LATIDIA · Robótica

RoboICL: aprendizaje en contexto incorporado con GPT-6 Astra

arXiv:2609.34261v1 Anuncio Tipo: nuevo Resumen: Los modelos de lenguaje de visión de propósito general ofrecen una vía prometedora para el control de robots sin entrenamiento previo: \gptastra{} destaca en la manipulación abierta y condicionada por el lenguaje o la imagen.

La noticia

arXiv:2609.34261v1 Announce Type: new Abstract: General-purpose vision-language models offer a promising way to zero-shot robot control: \gptastra{} excels at open-ended and language- or image-conditioned manipulation but remains substantially weaker on high-precision and long-horizon tasks. We introduce \emph{RoboICL}, an in-context robot-control framework that narrows these gaps without robot-specific parameter updates or a learned VLA. RoboICL separates \emph{demonstration context}, which provides recorded examples when available, from \emph{interaction memory}, which accumulates the model's own actions and observed outcomes. Both use a shared observation--action--receipt--observation grammar. To preserve experience across task stages, RoboICL combines sampled demonstration blocks with bounded anchored memory. Fixed anchors keep earlier rollout interactions available for in-context learning, while the latest interaction