Summary
Alina Leidinger is a final-year PhD student at the University of Amsterdam researching robustness, bias, and values in Large Language Models, with five first-author publications at top AI/NLP venues and media coverage including TechCrunch. She combines applied linguistics and ML engineering to analyze LLM instability and stereotyping, drawing on ethics, logic, and media studies for interdisciplinary methods. Her work includes collaborative projects with Hugging Face and a visiting research stint at LMU Munich tracing bias from pretraining data through alignment. With a strong mathematical background from TUM and Imperial College and hands-on industry experience at BMW, she bridges rigorous theory and practical evaluation of model safety. Notably, she not only develops experiments and datasets but has turned her findings into public-facing impact through publications and presentations.
2 years of coding experience
2 years of employment as a software developer
BSc Mathematics, Mathematics, First Class Honours, BSc Mathematics, Mathematics, First Class Honours at Imperial College London
MSc Mathematics in Data Science, Mathematics and Computer Science, High Distinction, MSc Mathematics in Data Science, Mathematics and Computer Science, High Distinction at Technische Universität München
European Baccalaureate, Abitur, 1.0 (top grade), 94.8/100, European Baccalaureate, Abitur, 1.0 (top grade), 94.8/100 at European School Munich
German, English, French, Italian, Dutch