A Survey of Bugs in AI-Generated Code

This comprehensive dataset for the Systematic Literature Review (SLR) and the survey responses on "A Survey of Bugs in AI-Generated Code" contains detailed information on selected studies in the SLR. It provides comprehensive information on the studies collected and selected throughout different stages of the review process, which includes detailed records of the selected studies, covering key attributes such as publication details, datasets, programming languages, bug detection methods, identified bug types and descriptions, models used, and mitigation strategies. In addition, the dataset documents the study selection process from initial search results to full-text screening, incorporating metadata from sources such as Google Scholar and Scopus to support transparency and reproducibility. It also summarizes several representative studies and their bug taxonomies that form the foundation of our classification framework. What's more, it outlines the coding guidelines used in our analysis, including the classification criteria for bug types and mitigation strategies.