研究人员开发了一种新方法，以提高基于大语言模型（LLM）的电力系统代码生成的可靠性，特别是在本地部署场景下。该方法通过引入一个名为PowerCodeBench的基准生成器和一种边界感知干预技术，来解决API知识边界错误，例如函数名称或参数不正确的问题。这种干预结合了API需求估计与文档注入和修正，显著提高了各种开源和商业LLM的准确性。 AI
影响 增强了LLM在关键基础设施代码生成方面的可靠性，实现了更安全的本地部署。
排序理由 该集群包含一篇详细介绍LLM代码生成新方法和基准的研究论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.31478v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used to automate power-system analysis, but many utilities and energy-research labs require on-premise serving for confidentiality, regulatory, reproducibility, and cost reasons. This …
Large language models (LLMs) are increasingly used to automate power-system analysis, but many utilities and energy-research labs require on-premise serving for confidentiality, regulatory, reproducibility, and cost reasons. This makes the reliability of open-weight models a depl…