I am an undergraduate student majoring in Cyberspace Security (Experimental Class) at Beijing University of Posts and Telecommunications (BUPT), advised by Prof. Jie Hao. Currently, I am a Research Intern at Duke University, advised by Prof. Neil Gong. Previously, I was fortunate to be advised by Prof. Ke Xu at Tsinghua University. I also collaborate closely with Dr. Yuxin Cao (from National University of Singapore).

Research Interests

My research vision is to build secure, trustworthy, and robust AI systems capable of interacting with the physical world. I systematically address vulnerabilities across the AI’s expanding capability stack:

• Perception (“The Ears”): Defending against input-level manipulation in multimodal perception, exploring modal-agnostic adversarial robustness beyond just the audio domain.
• Cognition (“The Brain”): Mitigating model-level exploits and ensuring safety alignment in Multimodal Large Language Models (MLLMs) and autonomous AI Agents.
• Memory (“The Knowledge”): Protecting Retrieval-Augmented Generation (RAG) systems and external knowledge bases from context-level injection and poisoning.

If you would like to learn more about my work, feel free to check out my recent publications below or contact me directly.

News

• 11/2025: I will serve as a reviewer for ICME 2026.
• 10/2025: I will serve as a reviewer for ICASSP 2026.
• 09/2025: I was awarded the Xiaomi Grand Prize Scholarship (the highest scholarship at BUPT, awarded to only 5 among all undergraduates).
• 09/2025: One first-author paper on protecting audio-language models against jailbreaks was accepted to NeurIPS 2025.
• 06/2025: One co-authored paper on iterative binary malware summarization was accepted to IEEE TIFS.
• 03/2025: One first-author paper on boosting the transferability of audio adversarial examples was accepted to ICME 2025.
• 01/2025: One first-author paper on speech privacy protection against eavesdroppers was accepted to USENIX Security 2025.
• 09/2024: We successfully received funding from the Beijing Natural Science Foundation Undergraduate “QiYan” Program.
• 04/2024: One first-author paper on audio adversarial attacks was accepted to SecTL 2024 (AsiaCCS Workshop).

Publications

• ALMGuard: Safety Shortcuts and Where to Find Them as Guardrails for Audio–Language Models
  Weifei Jin, Yuxin Cao, Junjie Su, Minhui Xue, Jie Hao, Ke Xu, Jin Song Dong, Derui Wang.
  To appear in the Thirty-ninth Annual Conference on Neural Information Processing Systems (NeurIPS), 2025.

• MALSIGHT: Exploring Malicious Source Code and Benign Pseudocode for Iterative Binary Malware Summarization
  Haolang Lu, Hongrui Peng, Guoshun Nan, Jiaoyang Cui, Cheng Wang, Weifei Jin, Songtao Wang, Shengli Pan, Xiaofeng Tao.
  In IEEE Transactions on Information Forensics and Security (TIFS), 2025.

• Boosting the Transferability of Audio Adversarial Examples with Acoustic Representation Optimization
  Weifei Jin, Junjie Su, Hejia Wang, Yulin Ye, Jie Hao.
  In IEEE International Conference on Multimedia & Expo (ICME), 2025.

• Whispering Under the Eaves: Protecting User Privacy Against Commercial and LLM-powered Automatic Speech Recognition Systems
  Weifei Jin, Yuxin Cao, Junjie Su, Derui Wang, Yedi Zhang, Minhui Xue, Jie Hao, Jin Song Dong, Yixian Yang.
  In the 34th USENIX Security Symposium (USENIX Security), 2025.

• Towards Evaluating the Robustness of Automatic Speech Recognition Systems via Audio Style Transfer
  Weifei Jin, Yuxin Cao, Junjie Su, Qi Shen, Kai Ye, Derui Wang, Jie Hao, Ziyao Liu.
  In the 2nd ACM Workshop on Secure and Trustworthy Deep Learning Systems (SecTL, AsiaCCS Workshop), 2024, pp. 47–55.

Awards

• Xiaomi Grand Prize Scholarship, (Highest scholarship in BUPT, Top 0.01%, 2025)
• Individual Award Representative, 2024-2025 BUPT Student Commendation Conference, BUPT (Top 0.1%, 2025)
• Outstanding Student Leader Award, BUPT (2025)
• Merit Student, BUPT (2024)
• Second Prize, 9th National Cryptography Technology Competition (2024)

Services

• Academic Reviewing: Served as a reviewer for TDSC, ICASSP 2026, ICME 2026/2025.
• Mentoring: Guided junior students in Cyberspace Security projects focusing on trustworthy Audio-Language Models.