We are thrilled to announce that X-VLA has been accepted to ICLR 2026. Successful generalist Vision–Language–Action (VLA) models depend on scalable, cross-platform training across diverse robotic ...