Jiafei Cao

DURA reveals that visually indistinguishable adversarial patches can exploit VLA models, posing a significant threat to their deployment in real-world robotics.

By preserving the semantics of pretrained models while achieving superior compositional generalization, InternVLA-A1.5 redefines how robots can learn and execute complex tasks.

Cortex outperforms traditional models by enabling zero-shot execution of complex long-horizon tasks, bridging the gap between high-level planning and low-level execution.

EventVLA's foresight-driven memory mechanism boosts long-horizon task success rates by 40% by dynamically capturing critical visual events before they vanish.