Leonard Salewski
PhD Researcher
Github Linkedin leonard.salewski (at) uni-tuebingen.de Google Scholar


Leonard Salewski is a Ph.D. candidate in the International Max-Planck Research School for Intelligent Systems (IMPRS-IS) under the supervision of Prof. Zeynep Akata and Prof. Hendrik P.A. Lensch. He received his bachelor’s degree in Aerospace Engineering at the University of Stuttgart in 2016 and focused on Information Technology and Flight Control and Systems Technology for his master’s degree in Aerospace Engineering. In 2018 he was a research intern at the Bosch Center for Artificial Intelligence.

His primary research interests lie in the intersection of computer vision and natural language processing as well as in the properties of large language models.

Additionally, he is working on scholarGPT, an academic chatbot, that does not hallucinate its sources, but instead gives reliable and traceable answers based on >2.25M arXiv pre-prints. Further applications for scholarGPT, can be found in law, journalism and education.


Zero-shot audio captioning with audio-language model guidance and audio context keywords.
Leonard Salewski, Stefan Fauth, A. Sophia Koepke and Zeynep Akata
NeurIPS 2023 Machine Learning for Audio Workshop (Oral)
Paper | Code


Zero-shot Translation of Attention Patterns in VQA Models to Natural Language.
Leonard Salewski, A. Sophia Koepke, Hendrik P.A. Lensch and Zeynep Akata
German Conference on Pattern Recognition, 2023
Paper | Code


In-Context Impersonation Reveals Large Language Models' Strengths and Biases.
Leonard Salewski, Stephan Alaniz, Isabel Rio-Torto, Eric Schulz and Zeynep Akata
NeurIPS 2023 (spotlight)
Paper | Poster


Diverse Video Captioning by Adaptive Spatio-temporal Attention.
Zohreh Ghaderi, Leonard Salewski and Hendrik P. A. Lensch
German Conference on Pattern Recognition, 2022
Paper | Code


CLEVR-X: A visual reasoning dataset for natural language explanations.
Leonard Salewski, A. Sophia Koepke, Hendrik P.A. Lensch and Zeynep Akata
Springer Lecture Notes on Artificial Intelligence, 2022
Paper | Project page | Code
This was also presented at the CVPR 2022 Workshop on Explainable AI for Computer Vision (XAI4CV).


e-ViL: A Dataset and Benchmark for Natural Language Explanations in Vision-Language Tasks.
Maxime Kayser, Oana-Maria Camburu, Leonard Salewski, Cornelius Emde, Virginie Do, Zeynep Akata and Thomas Lukasiewicz
IEEE International Conference of Computer Vision, ICCV 2021
Paper | Code


Relational Generalized Few-Shot Learning.
Xiahan Shi, Leonard Salewski, Martin Schiegg, Zeynep Akata and Max Welling
British Machine Vision Conference, 2020
This publication is the result of my master thesis.


For up-to-date information please also check: Semantic Scholar or Google Scholar.

Community Service


  • TPAMI 2021 / 2023
  • MULA 2022
  • CVPR 2023 (Emergency)
  • IJCV 2023 (2x)
  • BMVC 2023 (Emergency)


  • Digital Volunteer for the NeurIPS '21 workshop "ImageNet: Past, Present, Future"


Master Thesis