News
SEARCH
ALL NEWS
BIFOLD Research Reveals a Blind Spot in Software Security
Malicious code need not be concealed within software itself. It can be embedded in seemingly harmless documentation. BIFOLD researchers demonstrate how large language models can hide executable functionality in natural-looking files, enabling difficult-to-detect attacks during the software build process. Presented at ACM AsiaCCS 2026.
SIGMOD/PODS 2026 Conference Contributions
At the 2026 ACM SIGMOD/PODS Conference, held in Bengaluru, India, from May 31 to June 5, 2026, BIFOLD researchers will present two papers and receive two prestigious individual awards.
How the Brain Detects Relevance — and How AI Learns from It
The human brain’s ability to filter relevant information from the vast amount of data it continuously receives is known as attention. Researchers at BIFOLD at the Technical University of Berlin, in close collaboration with scientists from the Kording Lab at the University of Pennsylvania, have developed a new brain-inspired AI model of visual attention.
BIFOLD establishes Advisory Board
BIFOLD has officially established its Advisory Board on May 1, 2026. Four internationally renowned researchers in machine learning and data management, Nesime Tatbul, Cecilia Clementi, Yannis Ioannidis, and Masashi Sugiyama, will from now on support the advancement of the institute.
Klaus-Robert Müller Among Top 20 Computer Scientists Worldwide
Research.com's 2026 ranking places BIFOLD Co-Director Prof. Klaus-Robert Müller at 3 in Germany. He also received the Computer Science in Germany Leader Award for the fourth time, following 2022, 2023, and 2025.
Reviewing Girls' Day 2026 at BIFOLD
At Girls' Day 2026, BIFOLD welcomed eleven young women for a hands-on introduction to computer science and IT security. Across three stations, they explored digital forensics, cryptography, and programming, guided by researchers from BIFOLD's MLSec Research Group.
Complexity has to pay off
To test code before deployment, large language models (LLMs) are now frequently used. A research team from BIFOLD has shown in a recent study that the enormous technical effort behind these LLMs does not always pay off. The work was presented at the 48th IEEE/ACM International Conference on Software Engineering (ICSE) 2026.
ICDE 2026 Conference Contributions
BIFOLD researchers will be present at ICDE 2026 in Montréal, Canada (May 4–8), contributing keynotes, research papers, a tutorial, and program committee work at one of the world's leading conferences in data engineering.
Rethinking AI architecture for molecular simulations
Classical machine learning models for molecular dynamics embed physical principles, such as energy conservation and equivariance, directly into their architectures. These inductive biases have long been seen as essential for reliable simulations. A new study challenges this assumption with a surprisingly simple approach.
ICLR 2026 Conference Contributions
BIFOLD researchers will participate in the fourteenth International Conference on Learning Representations (ICLR), contributing seven workshop papers and seven posters.