LATIDIA · Robótica

SPARK: Post-entrenamiento simple para adaptar el conocimiento pRetrained al control de robots

arXiv: 2611.12102v1Announce Type: new Abstract: El aprendizaje de la representación a gran escala ha producido modelos de cimientos públicos que proporcionan fuertes características visuales de uso general. Sin embargo, sus objetivos previos a la capacitación no son

La noticia

arXiv:2611.12102v1 Announce Type: new Abstract: Large-scale representation learning has produced public foundation models that provide strong general-purpose visual features. However, their pretraining objectives are not designed for robot representation learning, so the resulting embeddings tend to emphasize global semantics rather than the temporally sensitive representations required for robot control. Training robot models from scratch is also undesirable due to the limited scale and high collection cost of robotics data, as well as the substantial computational burden of pre-training. In this paper, we propose SPARK (Simple Post-training for Adapting pRetrained Knowledge), a post-training method that adapts foundation models to robot control. SPARK is built on two principles: temporal abstraction, which encourages