τ₀-VLA Tackles Long-Horizon Robotic Tasks

The new embodied AI model τ₀-VLA leverages hierarchical planning and a world model to handle long-horizon robotic tasks. Trained on 40,000 hours of real-world data, it achieves up to 12 minutes of continuous autonomous operation.

2026-07-27 ~ 2026-07-28 · 2 related posts

• τ(0)-VLA uses 40,115 hours of robot data to tackle long-horizon tasks — 机器之心 · 2026-07-27
• τ₀-VLA shows 12-minute autonomous robot manipulation with hierarchical planning — chris_j_paxton · 2026-07-28