AI RESEARCH

Don't Blink: Evidence Collapse during Multimodal Reasoning

arXiv CS.AI

ArXi:2604.04207v1 Announce Type: new Reasoning VLMs can become accurate while progressively losing visual grounding as they think. This creates task-conditional danger zones where low-entropy predictions are confident but ungrounded, a failure mode text-only monitoring cannot detect. Evaluating three reasoning VLMs on MathVista, HallusionBench, and MMMU_Pro, we find a pervasive evidence-collapse phenomenon: attention to annotated evidence regions drops substantially, often losing over half of evidence mass, as reasoning unfolds.