Is VLA Reasoning Faithful? Probing Safety of Chain-of-Causation in Autonomous Driving Models 文章

ArXiv CS.CV2026-05-26NEWSen作者: Nicanor Mayumu, Xiaoheng Deng, Patrick Mukala
人工智能
查看原文 →
关系图谱

Is VLA Reasoning Faithful? Probing Safety of Chain-of-Causation in Autonomous Driving Models · 相关技术

相关技术

ORMODEPLAsafetydivide-and-conquer partitioningVerseVLAStraight-Through EstimatorSULSSESAFENATMaliForEffort Metric AttentionENAEFIANNAGIiOSTPU