Code Intelligence Task

Code intelligence tasks aim to leverage machine learning, particularly large pre-trained language models (PLMs), to automate various programming-related activities like code generation, summarization, and defect prediction. Current research focuses on improving the efficiency and robustness of these models, exploring techniques like prompt engineering (e.g., genetic algorithms for automatic prompt generation) and adapter tuning to reduce computational costs and mitigate vulnerabilities to adversarial attacks. These advancements are significant because they enhance the accuracy and reliability of code analysis tools, potentially leading to improved software development practices and increased software security.

Papers