Xiaozhi Wang

Xiaozhi Wang

Assistant Professor at SIGS, Tsinghua University

About me

I am a tenure-track assistant professor at the Shenzhen International Graduate School, Tsinghua University. Previously, I received my Ph.D. and Bachelor at the Department of Computer Science and Technology, Tsinghua Univeristy in 2025 and 2020. I was fortunately advised by Prof. Juanzi Li and also received mentorship from Prof. Zhiyuan Liu. I visited BLENDER Lab@UIUC and Mila in 2024 and 2019, where I fortunately worked with Prof. Heng Ji and Prof. Jian Tang. My CV is availiable here.

My research spans the areas of natural language processing, machine learning, and knowledge engineering. The research directions I am fascinated in and working on include:

  1. Understanding Lanaguge Models (Mechanistic Interpretability)
    • Understanding the inner workings of large models to explain their behaviors while exploring the scientific laws of artificial intelligence.
  2. Scientific Improvements in Large Models
    • Enhancing the architecture, training, and evaluation of large models based on the mechanistic understandings, thereby fundamentally improving the knowledge, efficiency, safety, and autonomy of large models.

Prospective Students: The group continuously seeks self-motivated Ph.D. and Master students, Postdocs, and Interns. Research topics include but not limited to mechanistic interpretability, LLM safety, and data-centric methods. If you are interested, please drop an email to xzwang@sz.tsinghua.edu.cn with your CV attached.

News

  • [Feb. 2025] Release a tookit OpenSAE for efficiently training sparse auto-encoders and interpreting LLMs. Technical report coming soon.
  • [Feb. 2025] The Precise Neuron-level Knowledge Editing got accepted at ICLR 2025.
  • [Oct. 2024] Will attend EMNLP in person. See you in Miami.
  • [May 2024] Got two papers accepted at ACL2024. See you in Bangkok!
  • [Mar. 2024] Start visiting BLENDER Lab@UIUC.
  • [Dec. 2023] The Robust Evaluation for Open IE paper was selected as outstanding paper of EMNLP.
  • [Oct. 2022] Release a nice event extraction toolkit OmniEvent. Welcome to try it!

Highlighted Publications

Please refer to publications or my Google Scholar for the full list.

  • Xiaozhi Wang, Hao Peng, Yong Guan, Kaisheng Zeng, Jianhui Chen, Lei Hou, Xu Han, Yankai Lin, Zhiyuan Liu, Ruobing Xie, Jie Zhou, Juanzi Li. MAVEN-Arg: Completing the puzzle of all-in-one event understanding dataset with event argument annotation. ACL 2024 [pdf] [code & data]
  • Xiaozhi Wang*, Kaiyue Wen*, Zhengyan Zhang, Lei Hou, Zhiyuan Liu, Juanzi Li. Finding Skill Neurons in Pre-trained Transformer-based Language Models. EMNLP 2022 [pdf] [code]
  • Xiaozhi Wang*, Yulin Chen*, Ning Ding, Hao Peng, Zimu Wang, Yankai Lin, Xu Han, Lei Hou, Juanzi Li, Zhiyuan Liu, Peng Li, Jie Zhou. MAVEN-ERE: A Unified Large-scale Dataset for Event Coreference, Temporal, Causal, and Subevent Relation Extraction. EMNLP 2022 [pdf] [code] [CodaLab]
  • Hao Peng*, Xiaozhi Wang*, Shengding Hu, Hailong Jin, Lei Hou, Juanzi Li, Zhiyuan Liu, Qun Liu. COPEN: Probing Conceptual Knowledge in Pre-trained Language Models. EMNLP 2022 [pdf] [code] [CodaLab]
  • Xiaozhi Wang, Tianyu Gao, Zhaocheng Zhu, Zhengyan Zhang, Zhiyuan Liu, Juanzi Li, Jian Tang. KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation. Transactions of the Association for Computational Linguistics (TACL), 2021. [pdf] [code] [dataset] (ESI Highly Cited Paper, TACL top 10 most cited paper)
  • Xiaozhi Wang, Ziqi Wang, Xu Han, Wangyi Jiang, Rong Han, Zhiyuan Liu, Juanzi Li, Peng Li, Yankai Lin, Jie Zhou. MAVEN: A Massive General Domain Event Detection Dataset. EMNLP 2020 [pdf] [code] [CodaLab] [leaderboard]

Professional Services

Student Mentoring

I am fortunate to have been mentored by several wonderful advisors and also feel obiligated to pass on the support. If you are a junior student interested in collaborating or seeking career advice, please just reach out and schedule a meeting.

Below is a (non-exhaustive) list of students who have collaborated with me.

Ziqi Wang (THU -> UIUC PhD)

Hao Peng (THU -> THU PhD)

Kaiyue Wen (THU -> Stanford PhD)

Yulin Chen (THU -> NYU Shanghai PhD)

Jianhui Chen (THU ->)