I’m a PhD student in the colabPotsdam, supervised by David Schlangen. My work is part of the SFB 1287, which tries to identfy the limits of variability in language from various angles. In our project, we interpret these limits as constraints imposed upon a next utterance by the preceeding context and study these different notions of coherence in more detail. I currently focus on the evaluation and analysis of large pre-trained language and dialogue models in terms of how these notions are (not yet) encoded and how they could be used to improve these models.

I have always been fascinated by languages and thanks to a series of lucky coincidences, I ended up studying Computational Linguistics at the University of Potsdam.
After finishing my Bachelor, I gathered some work experience in a translation company trying to figure out how machine translation research could be applied to the “real world” in order to actually assist human translators.
The rise of artificial neural network models brought me back to university and I completed my Master at the LMU in Munich while working at the Munich Center for Machine Learning on a project on representation learning for natural languages.
Having grown up in Berlin, I soon felt the urge to leave the south again, though, and now I’m happily back at where it all began (department-wise and thematically), focussing on the intersection between Linguistics and Natural Language Processing.