Weaving Context Across Images: Improving Vision-Language Models through Focus-Centric Visual Chains

Re-assign community