RITUAL: Random Image Transformations as a Universal Anti-hallucination Lever in Large Vision Language Models | Read Paper on Bytez