研究兴趣

我的研究方向是自适应人工智能，致力于构建能够在动态开放环境中持续学习、安全可信、自主适应的智能系统。研究重点包括数据稀缺和非平稳数据分布条件下的神经网络训练、持续学习、泛化能力提升与抗遗忘机制，并探索基于不确定性建模和自适应学习的高效训练框架，以提高人工智能模型在复杂动态环境中的泛化能力、鲁棒性、安全性和可靠性。此外，我长期关注基础模型的可信智能，围绕大语言模型、多模态大语言模型和扩散语言模型的安全性、可靠性与自适应推理开展研究，针对模型幻觉、自适应解码、高效推理以及模型安全等关键问题提出高效算法，为构建可信、高效、自适应的新一代基础模型提供理论基础与技术支撑。进一步地，我将前述理论应用于脑电信号分析任务，有效解决了该领域中普遍存在的抗干扰鲁棒性不足、标注数据匮乏以及非平稳分布等关键挑战。相关研究成果已发表于NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, ACL, EMNLP, TMLR, TPAMI等计算机领域顶级会议与期刊，累计发表论文40余篇。

教育背景

博士， 计算机科学,  纽约州立大学布法罗分校， 2018 - 2023     

硕士，韦仕敦大学， 计算机科学, 2016 - 2018

学士， 东北大学， 2011 - 2015

工作经历

计算机科学系助理教授, 中佛罗里大学，美国 2025-2026

博士后, 马里兰大学帕克分校， 2023-2025

Meta Reality lab (Pittsburgh), 研究科学家实习 2022 年 8月 – 2022 年12月

Amazon (Seattle, Washington), 应用研究科学家实习  2020年9月 –  2020年12月.

JD.com AI lab (Mountain View, CA), 自然语言研究实习 2020 年 6 月 –  2020 年 8月.

Tencent AI lab (Seattle, Washington), 自然语言研究实习  2019 年 6 月 – 2019 年 8 月

科研项目

2025  海外高层次青年人才， 项目负责人

代表性论文

(* indicates equal contribution, # indicates corresponding author)

1. CoRe: Context-Robust Remasking for Diffusion Language Models

Kevin Zhai, Sabbir Mollah, Zhenyi Wang, Mubarak Shah

International Conference on Machine Learning (ICML), 2026

2. AI Security in the Foundation Model Era: A Comprehensive Survey from a Unified Perspective

Zhenyi Wang, Siyu Luan

Transactions on Machine Learning Research (TMLR), 2026

3. Pixel-Perfect Puppetry: Precision-Guided Enhancement for Face Image and Video Editing

Yan Li, Zhenyi Wang, Guanghao Li, Wei Xue, Yike Guo, Wenhan Luo

International Conference on Learning Representations (ICLR), 2026

4. Adaptive Defense against Harmful Fine-Tuning for Large Language Models via Bayesian Data Scheduler (Spotlight)

Zixuan Hu, Li Shen, Zhenyi Wang, Yongxian Wei, Dacheng Tao

Conference on Neural Information Processing Systems (NeurIPS), 2025

5. Dynamic Neural Fortresses: An Adaptive Shield for Model Extraction Defense

Siyu Luan∗, Zhenyi Wang∗#, Li Shen, Zonghua Gu, Chao Wu, Dacheng Tao

International Conference on Learning Representations (ICLR), 2025

6. Open-Vocabulary Customization from CLIP via Data-Free Knowledge Distillation (Oral)

Yongxian Wei, Zixuan Hu, Li Shen, Zhenyi Wang, Chun Yuan, Dacheng Tao

International Conference on Learning Representations (ICLR), 2025

7. Defense against Model Extraction Attack by Bayesian Active Watermarking

Zhenyi Wang, Yihan Wu, Heng Huang

International Conference on Machine Learning (ICML), 2024

8. Representation Surgery for Multi-Task Model Merging

Enneng Yang, Li Shen, Zhenyi Wang, Guibing Guo, Xiaojun Chen, Xingwei Wang, Dacheng

Tao

International Conference on Machine Learning (ICML), 2024

9. A Unified and General Framework for Continual Learning

Zhenyi Wang, Yan Li, Li Shen, Heng Huang

International Conference on Learning Representations (ICLR), 2024

10. Improving Non-Transferable Representation Learning by Harnessing Content and Style (Spotlight)

Ziming Hong, Zhenyi Wang, Li Shen, Yu Yao, Zhuo Huang, Shiming Chen, Chuanwu Yang,

Mingming Gong, Tongliang Liu

International Conference on Learning Representations (ICLR), 2024

11. AdaMerging: Adaptive Model Merging for Multi-Task Learning

Enneng Yang, Zhenyi Wang, Li Shen, Shiwei Liu, Guibing Guo, Xingwei Wang, Dacheng Tao

International Conference on Learning Representations (ICLR), 2024

12. Defending against Data-Free Model Extraction by Distributionally Robust Defensive Training

Zhenyi Wang, Li Shen, Tongliang Liu, Tiehang Duan, Yanjun Zhu, Donglin Zhan, David Doer-

mann, Mingchen Gao

Conference on Neural Information Processing Systems (NeurIPS), 2023

13. Distributionally Robust Memory Evolution with Generalized Divergence for Continual Learning

Zhenyi Wang, Li Shen, Tiehang Duan, Qiuling Suo, Le Fang, Wei Liu, Mingchen Gao

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2023

14. Improving Task-free Continual Learning by Distributionally Robust Memory Evolution

Zhenyi Wang, Li Shen, Le Fang, Qiuling Suo, Tiehang Duan, Mingchen Gao

International Conference on Machine Learning (ICML), 2022

15. Towards Faithful Neural Table-to-Text Generation with Content-Matching Constraints

Zhenyi Wang, Xiaoyang Wang, Bang An, Dong Yu, Changyou Chen

Annual Conference of the Association for Computational Linguistics (ACL), 2020