SEBGen: A Self-Evolving Benchmark Generation Framework for Software Engineering Agents
Static software engineering benchmarks such as SWE-bench suffer from three fundamental limitations: task sets are fixed and become saturated as model capabilities rapidly improve; expert construction is costly and slow; and difficulty levels cannot adapt to evolving model abilities. We present SEBGen, a self-evolving b...
Jing-Jing Huang, Zhao-Wen Feng, Dao-De Xiong et al.
· Electronics · 0 citations