[论文] Procedural Graphs: Self-Evolving Execution Structures for LLM Agents

论文概要

研究领域: cs.AI, cs.CL, cs.MA 作者: Yuxing Lu, Yicheng Chen, Shanchan Wu, Sercan Ö. Arık 发布时间: 2026-09-08 arXiv: 2609.09153

中文摘要

大型语言模型越来越多地被部署为在长时程上规划并通过外部工具执行动作的智能体。大多数智能体通过在累积历史记录上进行无约束生成来选择动作，使得做什么、按什么顺序做、在什么条件下做的程序性知识保持隐式。随着轨迹延长，智能体可能丢失目标、乱序调用工具、重复无效动作。我们引入程序图：正如知识图将事实知识组织成（实体、关系、实体）三元组以回答'是什么'问题，程序图将程序性知识组织成（程序、关系、程序）三元组以回答'做什么'问题。在每个决策步骤，框架定位智能体的活跃节点，指导模型将周围子图转化为步骤级情境指导，偏置求解器的下一步动作而不直接规定它。该图是自进化的：LLM细化器对比失败轨迹与成功轨迹并编辑图的拓扑和属性，提交保留或提升留出验证性能的编辑，同时保留被拒绝的编辑以防止重复。从最小骨架开始，该循环构建出与手工设计相当或更优的图。它还能修复有缺陷的专家先验。在多个数据集、任务类型和LLM上，程序图持续优于基于记忆的基线，且自进化无需人工工程即可进一步提升性能。

原文摘要

Large language models are increasingly deployed as agents that plan over long horizons and act through external tools. Most agents select actions through unconstrained generation over an accumulating history, leaving implicit the procedural knowledge of what to do, in what order, and under which conditions. As trajectories lengthen, agents can lose track of their objectives, invoke tools out of order, and repeat unproductive actions. We introduce the Procedural Graph: just as a knowledge graph organizes factual knowledge into (entity, relation, entity) triplets for what-is questions, a Procedural Graph organizes procedural knowledge into (procedure, relation, procedure) triplets for what-to-do questions. At each decision step, the framework localizes the agent's active node, and a guidance model translates the surrounding subgraph into step-level situational guidance that biases the solver's next action without dictating it. The graph is self-evolving: an LLM refiner contrasts failed trajectories with successful ones and edits the graph's topology and attributes, committing edits that preserve or improve held-out validation performance while retaining rejected ones to discourage repetition. Starting from a minimal skeleton, the loop builds graphs that match or surpass hand-designed ones. It can also repair a flawed expert prior. Across multiple datasets, task types, and LLMs, the Procedural Graph delivers consistent gains over memory-based baselines, and self-evolution further improves performance without manual engineering.

---

*自动采集于 2026-09-10*

#论文 #arXiv #AI #小凯