Portrait of Sebastian Schuster

Sebastian Schuster

Assistant Professor


Faculty of Computer Science
University of Vienna


I am an assistant professor in natural language processing at the University of Vienna, where I head a WWTF-funded research group. I am also affiliated with the Linguistics Department at University College London.

Before that, I was a postdoc at Saarland University; a postdoc in Linguistics and Data Science at NYU as part of the 2020 Computing Innovation Fellow program, mentored by Tal Linzen; and a PhD student in Linguistics at Stanford and a member of the ALPS lab and the Stanford NLP Group, working with Judith Degen and Chris Manning.

My main research interests are in the evaluation of large language models and in computational semantics and pragmatics. I aim to develop more reliable natural language understanding systems and to develop psycholinguistic accounts of pragmatic language use. Please see my publications to find out more.

Recent News:

  • June 2025: New preprint on RExBench, an evaluation framework for LLM agents to assess their ability to autonomously implement research extensions.
  • April 2025: I started my new job as Assistant Professor in Natural Language Processing at the University of Vienna.
  • June 2024: New preprint on the effect of code-pretraining on entity tracking abilities in LLMs (joint work with Najoung Kim and Shubham Toshniwal).
  • June 2024: Our paper on scope ambiguities in language models appeared in TACL (joint work with Gaurav Kamath, Sowmya Vajjala, and Siva Reddy).
  • Spring 2024: Invited talks at ETH Zurich and Oxford.
  • May 2023: Najoung Kim and I presented our work on entity state tracking at ACL 2023.
  • April 2023: Our paper on variability in scalar implicature derivations was accepted to TACL (joint work with Jennifer Hu, Roger Levy, and Judith Degen).
  • Spring 2023: Area Chair for the "Psycholinguistics, cognitive linguistics and semantic processing" track at *SEM2023.