Torge Hinrichs
Researcher at the
Institute of Software Security
at
Hamburg University of Technology(TUHH).
About
I am a researcher at the Hamburg University of Technology (TUHH), working at the intersection of software security, software engineering, and machine learning.
My research focuses on the static analysis, prediction, and localization of software vulnerabilities, as well as mining vulnerability-contributing code changes and studying AI-assisted secure software development.
My work includes empirical research on tracing vulnerability-introducing commits in Java projects, investigating how task phrasing affects vulnerability detection with large language models, and contributing to online programming platforms for software engineering education. I am passionate about writing, reading, and analysing source code - not only as a practical research tool, but as a primary source of evidence for understanding how software evolves, fails, and can be made more secure.
Selected papers (Full list on Google Scholar)
-
VULTRITION: Nutrition Labels for Vulnerability Datasets
-
Back to the Roots: Assessing Mining Techniques for Java Vulnerability-Contributing Commits
-
Beyond Prompting: The Role of Phrasing Tasks in Vulnerability Prediction for Java