📄 中文摘要
生成搜索引擎标志着从传统的基于排名的检索向基于大型语言模型（LLM）的合成转变，优化目标从排名显著性转向内容包含性。生成引擎优化（GEO）旨在通过战略性地操控源内容，最大化黑箱摘要输出中的可见性和归属感。然而，现有方法依赖于静态启发式、单一提示优化或引擎偏好规则的提炼，容易导致过拟合，无法灵活适应多样化内容或生成引擎的变化行为。此外，有效优化这些策略需要不切实际的交互反馈量。该研究提出了一种自我进化的代理系统，旨在克服这些局限性，实现更高效的生成引擎优化。
AgenticGEO：一种自我进化的代理系统用于生成引擎优化
📄 English Summary
AgenticGEO: A Self-Evolving Agentic System for Generative Engine Optimization
Generative search engines signify a shift from traditional ranking-based retrieval to synthesis based on Large Language Models (LLMs), changing optimization goals from ranking prominence to content inclusion. Generative Engine Optimization (GEO) specifically aims to maximize visibility and attribution in black-box summarized outputs by strategically manipulating source content. However, existing methods rely on static heuristics, single-prompt optimization, or distillation of engine preference rules, which are prone to overfitting and lack flexibility to adapt to diverse content or changing behaviors of generative engines. Moreover, effectively optimizing these strategies requires an impractical amount of interaction feedback. This study introduces a self-evolving agentic system designed to overcome these limitations and achieve more efficient generative engine optimization.