τ₀-VLA: a Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation

Shanghai Innovation Institute

исследования офиц. + СМИ 3 ист. ~1 мин

Hierarchical robot foundation model that treats high-level subtask generation as a compute-scalable inference problem solved via world-model-guided beam search with execution memory; trained on 40,115 hours of real-world multimodal data and shows large closed-loop gains on long-horizon manipulation under distribution shift.

Почему это важно

HF Daily Paper with 519 upvotes (highest of the day); large real-data hierarchical VLA with test-time compute scaling

Важность: 3/5

HF Daily Paper with 519 upvotes

robotics world-models hierarchical test-time-compute vision-language-action vla

Источники

официальный τ₀-VLA: a Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation

официальный HF Daily Paper page for τ₀-VLA

СМИ Model card: sii-research/tau-0-vla