EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics

Hugging Face Daily Papers Papers
coding-agents robotics long-horizon dexterous-manipulation simulation-benchmark vla-training

Summary

The paper introduces EmbodiedSWE, a framework using coding agents to solve complex, long-horizon dexterous robotics tasks and generate demonstrations for training robot policies via a simulation benchmark.

We study coding agents for long-horizon, dexterous robotics and ask whether their solutions can provide scalable supervision for learning general robot policies. To test this, we develop EMBODIEDSWE-BENCH, a simulation benchmark for coding agents spanning contact-rich manipulation, deformable objects, and long-horizon tasks requiring up to half an hour of continuous interaction. We find that frontier coding agents can solve complex long-horizon tasks and transfer prior solutions across both tasks and embodiments. We also design supporting tools that help agents more effectively solve these tasks. However, the resulting solutions require substantial iterative interaction and are typically specialized to individual task instances. We therefore introduce EMBODIEDSWE-GEN, which expands a single solution from coding agent into large diverse trajectories for training a VLA. VLA performance improves with more generated demonstrations, and agent-aided diversification improves generalization to held-out task variations. We also show that a VLA finetuned solely on coding-agent-generated simulation demonstrations completes a long-horizon task on real robot. Together, our framework uses coding agents to solve complex robotics tasks and turn verified solutions into scalable supervision for robot policies.
Original Article
View Cached Full Text

Cached at: 09/24/26, 07:40 AM

Paper page - EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics

Source: https://huggingface.co/papers/2609.27308 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Westudycodingagentsforlong-horizon,dexterousroboticsandaskwhethertheirsolutionscanprovidescalablesupervisionforlearninggeneralrobotpolicies.Totestthis,wedevelopEMBODIEDSWE-BENCH,asimulationbenchmarkforcodingagentsspanningcontact-richmanipulation,deformableobjects,andlong-horizontasksrequiringuptohalfanhourofcontinuousinteraction.Wefindthatfrontiercodingagentscansolvecomplexlong-horizontasksandtransferpriorsolutionsacrossbothtasksandembodiments.Wealsodesignsupportingtoolsthathelpagentsmoreeffectivelysolvethesetasks.However,theresultingsolutionsrequiresubstantialiterativeinteractionandaretypicallyspecializedtoindividualtaskinstances.WethereforeintroduceEMBODIEDSWE-GEN,whichexpandsasinglesolutionfromcodingagentintolargediversetrajectoriesfortrainingaVLA.VLAperformanceimproveswithmoregenerateddemonstrations,andagent-aideddiversificationimprovesgeneralizationtoheld-outtaskvariations.WealsoshowthataVLAfinetunedsolelyoncoding-agent-generatedsimulationdemonstrationscompletesalong-horizontaskonrealrobot.Together,ourframeworkusescodingagentstosolvecomplexroboticstasksandturnverifiedsolutionsintoscalablesupervisionforrobotpolicies.

View arXiv pageView PDFProject pageGitHub16Add to collection

Get this paper in your agent:

hf papers read 2609\.27308

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/2609.27308 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.27308 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.27308 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

Self-Evolving Embodied Agents via Skill-Harness Evolution

arXiv cs.CL

This paper introduces SHAPER, a self-evolving framework for embodied agents that keeps model parameters frozen and improves performance by evolving reusable skills and context-code harnesses through target-environment rollouts. Evaluated on VLABench and ESI-Bench, it proposes a practical alternative to fine-tuning when training is expensive or unavailable.

SWE-Touch: Benchmarking Coding Agents When Users Touch the Code

Hugging Face Daily Papers

Introduces SWE-Touch, a benchmark framework that injects conflicting user edits during agent coding trajectories, showing that current coding agents significantly degrade in collaborative settings despite strong standalone benchmark performance.