arXiv · AI· Jiajun Wu, Leixin Sun, Zihan Tan, Yitao Liu, Shuo Li, Jiaru Qian, Yuxin Wu, Shanghaoran Quan, Chuangxin Zhao, Yangxu Liao, Yang Liu, Bin Chong, Guancheng Wan·· 2026-10-06AI 评分28
SWE-Prometheus:衡量真实代码仓库的工程治理改进
SWE-Prometheus: Measuring Engineering Governance Improvements in Real-World Repositories
AI 导读
研究者提出 SWE-Prometheus,一个面向仓库工程治理改进的基准,包含 60 个仓库,通过配对证据、干净环境探测、行为门控和两位独立教师评分评估六个治理维度。
来源:arXiv · AI · arxiv.org