大型語言模型的智能體環境工程：環境建模、合成、評估與應用綜述

摘要

環境作為基於大型語言模型（LLM）的代理在多樣場景中的互動系統，對於驅動模型能力的持續演進具有關鍵作用。儘管其重要性顯著，現有研究仍缺乏系統性的分類與深入分析。本文從環境工程生命週期的角度，系統性地探討當前代理環境的研究，涵蓋其建模、合成、評估與應用。具體而言，本文首先從八個屬性與八個領域的視角介紹代表性環境，詳細分析其發展路徑並凸顯其核心能力。其次，針對自動化環境合成，提出兩種範式，包括符號合成與神經合成。本文同時展示各範式中不同的環境評估方法。第三，從代理-環境共同演化視角探討相應的環境應用。具體而言，本文從四個互補視角描述動態環境中代理演化的主要途徑：以記憶為中心的經驗演化、以編排為中心的工作流程演化、以軌跡為中心的離線演化，以及以探索為中心的在線演化。同時，識別出三種環境演化範式，即神經驅動、難度驅動與規模驅動的方法。最後，討論數個具前景的未來方向，包括環境即服務（Environment-as-a-Service）、多代理環境（Multi-agent Environments）以及神經符號環境（Neural-Symbolic Environments）。

English

Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model capabilities. Despite this importance, existing work lacks a systematic categorization and deep analysis. This paper systematically studies current researches on agentic environments from the perspective of the environment engineering lifecycle, covering their modeling, synthesis, evaluation and application. Specifically, the paper first introduces representative environments from the perspectives of eight attributes and eight domains, providing detailed analyses of their development paths and highlighting their core capabilities. Second, for automated environment synthesis, two paradigms are introduced, such as symbolic synthesis and neural synthesis. This paper also shows different environment evaluation methods in each paradigm. Thirdly, the corresponding environment applications from the perspective of agent-environment co-evolution are discussed. In specific, the paper characterizes the primary pathways for agent evolution in dynamic environments from four complementary perspectives: memory-centric experience evolution, orchestration-centric workflow evolution, trajectory-centric offline evolution, and exploration-centric online evolution. And three paradigms of environment evolution are identified, namely neural-driven, difficulty-driven, and scaling-driven approaches. At last, several promising future directions are discussed, including Environment-as-a-Service, Multi-agent Environments, and Neural-Symbolic Environments.