LLMORPH: Automated Metamorphic Testing of Large Language Models

Collection of recent research tooling and empirical studies focused on testing, robustness, and security of AI-enabled and traditional software systems. Highlights include LLMORPH (metamorphic testing for LLM-based NLP tasks), LLMLOOP (iterative refinement of LLM-generated code and tests), a Detect–Repair–Verify study and EduCollab benchmark for evaluating vulnerability detection/repair on LLM-generated web apps, PerturbationDrive for ADAS robustness testing, generation of abstract penetration test cases from architecture models, a taxonomy and dataset of logging smells in ML code (including ‘