Your future responsibilities
The goal of this PhD is to further the potential of model-based formal verification of black-box systems by combining model-learning techniques with formal methods. This tight integration of formal methods into the learning process will help in producing high-quality models automatically, and steering the learning process towards areas that are of interest and/or lacking precision during the learning process. Your responsibilities will include:
As a PhD student in the CRYSTALLINE program, you will become a member of the SAL Doctoral College (SAL-DC) that actively fosters networking and collaboration among students and supervisors, as well as the discussion and exchange of ideas in a diverse interdisciplinary setting. For supporting our interdisciplinary environment, there will be dedicated events and platforms including our annual summit, our scientific forum, summer schools, social activity days, etc. This specific PhD position will be organized in cooperation with Graz University of Technology, Bernhard Aichernig serving as the main supervisor from TU Graz. The doctoral degree will be awarded within the doctoral school of computer science at TU Graz. Co-supervision will be performed Martin Leucker from Lübeck University.
- Research and development of novel model-learning approaches and their integration into the model-learning library AALpy (GitHub - DES-Lab/AALpy).
- Investigating the step from a learned model towards formal verification techniques such as invariant detection, model-checking, monitoring and shielding.
- Systems under verification include machine-learned systems, LLM-generated code, autonomous agents, embedded systems.
- Development and demonstration of a black-box verification workflow for one or two selected verification techniques in combination with model learning.
- Scientific collaboration with renowned universities and research organizations.
- Scientific publications in high-impact journals and presentations at major conferences.
- Possible specialization topics include LLMs, real-time, non-functional properties and verification of machine-learned agents.
As a PhD student in the CRYSTALLINE program, you will become a member of the SAL Doctoral College (SAL-DC) that actively fosters networking and collaboration among students and supervisors, as well as the discussion and exchange of ideas in a diverse interdisciplinary setting. For supporting our interdisciplinary environment, there will be dedicated events and platforms including our annual summit, our scientific forum, summer schools, social activity days, etc. This specific PhD position will be organized in cooperation with Graz University of Technology, Bernhard Aichernig serving as the main supervisor from TU Graz. The doctoral degree will be awarded within the doctoral school of computer science at TU Graz. Co-supervision will be performed Martin Leucker from Lübeck University.