This work explores the capabilitity of state-of-the-art LLMs, including closed-source options like OpenAI’s GPT-4 and open-source alternatives like Meta AI’s Codellama, to automatically generate tests and use these tests to validate and verify compiler implementations of a directive-based programming paradigm, OpenACC. We published a paper in FGCS Journal: https://arxiv.org/abs/2310.04963 For any questions, contact
Christian Munley - chrismun@udel.edu
Sunita Chandrasekaran - schandra@udel.edu