LATIDIA · Robótica

Del razonamiento a la acción: evaluación comparativa de modelos de IA de frontera en tareas de manipulación de robots LIBERO-10

arXiv: 2610.30051v1Tipo de anuncio: nuevo Resumen: Los recientes avances en los modelos de IA de frontera han demostrado capacidades notables en la comprensión visual, el razonamiento, la planificación y el uso de herramientas, lo que aumenta la posibilidad de usar

La noticia

arXiv:2610.30051v1 Announce Type: new Abstract: Recent advances in frontier AI models have demonstrated remarkable capabilities in visual understanding, reasoning, planning, and tool use, raising the possibility of using general-purpose AI models for robot manipulation without task-specific training. However, their manipulation capabilities have not been systematically evaluated on standardized robotics benchmarks. In this report, we present a unified closed-loop manipulation framework for evaluating frontier AI models on the LIBERO-10 benchmark. Our framework enables models to control a robot directly through visual observations, robot proprioception, and structured tool calls, primarily by specifying target end-effector poses and gripper commands, without task-specific fine-tuning or pretrained manipulation skills. We evaluate six frontier AI models, including