智能体工作流已成为构建复杂 AI 系统的主导范式，通过编排规划、推理、动作执行和反思等专用组件来处理复杂的现实任务。然而，由于复杂的组件间相互依赖关系以及缺乏有原则的归因方法，系统性地分析和优化这些工作流仍具挑战性。在本研究中，我们提出了 ShapleyFlow，这是首个采用合作博弈论来分析和优化智能体工作流的框架。通过应用 Shapley value 评估所有可能的组件配置，ShapleyFlow 能够实现对每个组件贡献的细粒度归因，并促进识别任务特定的最优配置。通过在导航、数学和 OS 等 7 个场景中评估的构建数据集，我们展示了 3 项关键贡献：(1) 理论框架：一种用于智能体工作流中贡献归因的有原则的博弈论方法。(2) 最优工作流发现：ShapleyFlow 识别出任务特定的组件配置，在所有测试任务中始终优于依赖单一 LLM 的工作流。(3) 综合分析：我们构建并分析了超过 1,500 个任务，为跨多个领域优化工作流提供了可操作的见解和设计指南。
英文摘要：
Agentic workflows have become the dominant paradigm for building complex AI systems, orchestrating specialized components, such as planning, reasoning, action execution, and reflection, to tackle sophisticated real-world tasks. However, systematically analyzing and optimizing these workflows remains challenging due to intricate component interdependencies and the lack of principled attribution methods. In this work, we introduce ShapleyFlow, the first framework that employs cooperative game theory to analyze and optimize agentic workflows. By applying the Shapley value to evaluate all possible component configurations, ShapleyFlow enables fine-grained attribution of each component's contribution and facilitates the identification of task-specific optimal configurations. Through a constructed dataset evaluated across 7 scenarios, such as navigation, math and OS, we demonstrate 3 key contributions: (1) Theoretical Framework: a principled game-theoretic approach for the attribution of contributions in agentic workflows. (2) Optimal Workflow Discovery: ShapleyFlow identifies task-specific component configurations that consistently outperform workflows relying on a single LLM across all tested tasks. (3) Comprehensive Analysis: we construct and analyze over 1,500 tasks, providing actionable insights and design guidelines for optimizing workflows across multiple domains.
GNNs as Predictors of Agentic Workflow Performances
Shanghai Jiao Tong University（上海交通大学）；Beijing Institute for General Artificial Intelligence（北京通用人工智能研究院）；State Key Laboratory of General Artificial Intelligence, BIGAI（北京通用人工智能研究院通用人工智能国家重点实验室）；University of Oxford（牛津大学）；Peking University（北京大学）
Difficulty-Aware Agentic Orchestration for Query-Specific Multi-Agent Workflows
South China Normal University（华南师范大学）；The University of Texas Health Science Center（德克萨斯大学健康科学中心）；The Hong Kong University of Science（香港科学大学）；University of California（加州大学）；Peking University（北京大学）；University of Toronto（多伦多大学）
The Bitter Lesson of Diffusion Language Models for Agentic Workflows: A Comprehensive Reality Check
Southeast University（东南大学）；Alibaba（阿里巴巴）；Southeast University Shenzhen Research Institute（东南大学深圳研究院）；College of Computing and Data Science at Nanyang Technological University（南洋理工大学计算机与数据科学学院）
DeepWisdom（深度智慧）；The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；Renmin University of China（中国人民大学）；Nanjing University（南京大学）；Fudan University（复旦大学）；King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）；Mila（米拉研究所）；The Hong Kong University of Science and Technology（香港科技大学）
Do We Always Need Query-Level Workflows? Rethinking Agentic Workflow Generation for Multi-Agent Systems
State Key Laboratory of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences（人工智能安全国家重点实验室，计算技术研究所，中国科学院）；University of Chinese Academy of Sciences（中国科学院大学）
EvoFlow: Evolving Diverse Agentic Workflows On The Fly
The Chinese University of Hong Kong（香港中文大学）；Tongji University（同济大学）；Wuhan University（武汉大学）；Tsinghua University（清华大学）；Nanyang Technological University（南洋理工大学）；Shanghai AI Laboratory（上海人工智能实验室）
Composing Verifiable Conceptual Models via Building Blocks: Towards Design-Time Verification of Agentic AI Workflows
Team EVERGREEN Inria Centre Inria d’Université Côte d’Azur（法国国家信息与自动化研究所蔚蓝海岸大学中心EVERGREEN团队）；Department of Data Science William & Mary（威廉与玛丽学院数据科学系）；Office of Enterprise Research and Innovation Old Dominion University（欧道明大学企业研究与创新办公室）
A Practical Guide for Designing, Developing, and Deploying Production-Grade Agentic AI Workflows
Old Dominion University（旧 Dominion 大学）；Florida International University（佛罗里达国际大学）；Nanyang Technological University（南洋理工大学）；University of Colombo（科伦坡大学）
A Sober Look at Agentic Misalignment in Automated Workflows
University of Virginia（弗吉尼亚大学）；Georgia Institute of Technology（佐治亚理工学院）；University of Utah（犹他大学）；University of Notre Dame（圣约翰大学）；George Washington University（乔治华盛顿大学）
Agentic workflow enables the recovery of critical materials from complex feedstocks via selective precipitation
Physical and Computational Sciences Directorate, Pacific Northwest National Laboratory（太平洋西北国家实验室物理与计算科学部门）；Department of Chemical Engineering, University of Washington（华盛顿大学化学工程系）
Internet of Agentic AI: Incentive-Compatible Distributed Teaming and Workflow
Department of Electrical and Computer Engineering, Tandon School of Engineering, New York University, Brooklyn, NY, USA（电气与计算机工程系，工程学院，纽约大学，布鲁克林，纽约，美国）
HyEvo: Self-Evolving Hybrid Agentic Workflows for Efficient Reasoning
East China Normal University（华东师范大学）；Beihang University（北京航空航天大学）；Shanghai University of International Business and Economics（上海对外经贸大学）；Fudan University（复旦大学）
