AI RESEARCH
Downscaling Intelligence: Exploring Perception and Reasoning Bottlenecks in Small Multimodal Models
arXiv CS.CV
•
ArXi:2511.17487v2 Announce Type: replace Scaling up multimodal models has enabled remarkable advances in visual understanding and reasoning, but practical demands call for smaller, efficient systems. In this work, we conduct a principled analysis of downscaling intelligence in multimodal models, examining how reduced large language model (LLM) capacity affects multimodal capabilities. Our initial findings reveal an interesting trend: LLM downscaling disproportionately affects visual capabilities, rather than abilities inherited from the.