Fetching the paper…

Perception Tokens Enhance Visual Reasoning in Multimodal Language Models · Around