Welcome to the Schütze Lab

We are a dynamic research group at the Center for Information and Language Processing at Ludwig Maximilian University Munich, under the supervision of Prof. Hinrich Schütze. Our research areas include:

  • Large Language Models (LLMs): We explore the behavior, structure, and potential of LLMs, examining their capabilities, biases, and self-assessment mechanisms to improve reliability and interpretability.
  • Knowledge Expansion in NLP Models: We investigate how models can acquire and integrate new knowledge over time, using techniques that help improve their comprehension and generation abilities.
  • Representation Learning and Interpretability: We study how language models represent linguistic and conceptual information by analyzing neurons and internal circuits to better understand and refine model behavior.
  • Multilingual NLP: We address challenges in processing and evaluating multiple languages by developing benchmarks and methods for multilingual evaluation, including work on low-resource languages.
  • Intersection of NLP and Robotics: We integrate language understanding into robotic systems to enable natural, adaptable interaction in multimodal environments.

You can find resources (data, code, repositories) released by CIS lab members on our GitHub page and HuggingFace page.

We are grateful for funding from the Munich Center for Machine Learning (MCML), the Elitenetzwerk Bayern (International Doctorate Program), the Deutsche Forschungsgemeinschaft (DFG), and the European Research Council (ERC Advanced Grant).

News

01 December, 2024

We have 2 papers accepted by COLING 2025! Congrats to all the authors!

15 October, 2024

We have 7 papers accepted by EMNLP 2024! Congrats to all the authors!

15 June, 2024

We have 5 papers accepted by ACL 2024! Congrats to all the authors!

15 April, 2024

We have 2 papers accepted by NAACL 2024! Congrats to all the authors!

18 January, 2024

We have 3 papers accepted by EACL 2024! Congrats to all the authors!

15 November 2023

We have 8 papers accepted by EMNLP 2023! Congrats to all the authors!

... see all News