Cheng Yu, M.Eng.
Doctoral Candidate / Research Associate
Academic Career and Research Areas
Cheng Yu is a doctoral researcher at the Technical University of Munich (TUM), where she is affiliated with the Chair of Societal Computing. Her research focuses on the development and evaluation of intelligent systems, with particular emphasis on large language models and AI agents. Her work seeks to advance the understanding of how AI systems can be designed to be more transparent, robust, and aligned with human values, and lies at the intersection of machine learning and computational social science.
Cheng Yu holds both a Bachelor’s and a Master’s degree in Computer Science from ShanghaiTech University. In her previous work, she focused on the pre-training of speech language models.
Key Publications
C Yu, S Engelmann, R Cao, D Ali, O Papakyriakopoulos “How Should AI Safety Benchmarks Benchmark Safety?” arXiv preprint arXiv:2601.23112, 2026
C Yu, B Stroebl, D Yang, O Papakyriakopoulos “Information Retrieval Induced Safety Degradation in AI Agents” The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025
