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