AI RESEARCH

LoMo: Local Modality Substitution for Deeper Vision-Language Fusion

arXiv CS.CV

ArXi:2605.30265v1 Announce Type: new Vision-Language Models (VLMs) have achieved substantial progress across a wide range of understanding and reasoning tasks, driven by large-scale image-text