LATIDIA · Robótica

¿Saben los VLA que fracasarán? Patrones interpretables de éxito y fracaso en sistemas físicos de IA

arXiv: 2610.32208v1Tipo de anuncio: nuevo Resumen: Los modelos de Visión-Lenguaje-Acción (VLA) han surgido como poderosas políticas de robots de propósito general, pero sus éxitos y fracasos siguen siendo difíciles de diagnosticar y comprender. I

La noticia

arXiv:2610.32208v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have emerged as powerful general-purpose robot policies, but their successes and failures remain difficult to diagnose and understand. In this work, we show that sparse autoencoder (SAE) representations of OpenVLA activations contain outcome-predictive information and can be used to predict task failure before the robot reaches an unrecoverable state. Moreover, to provide insights into what common behaviors may contribute to these outcomes, we identify individual SAE features whose strong activations are associated with specific stages of success or failure, and uncover recurring behavioral patterns by localizing temporal intervals where these features exhibit significant activation changes. We further develop a VLM-based interpretation pipeline