Dawn Song - arXivDaily 作者

2505.23135 2026-03-18 cs.LG cs.AI cs.LO cs.PL cs.SE

VERINA: Benchmarking Verifiable Code Generation

VERINA：可验证代码生成基准测试

Zhe Ye, Zhengxu Yan, Jingxuan He, Timothe Kasriel, Kaiyu Yang, Dawn Song

机构 * University of California, Berkeley（加州大学伯克利分校）； Meta FAIR

AI总结本文提出VERINA基准测试，用于全面评估代码、规范及证明生成能力，揭示LLM在可验证代码生成中的挑战，特别是证明生成的困难。

2603.15798 2026-03-18 cs.AI

CUBE: A Standard for Unifying Agent Benchmarks

CUBE：统一智能体基准的标准化

Alexandre Lacoste, Nicolas Gontier, Oleh Shliazhko, Aman Jaiswal, Kusha Sareen, Shailesh Nanisetty, Joan Cabezas, Manuel Del Verme, Omar G. Younis, Simone Baratta, Matteo Avalle, Imene Kerboua, Xing Han Lù, Elron Bandel, Michal Shmueli-Scheuer, Asaf Yehudai, Leshem Choshen, Jonathan Lebensold, Sean Hughes, Massimo Caccia, Alexandre Drouin, Siva Reddy, Tao Yu, Yu Su, Graham Neubig, Dawn Song

AI总结本文提出CUBE标准，通过MCP和Gym构建统一协议，实现智能体基准的标准化，减少碎片化，提升研究效率。

Comments Position paper. 10 pages. Reference implementation: https://github.com/The-AI-Alliance/cube-standard