智能体
x.com ·
@omarsar0
Google 新论文提出 Procedural Graph：为长程 LLM Agent 显式化过程性知识
该帖介绍 Google 一篇关于长程 agent 记忆的论文。其核心是 Procedural Graph：以 procedure-relation-procedure 三元组存储过程性知识，使 agent 可查询下一步动作及触发条件。框架在每步定位 agent 的活跃节点，由 guidance model 将邻域子图转为步骤级引导，偏置而非强制下一步动作。图可自我改写：LLM refiner 对比失败与成功轨迹，编辑拓扑与属性，仅提交在留出验证集上成立的修改，并保留被拒编辑以避免重复提议。作者称从最小骨架出发可构建与手工设计相当或更优的图，并能修正而非继承有缺陷的专家先验。
AI推荐理由 该帖转述了一篇 Google 论文（Procedural Graph），核心新信息是把 agent 的过程性知识显式化为 procedure-relation-procedure 三元组，并让 LLM refiner 依据失败/成功轨迹对比自动改写图结构，仅在留出验证集上通过的编辑才提交。对长程 agent 记忆与自演化 agent 有直接参考价值。局限：仅为 X 帖转述，未提供论文原文细节、实验规模或复现证据，'match or beat hand-designed ones' 等结论无法独立核实；互动量只作传播信号，不折算为质量分。
78
浏览 46,464
点赞 531
评论 19
转发 76
正文
Another banger paper from Google.
If you build memory for long-horizon agents, this one is worth your time.
(bookmark it)
Really nice to see how knowledge graphs are being explored in creative ways for agents. This has lots of implications for self-evolving agents.
Technical summary below:
Agents usually pick actions by generating over an accumulating history, which leaves the procedural knowledge implicit. As trajectories get longer they lose track of objectives, call tools out of order, and repeat actions that did not work.
The Procedural Graph helps to make that knowledge explicit.
A knowledge graph stores facts as entity-relation-entity triplets. A Procedural Graph stores procedures as procedure-relation-procedure triplets, so the agent can query what to do next and under which conditions.
At each step, the framework localizes the agent's active node, and a guidance model turns the surrounding subgraph into step-level guidance that biases the next action without dictating it.
The graph rewrites itself.
An LLM refiner compares failed trajectories against successful ones and edits the topology and attributes, committing only edits that hold up on held-out validation, and keeping the rejected ones on file so the same change is not proposed twice.
Starting from a minimal skeleton it builds graphs that match or beat hand-designed ones, and it helps to repair a flawed expert priors instead of inheriting them.
Paper: https://academy.dair.ai/papers/procedural-graphs-self-evolving-execution-structures-for-llm-agents-2609.09153