Scholar
Yuji Zhang
Google Scholar ID: meTtEDcAAAAJ
Postdoc@University of Illinois at Urbana-Champaign
Citations & Impact
Citations
282
H-index
10
i10-index
10
Publications
20
Co-authors
8
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Published several papers in the field of NLP, including 'Atomic Reasoning for Scientific Table Claim Verification', 'The Law of Knowledge Overshading: Towards Understanding, Predicting, and Preventing LLM Hallucination', and more. For a full list, see his Google Scholar profile.
Research Experience
- Works as a postdoc researcher at UIUC, focusing on NLP, LLM theoretical interpretation, and trustworthiness. Organized and participated in various workshops, talks, and tutorials at conferences like ACL 2025, AAAI 2025, etc.
Education
- Postdoc, University of Illinois at Urbana-Champaign (UIUC), supervised by Prof. Heng Ji and Prof. Chengxiang Zhai
Background
- Research interests include Natural Language Processing (NLP), theoretical interpretation of large language models (LLMs), and trustworthy LLMs. Focuses on understanding the knowledge mechanisms of LLMs, such as how they acquire, store, represent, and utilize knowledge, and leveraging these insights to enhance model reliability and performance. Specifically interested in: 1. Interpreting, predicting, and preventing hallucination through the lens of knowledge interaction; 2. Updating knowledge while preserving model robustness and reliability; 3. Improving knowledge acquisition mechanisms to boost model intelligence.
Miscellany
- Invited to give talks at institutions such as the Chinese Academy of Sciences, University of Texas at Austin, among others, covering topics related to the impact of knowledge overshadowing on large language models and potential solutions.
