從靜態模板到動態運行時圖譜：大型語言模型代理的工作流優化綜述

From Static Templates to Dynamic Runtime Graphs: A Survey of Workflow Optimization for LLM Agents

March 23, 2026

作者: Ling Yue, Kushal Raj Bhandari, Ching-Yun Ko, Dhaval Patel, Shuxin Lin, Nianjun Zhou, Jianxi Gao, Pin-Yu Chen, Shaowu Pan

cs.AI

摘要

基於大型語言模型（LLM）的系統正日益普及，其通過構建可執行的運算流程來解決任務，這些流程交織了LLM調用、資訊檢索、工具使用、程式碼執行、記憶體更新與驗證等環節。本綜述回顧了近期關於設計與優化此類流程的方法，我們將其視為能動性計算圖（ACGs）。我們根據流程結構確定的時機來組織文獻，其中「結構」指代組件或智能體的構成、彼此間的依賴關係以及資訊流動方式。這一視角區分了靜態方法（在部署前固定可複用的流程框架）與動態方法（在執行前或執行中針對特定任務選擇、生成或修訂流程）。我們進一步沿三個維度梳理現有研究：結構確定的時機、流程中優化的具體部分，以及指導優化的評估信號（如任務指標、驗證器信號、偏好或軌跡反饋）。同時，我們區分了可複用的流程模板、運行時實例化的具體圖結構，以及執行軌跡，從而將可複用的設計選擇與實際部署的運行結構及實時行為分離。最後，我們提出一種結構感知的評估視角，在任務下游指標之外，兼顧圖層級屬性、執行成本、魯棒性及跨輸入的結構變異性。本文旨在為LLM智能體的流程優化研究提供清晰的術語體系、統一的方法定位框架、更具可比性的文獻視角，以及更可重現的評估標準。

English

Large language model (LLM)-based systems are becoming increasingly popular for solving tasks by constructing executable workflows that interleave LLM calls, information retrieval, tool use, code execution, memory updates, and verification. This survey reviews recent methods for designing and optimizing such workflows, which we treat as agentic computation graphs (ACGs). We organize the literature based on when workflow structure is determined, where structure refers to which components or agents are present, how they depend on each other, and how information flows between them. This lens distinguishes static methods, which fix a reusable workflow scaffold before deployment, from dynamic methods, which select, generate, or revise the workflow for a particular run before or during execution. We further organize prior work along three dimensions: when structure is determined, what part of the workflow is optimized, and which evaluation signals guide optimization (e.g., task metrics, verifier signals, preferences, or trace-derived feedback). We also distinguish reusable workflow templates, run-specific realized graphs, and execution traces, separating reusable design choices from the structures actually deployed in a given run and from realized runtime behavior. Finally, we outline a structure-aware evaluation perspective that complements downstream task metrics with graph-level properties, execution cost, robustness, and structural variation across inputs. Our goal is to provide a clear vocabulary, a unified framework for positioning new methods, a more comparable view of existing body of literature, and a more reproducible evaluation standard for future work in workflow optimizations for LLM agents.