GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks

Hugging Face Daily Papers Papers

Summary

Introduces Graph-as-Policy (GaP), a multi-agent coding harness that generates directed computation graphs from a modular robot skill library and uses parallel simulation to iteratively refine task execution, achieving significantly higher success rates on variable automation tasks compared to baselines.

For robots to work reliably in commercial and industrial applications, can recent advances in agentic coding systems combine interpretable robot programming with the open-world adaptability of model-free policies? We focus on "Variational Automation" (VA), a class of tasks that have larger variations in object geometry and pose than fixed automation. Model-free policies often struggle to close the reliability gap for VA tasks, which must be executed persistently and reliably in commercial and industrial applications. Motivated by prior work on Task and Motion Planning (TAMP) and the Robot Operating System (ROS), we introduce Graph-as-Policy (GaP), a multi-agent coding harness that generates directed computation graphs with perception, planning, and control nodes from a Modular Open Robot Skill Library (MORSL). GaP then generates an internal simulation environment to rehearse task instances with different graphs in parallel to iteratively refine the graph structure and parameters to improve success rates and throughput. Evaluation with 8 new open VA task benchmarks, 4 in-simulation and 4 in real-world, suggests that GaP can achieve success rates that significantly outperform baselines. Details, code, and data can be found online: https://graph-robots.github.io/gap
Original Article
View Cached Full Text

Cached at: 07/07/26, 10:45 PM

Paper page - GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks

Source: https://huggingface.co/papers/2607.05369 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Graph-as-Policy system combines modular robot skills with multi-agent coding to improve reliability in variable automation tasks through parallel simulation refinement.

For robots to work reliably in commercial and industrial applications, can recent advances inagentic coding systemscombine interpretable robot programming with the open-world adaptability ofmodel-free policies? We focus on “Variational Automation” (VA), a class of tasks that have larger variations in object geometry and pose than fixed automation.Model-free policiesoften struggle to close the reliability gap for VA tasks, which must be executed persistently and reliably in commercial and industrial applications. Motivated by prior work onTask and Motion Planning(TAMP) and theRobot Operating System(ROS), we introduce Graph-as-Policy (GaP), a multi-agent coding harness that generatesdirected computation graphswith perception, planning, and control nodes from aModular Open Robot Skill Library(MORSL). GaP then generates aninternal simulation environmentto rehearse task instances with different graphs in parallel to iteratively refine the graph structure and parameters to improve success rates and throughput. Evaluation with 8 new open VA task benchmarks, 4 in-simulation and 4 in real-world, suggests that GaP can achieve success rates that significantly outperform baselines. Details, code, and data can be found online: https://graph-robots.github.io/gap

View arXiv pageView PDFProject pageGitHub13Add to collection

Get this paper in your agent:

hf papers read 2607\.05369

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/2607.05369 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.05369 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.05369 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

Agent as Policy for Robotic Manipulation

arXiv cs.CL

This paper introduces Agent as Policy (AGP), a framework that enables general-purpose agents to directly control physical robots for manipulation tasks without task-specific training, achieving high success rates across various real-world scenarios.

In-Context Robot Learning with VLM Agents

Hugging Face Daily Papers

This paper introduces GPT-Policy, a framework for in-context robot learning using vision-language models, enabling robots to learn from demonstrations without gradient updates. It evaluates the framework in real-robot trials, showing improved task completion.

Geometric Action Model for Robot Policy Learning

Hugging Face Daily Papers

The Geometric Action Model (GAM) repurposes a pretrained geometric foundation model (GFM) as a unified backbone for language-conditioned robot manipulation, achieving higher accuracy, robustness, and efficiency than existing foundation-model-scale baselines across simulation and real-world benchmarks.