arXiv · cs.AI · Rujun Han, Yufan Zhuang, Arijit Ray, Han Yu, Zhongying CuiZhu, George Lee, Yifei Ming, Binz Roy, Varun Sakalkar, Vishy Tirumalashetty, Burak Gokturk, Tomas Pfister, Chen-Yu Lee · 9 小时前模型评分57
arXiv 发布 PLE 基准，测试语言模型能否完成真实物理工程交付
PLE: Physical-Engineering's Last Exam
导读
研究团队发布 PLE（Physical-Engineering's Last Exam）基准，包含 85 个长周期智能体任务，覆盖电气、土木、结构和机械工程，解题即完成一份可计费的真实工程交付物。
分类置信度
中 · 待人工复核
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
应用领域：制造与工业、科学研究
事件类型：研究发布
证据摘录
- “We introduce PLE, a benchmark of 85 long-horizon, agentic tasks in professional physical engineering consisting of electrical, civil, structural and mechanical engineering, where solving a task means completing a realistic, billable engineering deliverable.”
- “Even the strongest configuration reaches a mean score of only 50.6%, while spending 27M tokens and 63 minutes per task on average, demonstrating the substantial gap between current agents and autonomous physical engineering design.”
来源：arXiv · cs.AI · arxiv.org