Pengju An
Papers
5
Total Citations
37
H-Index
3
About
Pengju An is at the forefront of robotic manipulation, pioneering the integration of multimodal large language models (MLLMs) with physical action. His research centers on creating unified, efficient brain models that bridge the gap between abstract reasoning and concrete robotic control. An’s major contributions include the development of **RoboMamba**, a Vision-Language-Action (VLA) model that enhances reasoning and manipulation efficiency, and **RoboBrain**, a unified architecture tackling long-horizon tasks where traditional MLLMs fall short. He also introduced **RoboMIND**, a comprehensive benchmark and dataset of 107k demonstration trajectories across 479 tasks and 96 object classes, collected via human teleoperation to standardize multi-embodiment robot learning. These works have rapidly accumulated citations (over 30 across recent papers), reflecting their immediate impact on the field. By addressing critical limitations in robotic reasoning and data standardization, An is shaping the next generation of intelligent, adaptable robots capable of executing complex, real-world manipulation tasks.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5