About Me

I am a first-year PhD student at the Chair of Responsible Data Science at TUM, supervised by Prof. Dr. Gjergji Kasneci. I hold a Bachelor’s degree in Physics from Wuhan University and Université Lyon 1, as well as a Master’s degree in Digital Humanities from École Polytechnique Fédérale de Lausanne (EPFL).

During my Master’s studies, I conducted research at the National Institute of Informatics (Japan), EPFL, and ETH Zurich, working on a diverse range of topics, including Computational Social Science, Point Cloud Semantic Segmentation, Contrastive Learning for Image Similarity Detection, and Single-Image 3D Reconstruction with Stable Diffusion. Alongside my academic experience, I worked for a year as a Data Scientist and Machine Learning Engineer. In general, I love to do research that helps to solve complex real-world problems.
 

Research Focus

My research focuses on enhancing the reliability, robustness, and safety of AI models, with a particular interest in LLM robustness, factuality, and adversarial resilience. Currently, I am developing a lightweight language model designed to generate factually reliable text summaries, while also exploring methods to prevent LLMs from being manipulated into producing misleading or harmful content. My broader interests include evaluating and mitigating risks in AI systems and adversarial robustness to ensure safer and more trustworthy AI deployment.