✨ About Me

I am currently a first-year PhD student in system engineering at CUHK under the supervision of Prof. Xixin Wu. Previously, I received an MPhil AI degree from HKUST(GZ), under the supervision of Prof. Li Liu and the co-supervision of Prof. Yi R. (May) Fung, plus an MSc CS degree from HKU. Between the two master’s studies, I worked as an RA at ASTAPLE of PolyU with Prof. Haibo Hu. I got my BEng degree in software engineering at ECNU, where I began my research career under the supervision of Prof. Dehui Du. It is also my great pleasure to work with close co-authors Zi Liang (PolyU), Ziyu Zhou (HKUST), Yu Huang (CityU), Yiming Huang (HKUST), Weilin Lin (HKUST(GZ)), and Nanjun Zhou (ZJU).

My research interests include AI Security and Multimodal Large Language Models. Specifically, I used to study adversarial attacks and defenses across time series, vision, and speech. Together with close co-authors, I have also dabbled in model extraction, backdoor, data poisoning, agent security, etc. Recently, I have been interested in MLLM post-training and inference-time security, especially regarding safety alignment and modality gap under multimodal inputs.

I would always be open to discussions and potential collaborations. Please feel free to email me!

🔥 News

• 2026.08: Reported for duty at CUHK! Let’s write the new story together ⭐
• 2026.07: Graduated from HKUST(GZ)! My sincere thanks to everyone for all the help and support, especially to my respected supervisor, Prof. Liu. Will always miss the precious two years 😭
• 2026.07: One paper benchmarking text-to-audio-video accepted by ACMMM’26 as Oral. Congratulations to Prof. Liu, Tianxin, and Wentao 🎉
• 2026.05: One paper on universal jailbreak of audio language models accepted by ICML’26. Many thanks to Prof. Liu, Prof. Wu, Yu, and Zi 🎉
• 2026.03: Been awarded the Lizhi Innovative Talent Scholarship! My great thanks to HKUST(GZ) Info Hub, Lizhi Group, and Prof. Liu 🙌
• 2026.03: One paper on accuracy-robustness trade-off in adversarial training accepted by CVPR’26 as Findings. Many thanks to Prof. Hu, Prof. Liu, Prof. Ye, and Zi 🎉
• 2026.01: One paper on adversarial fine-tuning of speech foundation models accepted by ICASSP’26 as Oral. Many thanks to Prof. Liu 🎉

📝 Publications

As 1st Author:

ICML'26

Acoustic Interference: A New Paradigm Weaponizing Acoustic Latent Semantic for Universal Jailbreak against Large Audio Language Models

Yanyun Wang*, Yu Huang*, Zi Liang, Xixin Wu, Li Liu^†

In 43rd International Conference on Machine Learning, 2026.

[Project Page]    [Slides]    [Poster]    [Code]

CVPR'26 Findings

Robust Alignment: Harmonizing Clean Accuracy and Adversarial Robustness in Adversarial Training

Yanyun Wang, Qingqing Ye, Li Liu, Zi Liang, Haibo Hu^†

In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026.

[Poster]    [Code]

ICASSP'26 Oral

Adversarial Fine-Tuning on Speech Foundation Model with Vulnerable Attention Consistency Regularization for Robust Speech Recognition

Yanyun Wang, Baoyuan Wu, Li Liu^†

In IEEE International Conference on Acoustics, Speech and Signal Processing, 2026.

[Slides]    [Code]

ICCV'25

Failure Cases Are Better Learned But Boundary Says Sorry: Facilitating Smooth Perception Change for Accuracy-Robustness Trade-Off in Adversarial Training

Yanyun Wang, Li Liu^†

In IEEE/CVF International Conference on Computer Vision, 2025.

[Poster]    [Code]

ECAI'24 Oral

TSFool: Crafting Highly-Imperceptible Adversarial Time Series through Multi-Objective Attack

Yanyun Wang, Dehui Du^†, Haibo Hu^†, Zi Liang, Yuanhao Liu

In 27th European Conference on Artificial Intelligence, 2024.

[Poster]    [Code]

• [Preprint] New Paradigm of Adversarial Training: Releasing Accuracy-Robustness Trade-Off via Dummy Class, Yanyun Wang, Li Liu^†, Zi Liang, Yi R. Fung, Qingqing Ye, Haibo Hu. Arxiv Preprint, 2024.

• [SMC’23 Oral] Meta Pattern Concern Score: A Novel Evaluation Measure with Human Values for Multi-classifiers, Yanyun Wang, Dehui Du^†, Yuanhao Liu. In IEEE International Conference on Systems, Man, and Cybernetics, 2023.

As Co-Author:

• [Preprint] CAITLYN: Can LLM Agents Autonomously Synthesize Defenses against Emerging Injection Attacks?, Zi Liang*, Xiaoyu Xu*, Yanyun Wang, Minxin Du, Qingqing Ye, and Haibo Hu^†. Arxiv Preprint, 2026.

• [Preprint] Adversarial Attacks on Deep OCR Systems, Wenbo Sun*, Hongzong LI*, Yanyun Wang, Jiahao MA, Shuxin Zhuang, Rong Feng, Shiqin Tang, and Zi Liang^†. Arxiv Preprint, 2026.

• [Preprint] Can a Single Message Paralyze the AI Infrastructure? The Rise of AbO-DDoS Attacks through Targeted Mobius Injection, Zi Liang, Ronghua Li, Yanyun Wang, Qingqing Ye, and Haibo Hu^†. Arxiv Preprint, 2026.

