LATIDIA · Robótica

Enseñanza de un árbol para recuperar: Destilación en línea de un supervisor de LLM con repetición contrafactual

arXiv:2610.39775v1 Tipo de anuncio: nuevo Resumen: Los robots móviles autónomos suelen emparejar un planificador global con un controlador predictivo de modelo local (MPC), pero la evitación de obstáculos no convexos y un horizonte limitado pueden hacer que el

La noticia

arXiv:2610.39775v1 Announce Type: new Abstract: Autonomous mobile robots commonly pair a global planner with a local model predictive controller (MPC), but nonconvex obstacle avoidance and a limited horizon can make the robot freeze in complex layouts. Deciding whether to replan, wait, or call for help is a supervisory decision. Recent work queries a large language model (LLM) supervisor on every anomaly, which is slow and costly. We instead distill the LLM online into a decision tree. After each training episode, counterfactual replay labels recoveries as needed, unnecessary, or harmful. These labels retire unhelpful rules. In simulation, on 20 unseen layouts where the MPC alone freezes, the tree handles at least