Fetching the paper…

Text as Images: Can Multimodal Large Language Models Follow Printed Instructions in Pixels? · Around