• [ACMMM’26 Oral] PhyAVBench: A Challenging Audio Physics-Sensitivity Benchmark for Physically Grounded Text-to-Audio-Video Generation, Tianxin Xie, Wentao Lei, Kai Jiang, Guanjie Huang, Pengfei Zhang, Chunhui Zhang, Fengji Ma, Haoyu He, Han Zhang, Jiangshan He, Jinting Wang, Linghan Fang, Lufei Gao, Orkesh Ablet, Peihua Zhang, Ruolin Hu, Shengyu Li, Weilin Lin, Xiaoyang Feng, Xinyue Yang, Yan Rong, Yanyun Wang, Zihang Shao, Zelin Zhao, Chenxing Li, Shan Yang, Wenfu Wang, Meng Yu, Dong Yu, and Li Liu^†. In 34th ACM International Conference on Multimedia, 2026.

• [AAAI’26] Revitalizing Canonical Pre-Alignment for Irregular Multivariate Time Series Forecasting, Ziyu Zhou, Yiming Huang, Yanyun Wang, Yuankai Wu, James Kwok, and Yuxuan Liang^†. In 40th AAAI Conference on Artificial Intelligence, 2026.

• [NeurIPS’25 Spotlight] Virus Infection Attack on LLMs: Your Poisoning Can Spread “VIA” Synthetic Data, Zi Liang, Qingqing Ye, Xuan Liu, Yanyun Wang, Jianliang Xu, and Haibo Hu^†. In 39th Annual Conference on Neural Information Processing Systems, 2025.

• [NeurIPS’25] BackdoorDM: A Comprehensive Benchmark for Backdoor Learning in Diffusion Model, Weilin Lin, Nanjun Zhou, Yanyun Wang, Jianze Li, Hui Xiong, and Li Liu^†. In 39th Annual Conference on Neural Information Processing Systems, 2025.

• [ACL’25 Main] “Yes, My LoRD.” Guiding Language Model Extraction with Locality Reinforced Distillation, Zi Liang, Qingqing Ye, Yanyun Wang, Sen Zhang, Yaxin Xiao, Ronghua Li, Jianliang Xu, and Haibo Hu^†. In 63rd Annual Meeting of the Association for Computational Linguistics, 2025.

• [AAAI’22 Workshop] Efficient Adversarial Sequence Generation for RNN with Symbolic Weighted Finite Automata, Mingjun Ma, Dehui Du^†, Yuanhao Liu, Yanyun Wang, and Yiyang Li. In SafeAI Workshop @ 36th AAAI Conference on Artificial Intelligence, 2022.

📖 Educations

• 2026.08 - Now, Ph.D. in System Engineering, The Chinese University of Hong Kong.
• 2024.09 - 2026.07, M.Phil. in Artificial Intelligence, The Hong Kong University of Science and Technology (Guangzhou).
• 2022.09 - 2023.10, M.Sc. in Computer Science, The University of Hong Kong.
• 2018.09 - 2022.07, B.Eng. in Software Engineering, East China Normal University.

💻 Internships

• 2026.06 - 2026.08, LLM Research Intern, Bairong Intelligence, Beijing.
• 2026.03 - 2026.06, MLLM Data Intern, ByteDance Seed, Beijing.
• 2023.10 - 2024.06, Research Assistant, Applied Security, Trust And Privacy Lab for Enterprise, Hong Kong.
• 2021.08 - 2021.11, NLP Algorithm Intern, Ping An Technology, Shanghai.
• 2020.07 - 2020.08, Software Development Intern, Dareway Intelligent Technology, Jinan.

🎖 Honors and Awards

• 2026, Postgraduate Studentship Award (48 months), The Chinese University of Hong Kong.
• 2026, Lizhi Innovative Talent Scholarship, The Hong Kong University of Science and Technology (Guangzhou) & Lizhi Group.
• 2025, Second-Class Award, The 15th Guangdong-Hong Kong-Macao Conference on Image and Graphics.
• 2024, Postgraduate Studentship Award (24 months), The Hong Kong University of Science and Technology (Guangzhou).
• 2022, Excellent Graduate Award, East China Normal University.
• 2022, Excellent Bachelor’s Degree Thesis Award, Software Engineering Institute, East China Normal University.
• 2021, People’s Choice Award, School of Computing Summer Workshop, National University of Singapore.
• 2021, Excellent Undergraduate Student, East China Normal University.
• 2021, Second-class Scholarship, East China Normal University.
• 2020, Excellent Undergraduate Student, East China Normal University.
• 2020, First-class Scholarship, East China Normal University.

💬 Services

• Conference Program Committee:
  • AAAI Conference on Artificial Intelligence (AAAI): 2027, 2026, 2025
• Conference Reviewer:
  • IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR): 2026, 2025, 2024
  • ACM International Conference on Multimedia (ACMMM): 2026
  • IEEE/CVF International Conference on Computer Vision (ICCV): 2025
  • European Conference on Computer Vision (ECCV): 2026
  • British Machine Vision Conference (BMVC): 2026
  • RSI Workshop @ International Conference on Learning Representations (ICLR): 2026
• Journal Reviewer:
  • IEEE Transactions on Knowledge and Data Engineering (TKDE)
  • Neural Networks (NN)