025: 🏆 Best Innovation Paper at IEEE S&P Workshops (SPW) 2025.
- 2025: IllusionCAPTCHA accepted at The Web Conference (WWW 2025).
Selected Works
- WWW 2025 CORE A* IllusionCAPTCHA: A CAPTCHA based on Visual IllusionZiqi Ding, Gelei Deng, Yi Liu, Junchen Ding, Jieshan Chen, Yulei Sui, Yuekang Li.Proceedings of the ACM Web Conference, 2025.
Media coverage: New Scientist; PetaPixel; ITmedia; Creative Bloq; Cybersecurity News and more. Course material: included in Rutgers University's CS 419 Computer Security study guide.
Impact of IllusionCAPTCHA
Publications
2026
- How Do Agentic LLMs Decide to Call Tools? A Tool-Call Vector Shaped by Suppression CORE A*X. Gong, T. Han, J. Zhang, W. Song, Z. Ding, H. Yan, Y. Sun, L. HuNeurIPS 2026
- Help or Hurdle? Rethinking Model Context Protocol-Augmented Large Language Models CORE A*W. Song*, H. Zhong*, Z. Ding*, Z. Xing, Y. Sui, J. Xue, Y. LiEMNLP 2026* Equal contribution (co-first authors).
- NgCaptcha: A CAPTCHA Bridging the Past and the Future CORE A*Z. Ding, S. Xu, W. Song, Y. LiWWW 2026 Companion
- PUFFERDOS: Efficient and Effective Attack String Generation for Regular Expression Denial of Service Vulnerabilities CORE A*S. Xu, Z. Ding, X. Cheng, Y. Li, N. Sun, B. Turnbull, S. Kan, S. MaIEEE S&P 2026
- Membership Inference Attacks Against Video Large Language Models CORE BW. Song, Y. Cao, Z. Ding, Y. Liu, G. Deng, Y. LiSiMLA 2026
- Debiasing LLMs by Masking Unfairness-Driving Attention Heads CORE A*T. Han, W. Song, Z. Ding, Z. Li, C. Fang, Y. Li, D. Liu, Z. Chen, Z. WangACL Findings 2026
2025
- TOMBRAIDER: Entering the Vault of History to Jailbreak Large Language Models CORE A*J. Ding, J. Zhang, Y. Liu, Z. Ding, G. Deng, Y. LiEMNLP 2025
- A Rusty Link in the AI Supply Chain: Detecting Evil Configurations in Model Repositories CORE A* 🏆 Best Innovation PaperZ. Ding, Q. Fu, J. Ding, G. Deng, Y. Liu, Y. LiIEEE S&P Workshops (SPW) 2025
- IllusionCAPTCHA: A CAPTCHA based on Visual Illusion CORE A*Z. Ding, G. Deng, Y. Liu, J. Ding, J. Chen, Y. Sui, Y. LiWWW 2025
- Mutual Information Guided Backdoor Mitigation for Pre-trained Encoders CORE AT. Han, W. Sun, Z. Ding, C. Fang, H. Qian, J. Li, Z. Chen, X. ZhangIEEE TIFS 2025
2024
- Exploring Large Language Models for Method Name Prediction CORE BH. Qian, T. Xu, Z. Ding, W. Liu, S. ZhuQRS 2024
- A Survey of Learning-based Method Name Prediction CORE A*H. Qian, W. Liu, T. Xu, J. Yin, X. Feng, W. Sun, Z. Ding, et al.ACM TOSEM 2024
2023
- Abstract Syntax Tree for Method Name Prediction: How Far Are We? CORE BH. Qian, W. Liu, Z. Ding, W. Sun, C. FangQRS 2023