τ_0-VLA: a Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation

Hugging Face Daily Papers Papers

Summary

A hierarchical vision-language-action model improves long-horizon robot manipulation by using world-model-guided test-time search to scale computation for high-level subtask decisions.

Long-horizon robot manipulation requires a robot to both execute individual skills reliably and sequence them coherently over extended tasks. Most hierarchical vision-language-action (VLA) models make each such decision with a single forward pass, leaving no mechanism to allocate additional computation to difficult or consequential choices. We introduce τ_0-VLA, a hierarchical robot foundation model that formulates high-level subtask generation as a compute-scalable inference problem through world-model-guided test-time computation. At each inference step, the high-level policy uses execution memory to generate a subtask and, when needed, searches over alternatives before committing to its output. A low-level policy then executes the generated subtask across multiple robot embodiments. The policy is trained on 40,115 hours of heterogeneous real-world data with multimodal co-training. Across in-domain and distribution-shifted settings, allocating additional test-time computation substantially improves next-subtask prediction accuracy, and these gains translate into higher closed-loop success on long-horizon robot manipulation tasks.
Original Article
View Cached Full Text

Cached at: 08/21/26, 04:10 PM

Paper page - τ_0-VLA: a Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation

Source: https://huggingface.co/papers/2608.16885 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

A hierarchical vision-language-action model improves long-horizon robot manipulation by using world-model-guided test-time search to scale computation for high-level subtask decisions.

Long-horizon robot manipulation requires a robot to both execute individual skills reliably and sequence them coherently over extended tasks. Most hierarchicalvision-language-action(VLA) models make each such decision with a single forward pass, leaving no mechanism to allocate additional computation to difficult or consequential choices. We introduce τ_0-VLA, a hierarchical robot foundation model that formulates high-levelsubtask generationas a compute-scalable inference problem throughworld-model-guided test-time computation. At each inference step, the high-level policy usesexecution memoryto generate a subtask and, when needed, searches over alternatives before committing to its output. Alow-level policythen executes the generated subtask across multiple robot embodiments. The policy is trained on 40,115 hours of heterogeneous real-world data withmultimodal co-training. Across in-domain and distribution-shifted settings, allocating additional test-time computation substantially improves next-subtask prediction accuracy, and these gains translate into higher closed-loop success on long-horizon robot manipulation tasks.

View arXiv pageView PDFProject pageGitHubAdd to collection

Get this paper in your agent:

hf papers read 2608\.16885

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

Models citing this paper1

#### sii-research/tau-0-vla Robotics• 3B• Updated3 days ago • 196 • 3

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2608.16885 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2608.16885 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

HiVLA: A Visual-Grounded-Centric Hierarchical Embodied Manipulation System

Hugging Face Daily Papers

HiVLA introduces a hierarchical vision-language-action framework that decouples semantic planning from motor control using a diffusion transformer action expert for improved robotic manipulation. The system combines a VLM planner for task decomposition and visual grounding with a specialized DiT action expert using cascaded cross-attention, outperforming end-to-end baselines particularly in long-horizon tasks and fine-grained manipulation.