DexJoCo: A Benchmark and Toolkit for Task-Oriented Dexterous Manipulation on MuJoCo
Summary
DexJoCo introduces a benchmark and toolkit for task-oriented dexterous manipulation in MuJoCo, featuring 11 functional tasks, a low-cost data collection system, and comprehensive evaluations that highlight limitations in current dexterous manipulation policies.
View Cached Full Text
Cached at: 05/18/26, 06:24 AM
Paper page - DexJoCo: A Benchmark and Toolkit for Task-Oriented Dexterous Manipulation on MuJoCo
Source: https://huggingface.co/papers/2605.16257 Published on May 15
#2 Paper of the day Authors:
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
DexJoCo presents a benchmark and toolkit for dexterous manipulation with 11 functional tasks evaluating tool-use, bimanual coordination, and long-horizon execution, along with a low-cost data collection system and comprehensive model evaluation.
Achieving human-level manipulation requires dexterous robotic hands capable of complex object interactions. Advancing such capabilities further demands standardizedbenchmarks for systematic evaluation. However, existing dexterousbenchmarks lack tasks that reflect the unique manipulation capabilities of dexterous hands over parallel grippers, as well as comprehensive evaluation pipelines. In this paper, we present DexJoCo, abenchmarkandtoolkitfor task-orienteddexterous manipulation, comprising 11functionally grounded tasksthat evaluatetool-use,bimanual coordination,long-horizon execution, and reasoning. We develop a low-cost data collection system and collect 1.1K trajectories across these tasks, with support fordomain randomizationto assess robustness. Webenchmarkmodern models under diverse settings, including visual anddynamics randomization,multi-task training, andaction-head adaptation. Through extensive empirical analysis, we identify several important insights and common limitations of current policies indexterous manipulation, highlighting key challenges for future research in dexterous hand robot learning. Project page available at: https://dexjoco.github.io
View arXiv pageView PDFProject pageGitHub7Add to collection
Get this paper in your agent:
hf papers read 2605\.16257
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/2605.16257 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2605.16257 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2605.16257 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
MuJoCo – Advanced Physics Simulation
Google DeepMind maintains MuJoCo, a high-performance open-source physics engine with C/Python APIs and Unity plugin for robotics and ML research.
DeVI: Physics-based Dexterous Human-Object Interaction via Synthetic Video Imitation
DeVI introduces a framework that turns text-conditioned synthetic videos into physically plausible dexterous robot control via a hybrid 3D-2D tracking reward, enabling zero-shot generalization to unseen objects.
TOBench: A Task-Oriented Omni-Modal Benchmark for Real-World Tool-Using Agents
TOBench is a new benchmark for evaluating AI agents on real-world, task-oriented tool use with multimodal inputs and closed-loop verification. Experiments show top models like Qwen 3.5 Plus achieve only 41% success, far below the 94% human benchmark, highlighting a significant gap.
@Modular: Mojo has minimal boilerplate, a strict type system, and compile-time validation of code, all things that make it well-s…
Modular releases open-source Mojo agent skills to help AI coding agents produce correct, idiomatic Mojo code, including a demo translating CUDA kernel code to Mojo. Mojo's minimal boilerplate, strict type system, and compile-time validation make it suitable for agentic workflows.
CoCoDA: Co-evolving Compositional DAG for Tool-Augmented Agents
This paper introduces CoCoDA, a framework that uses a co-evolving compositional Directed Acyclic Graph (DAG) to manage tool libraries for augmented agents. It enables small language models to efficiently retrieve and compose tools, allowing an 8B model to match or exceed the performance of a 32B model on reasoning benchmarks.