两篇新研究论文探讨了提示词优化对于大型语言模型（LLMs）的有效性和可解释性。第一篇论文 iPOE 介绍了一种使用自动生成的标注决策指南来使提示词优化透明化并提高性能高达 39% 的方法。第二篇论文分析了提示词优化有时为何会失败，发现某些类型的编辑会负面影响推理任务，而另一些则会改善它们，这表明需要进行面向任务的优化器设计。 AI
影响 这些论文通过改进提示词工程和理解当前优化方法的局限性，为提高 LLM 性能提供了见解。
排序理由 该集群包含两篇讨论 LLM 提示词优化技术的学术论文。
在 arXiv cs.NE (Neural & Evolutionary) 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2605.18113v2 Announce Type: replace Abstract: Prompt optimization has often been framed as a discrete search problem to find high-performing and robust instructions for an LLM. However, the search result might not make it transparent why and where specific prompt changes le…
arXiv:2605.26655v1 Announce Type: new Abstract: Automated prompt optimization methods (e.g., DSpy, TextGrad) can substantially improve the performance of large language model (LLM), however, their generalization ability across different tasks remains underperformed. In practice, …
Automated prompt optimization methods (e.g., DSpy, TextGrad) can substantially improve the performance of large language model (LLM), however, their generalization ability across different tasks remains underperformed. In practice, the superiority of the optimized prompt on one b…