Flaky Test

Flaky tests, which intermittently pass and fail without code changes, significantly hinder software development and testing. Current research focuses on automatically categorizing flaky tests to understand their root causes, employing machine learning models to predict and even automatically repair them, often leveraging techniques like contrastive learning and large language models. This work aims to improve software reliability and efficiency by reducing the time and effort spent debugging unreliable tests, ultimately leading to more robust and trustworthy software systems.