Scholar
Jonas Geiping
Google Scholar ID: 206vNCEAAAAJ
ELLIS Institute Tübingen & Max Planck Institute for Intelligent Systems
Citations & Impact
Citations
9,130
H-index
38
i10-index
51
Publications
20
Co-authors
52
list available
Publications
Resume (English only)
Background
- Machine Learning researcher based in Tübingen, Germany, leading the research group for safety- & efficiency-aligned learning.
- Fascinated by questions of safety and efficiency in modern machine learning.
- On safety: investigates data poisoning, jailbreaks, adversarial attacks, watermarking for generative models, privacy guarantees, and the technical definition of 'safety'.
- On efficiency: studies systems that do more with less, including weight averaging and recursive computation, with a focus on reasoning in intelligent systems and efficient language modeling.
- Core research questions: Can models reason well without sacrificing safety? How do computational constraints affect safety? Can intelligence and safety reinforce each other?
- Main research areas: Safety, security, and privacy in ML; understanding and implementing reasoning in intelligent systems; efficient ML (especially language modeling); deep learning as a natural science.
