Peer-reviewed and preprint work on robotic manipulation robustness, vision-language-action models, and 6D pose estimation.
-
How VLAs (Really) Work In Open-World Environments
Paper • 2604.21192 • Published • 1 -
Do World Action Models Generalize Better than VLAs? A Robustness Study
Paper • 2603.22078 • Published • 9 -
Distracted Robot: How Visual Clutter Undermine Robotic Manipulation
Paper • 2511.22780 • Published -
Improving Robotic Manipulation Robustness via NICE Scene Surgery
Paper • 2511.22777 • Published