Cass Zhixue Zhao

Dr. Cass Zhixue Zhao is a Lecturer (Assistant Professor) in Natural Language Processing at the Department of Computer Science, University of Sheffield. She serves as Student Services Liaison (Library, ELTC) and leads research in model interpretability, hallucination, and compression. Her work bridges theoretical and applied NLP, with a focus on trustworthy and efficient large language models (LLMs).

• PhD from University of Sheffield (funded by university scholarships)
• Postdoc researcher (2020–2024) on model interpretability and responsible AI
• Co-PI for ExU project (2023–2025), funded by European Media and Information Fund (€399,926)

Her research prioritizes enabling trustworthy NLP models through:

• Model explanation frameworks
• Hallucination and bias mitigation
• Efficient model compression techniques
• Knowledge editing detection and reversal
• Multilingual disinformation analysis
• Scientific text-to-image generation

Recent publications (2025–2024) demonstrate expertise in:

• Formal logic reasoning for LLMs
• Security risks in knowledge editing
• Multilingual hallucination consistency
• FAITH metrics for multilingual models
• Pruned model hallucination analysis
• Visual knowledge reasoning pipelines

Dr. Zhao actively supervises PhD students, with a UKRI-funded position available in 2026. She contributes to program committees (AAAI, ACL, EMNLP) and maintains active collaborations with researchers like Nikolaos Aletras, Paul Youssef, and Lucie Flek.