Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models 事件

Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models arXiv:2605.12374v4 Announce Type: replace Abstract: Visual latent reasoning lets a multimodal large language model (MLLM) create intermediate visual evidence as continuous tokens, avoiding external tools or image generators. However, existing methods usually follow an output-as-input latent paradigm and yield unstable gains. We identify evidence for a feature-space mismatch that can contribute t