Procedural Graphs: Self-Evolving Execution Structures for LLM Agents

Hugging Face Daily Papers Papers
procedural-graphs llm-agents self-evolving execution-structures tool-use research-framework

Summary

This paper introduces Procedural Graph, a framework that organizes LLM agent actions into structured triplets for improved long-horizon tool use, with self-evolving topology to enhance performance.

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.
Original Article
View Cached Full Text

Cached at: 09/09/26, 08:33 AM

Paper page - Procedural Graphs: Self-Evolving Execution Structures for LLM Agents

Source: https://huggingface.co/papers/2609.09153

Abstract

A procedural graph framework organizes agent actions into structured relational triplets, providing situational guidance and self-evolving topology to improve long-horizon tool use.

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 theprocedural knowledgeof 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 theProcedural Graph: just as a knowledge graph organizes factual knowledge into (entity, relation, entity) triplets for what-is questions, aProcedural Graphorganizesprocedural knowledgeinto (procedure, relation, procedure) triplets for what-to-do questions. At each decision step, the framework localizes the agent’s active node, and aguidance modeltranslates the surrounding subgraph into step-level situational guidance that biases the solver’s next action without dictating it. The graph is self-evolving: anLLM refinercontrasts 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, theProcedural Graphdelivers consistent gains over memory-based baselines, and self-evolution further improves performance without manual engineering.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2609\.09153

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.09153 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.09153 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.09153 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles