I’m a researcher at Alibaba’s Tongyi Lab, working on the safety, trustworthinese, and cultural inclusiveness of large language models. In parallel, I study speech understanding, generation and real-time interpretation.
I earned my Ph.D. in 2024 from Zhejiang University, under the supervision of Prof. Kui Ren and Prof. Zhongjie Ba, with joint guidance from Prof. Yuan Hong at the University of Connecticut. My research focuses on trustworthy AI, aiming to improve the safety, robustness, and privacy of multimodal models.
I’m always open to collaboration. Feel free to reach out at ixinyuzhang AT gmail.com. We’re also recruiting intern researchers.
🪐 News
- 2026.04: Qwen3.5-Omni is live: a native omni-modal LLM supporting 113 languages for speech recognition.
- 2026.01: Introducing open-source Qwen3-ASR: robust, streaming, multilingual ASR.
- 2026.01: Our Qwen3-TTS family is now open sourced!
- 2025.09: Our Qwen3-Live-Translate multimodal simultaneous translation API has been officially released!
- 2025.08: Our paper CultureSynth accepted at EMNLP 2025 📍 Suzhou, China — see you there!
- 2025.06: Our paper PrivacyAsst has been selected as a Popular Article in the IEEE TDSC journal!
🥝 Publications
Conference & Journal Papers (Selected)
-
CultureSynth: A Hierarchical Taxonomy-Guided and Retrieval-Augmented Framework for Cultural Question-Answer Synthesis
, Pei Zhang, Shuang Luo, Jialong Tang, Yu Wan, Baosong Yang, Fei Huang
EMNLP 2025
[PDF] [BENCHMARK] -
Certifiable Black-Box Attacks with Randomized Adversarial Examples: Breaking Defenses with Provable Confidence
Hanbin Hong, , Binghui Wang, Zhongjie Ba, Yuan Hong
ACM CCS 2024 (Security Top 4)
[PDF] [CODE] [MEDIA] -
Fltracer: Accurate poisoning attack provenance in federated learning
, Qingyu Liu, Zhongjie Ba, Yuan Hong, Tianhang Zheng, Feng Lin, Li Lu, Kui Ren
IEEE TIFS 2024
[PDF] [CODE] -
Privacyasst: Safeguarding user privacy in tool-using large language model agents
, Huiyu Xu, Zhongjie Ba, Zhibo Wang, Yuan Hong, Jian Liu, Zhan Qin, Kui Ren
IEEE TDSC 2024
[PDF] [CODE] -
Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial Attacks
, Hanbin Hong, Yuan Hong, Peng Huang, Binghui Wang, Zhongjie Ba, Kui Ren
IEEE S&P 2024 (Security Top 4)
[PDF] [CODE] [MEDIA] -
An efficient e2e crowd verifiable e-voting system
, Bingsheng Zhang, Aggelos Kiayias, Thomas Zacharias, Kui Ren
IEEE TDSC 2021
[PDF]
-
Learning-based Practical Smartphone Eavesdropping with Built-in Accelerometer
Zhongjie Ba, Tianhang Zheng, , Zhan Qin, Baochun Li, Xue Liu, Kui Ren
NDSS 2020 (Security Top 4)
[PDF] -
CFP: Enabling camera fingerprint concealment for privacy-preserving image sharing
Zhongjie Ba, , Zhan Qin, Kui Ren
IEEE ICDCS 2019
[PDF]
Technical Reports
- Qwen3.5-Omni Technical Report, Qwen Team, 2026.
- Qwen3-ASR Technical Report, Qwen Team, 2026.
- Qwen3-TTS Technical Report, Qwen Team, 2026.
- Qwen3-Omni Technical Report, Qwen Team, 2025.
- Qwen3Guard Technical Report, Qwen Team, 2025.
- Qwen3 Technical Report, Qwen Team, 2025.
- Qwen2 Technical Report, Qwen Team, 2024.
🏆 Honors and Awards
- 2024, Excellent Graduate of Zhejiang University (Graduate level)
- 2020-2023, Excellent Graduate Student, Zhejiang University
- 2019, Doctoral Freshman Entrance Scholarship, Zhejiang University
- 2019, Outstanding Graduate of Xidian University (Undergraduate level)
- 2018, National Scholarship (Undergraduate, Top 1%)
- 2018, Cybersecurity Scholarship - China Internet Development Foundation
🌱 Educations
- 2019.06 - 2024.12, Ph.D., College of Computer Science and Technology, Zhejiang University, Hangzhou.
- 2015.09 - 2019.06, B.Eng., School of Cyber Engineering (Elite Program), Xidian Univeristy, Xi’an.
✈️ Internships
- 2024.04 – 2024.10, Research Intern, Tongyi Lab, Alibaba Group, Hangzhou, China
Supervisor: Baosong Yang and Pei Zhang - 2022.09 – 2023.08, Visiting Researcher, University of Connecticut (UConn), Storrs, CT, USA
Supervisor: Prof. Hong Yuan - 2018.08 – 2019.01, Visiting Student Researcher, Zhejiang University, Hangzhou, China
Supervisor: Prof. Kui Ren and Prof. Zhan Qin