Can Large Language Models Improve SE Active Learning via Warm-Starts? (2501.00125v1)

Abstract: When SE data is scarce, "active learners" use models learned from tiny samples of the data to find the next most informative example to label. In this way, effective models can be generated using very little data. For multi-objective software engineering (SE) tasks, active learning can benefit from an effective set of initial guesses (also known as "warm starts"). This paper explores the use of LLMs for creating warm-starts. Those results are compared against Gaussian Process Models and Tree of Parzen Estimators. For 49 SE tasks, LLM-generated warm starts significantly improved the performance of low- and medium-dimensional tasks. However, LLM effectiveness diminishes in high-dimensional problems, where Bayesian methods like Gaussian Process Models perform best.