IMPACT-VLA: Interaction-aware Multimodal Propagation Attribution via Counterfactual Trajectories for Vision-Language-Action Policies

תקציר מקורי באנגליתarXiv:2609.15005v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) policies perform robot manipulation tasks using multimodal inputs such as visual observations, proprioceptive states, and language instructions. However, it remains unclear at which execution stages each modality contributes to final task success and how input interventions propagate through subsequent states, observations, and actions. Existing attribution approaches primarily measure local sensitivity or temporally aggregated importance, limiting their ability to capture phase-dependent contributions and cross-phase dependencies. We propose Interaction-aware Multimodal Propagation Attribution via Counterfactual Trajectories for Vision-Language-Action Policies (IMPACT-VLA). IMPACT-VLA constructs behavioral phas