I worked on this model at Liquid and will be checking in to answer any questions/discussions, both about this vision model and LiquidAI more broadly.