Large Language Models Adaptation for Cyber-Physical System Testing
Dive into the challenge of testing Cyber-Physical Systems (CPSs) by optimizing and leveraging the potential of Large Language Models (LLMs).
This research project proposes incorporating Large Language Models (LLMs) with advancing technologies to enhance testing methods of Cyber-Physical Systems (CPS), such as self-driving cars and autonomous ships. With their natural language processing and human-like reasoning, LLMs offer the potential to handle diverse and complex tasks, improve test coverage, and evaluate the realism of a test scenario. Nevertheless, LLMs have issues, such as hallucinations and knowledge redundancy, which can compromise their reliability and trustworthiness. This master's project aims to develop methods to reduce the limitations of LLMs and find methods to incorporate LLMs with other advanced ML and testing techniques. You can choose between one of the following topics: Topic 1: Enhancing LLMs by limiting their issues: You will focus on developing methods to reduce the known limitations of LLMs. Topic 2: LLM combined with advanced techniques: You will focus on combining LLMs with advancing ML and AI techniques and system testing methods. Topic 3: Testing CPS with LLMs. You will focus on developing efficient methods and a guideline to test CPS with LLMs. You have the freedom to choose a CPS that you find interesting.
Goal
The primary goal of this master project is to utilize Large Language Models in conjunction with advanced technologies to improve the testing methods of Cyber-Physical Systems.
Learning outcome
Advanced knowledge and deeper understanding about:
- Large Language Models
- Applied Machine Learning and Artificial Intelligence
- Testing Cyber-Physical Systems
Qualifications
- Programming skills
- Knowledge of Machine Learning frameworks and Large Language Models
Supervisors
- Karoline Nylænder
- Shaukat Ali
