TS-Mask VLA: 2D Temporal-Spatial Masking Enables Efficient Vision-Language-Action Models for Robotic Manipulation

Jul 14, 2026

A new framework, TS-Mask VLA, introduces a 2D temporal-spatial masking strategy and a Discrete Diffusion Action Expert with Bridge Attention for vision-language-action (VLA) models in robotic manipulation. The model achieves a 95.7% average success rate on the LIBERO benchmark with only 0.5B parameters, outperforming much larger models, and sets a new best average sequence length of 4.19 on the CALVIN benchmark. Extensive experiments and ablation studies support the effectiveness of the proposed approach.

Why it matters: This work shows that efficient spatiotemporal modeling can allow smaller VLA models to surpass larger ones in complex robotic manipulation tasks, potentially reducing computational requirements for real-world deployment.

Full story at: arXiv Robotics ↗