AgenticGEO：用于生成式引擎优化的自演化智能体系统

中文速览

生成式搜索把传统“争排名”变成了让内容被大模型答案采纳并引用，但现有生成式引擎优化（GEO）方法依赖固定套路、容易过拟合，而且需要大量昂贵的引擎反馈。AgenticGEO将问题视为根据内容特点选择策略的动态控制任务，用MAP-Elites维护一个不断进化且多样化的改写策略库，并训练一个协同进化的轻量评估器来模拟黑箱搜索引擎反馈、挑选策略和规划多轮改写。实验显示，它在两个代表性生成式引擎和三个数据集上击败14种基线，平均提升46.4%，在仅使用41.2%反馈的情况下仍保留98.1%的效果，并能迁移到新领域。其重要性在于，它让内容优化不再依赖一套僵化规则，而是能以较低交互成本适应不同内容和不断变化的搜索引擎。

摘要

Generative search engines represent a transition from traditional ranking-based retrieval to Large Language Model (LLM)-based synthesis, transforming optimization goals from ranking prominence towards 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 engine preference rule distillation that is prone to overfitting. They cannot flexibly adapt to diverse content or the changing behaviors of generative engines. Moreover, effectively optimizing these strategies requires an impractical amount of interaction feedback from the engines. To address these challenges, we propose AgenticGEO, a self-evolving agentic framework formulating optimization as a content-conditioned control problem, which enhances intrinsic content quality to robustly adapt to the unpredictable behaviors of black-box engines. Unlike fixed-strategy methods, AgenticGEO employs a MAP-Elites archive to evolve diverse, compositional strategies. To mitigate interaction costs, we introduce a Co-Evolving Critic, a lightweight surrogate that approximates engine feedback for content-specific strategy selection and refinement, efficiently guiding both evolutionary search and inference-time planning. Through extensive in-domain and cross-domain experiments on two representative engines, AgenticGEO achieves state-of-the-art performance and demonstrates robust transferability, outperforming 14 baselines across 3 datasets. Our code and model are available at: https://github.com/AIcling/agentic_geo.

术语表

Generative search engine

生成式搜索引擎

Large Language Model (LLM)

大语言模型（LLM）

Generative Engine Optimization (GEO)

生成式引擎优化（GEO）

Search Engine Optimization (SEO)

搜索引擎优化（SEO）

Visibility

可见性

Attribution

归因

black-box engine

黑盒引擎

content-conditioned control problem

内容条件控制问题

content-conditioned control policy

内容条件控制策略

MAP-Elites

MAP-Elites

Quality-Diversity (QD) Archive

质量-多样性（QD）档案库

Co-Evolving Critic

协同进化批评器

surrogate critic

代理批评器

self-evolving agentic framework

自演化智能体框架

agentic system

智能体系统

non-stationary black-box environment

非平稳黑盒环境

strategy selection

策略选择

multi-step rewrite

多步重写

inference-time planning

推理时规划

engine feedback

引擎反馈

sparse feedback

稀疏反馈

retrieval-grounded answer synthesis

检索增强答案合成

ranking-based retrieval

基于排序的检索

Search Engine Results Pages (SERPs)

搜索引擎结果页面（SERPs）

PageRank

PageRank

retrieval signals

检索信号

link-analysis signals

链接分析信号

content inclusion

内容纳入