Yiyi Chen is completing her PhD in language model security and multilingual NLP at Aalborg University. Her research focuses on embedding inversion in multilingual and multimodal models, as well as modeling language graphs in linguistic typology. She is currently investigating how humans process cross-modal meaning through eye-tracking experiments, with the aim of contributing to safer and more interpretable multimodal AI. During her PhD, she has published at venues such as ACL, AAAI, CHI, EMNLP, NAACL, and NoDaLiDa, and received the Danish EliteForsk travel grant for talented young researchers. She has a background in philosophy, computer science, and computational linguistics, and has been a visiting scholar at Saarland University and Macquarie University, working at the intersection of multimodal AI, safety, and security.