About Me

I am a PhD student at the chair for Responsible Data Science under the supervision of Prof. Dr. Gjergji Kasneci. I obtained my Bachelor’s and Master’s degrees from Ludwig-Maximilians-University Munich, where I've developed theoretical and hands-on expertise around Natural Language Processing. Specific projects I have worked on include named entity recognition of interactions between different medications and the investigation of embedding properties of large language models in the context of semantic similarity search.

Research Focus

Currently I am interested in the fields of robustness and uncertainty estimation in deep neural networks. Robustness generally aims to make models stable and reliable, concerning itself with topics such as adversarial attacks or out-of-distribution generalization. Uncertainty quantification on the other hand deals with a system's confidence in its output, which plays a big role in its trustworthiness and is therefore an important field to address in the scope of responsible AI.

Publications

"Understanding Knowledge Drift in LLMs through Misinformation." Alina Fastowski and Gjergji Kasneci. 2024. DELTA workshop at KDD, 2024.

“Exploring Anisotropy and Outliers in Multilingual Language Models for Cross-Lingual Semantic Sentence Similarity.“ Katharina Hämmerl, Alina Fastowski, Jindřich Libovický, and Alexander Fraser. 2023. Findings of ACL, 2023.