SOURCE-LINKED INTELLIGENCE

Learning Where to Focus: Self-Supervised Multi-Scale ViTs for Histopathology

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

Pathologists diagnose diseases by first locating suspicious tissue and then examining it at higher magnification, whereas self-supervised vision transformers (ViTs) allocate the same spatial resolution to every image region despite diagnostic evidence being sparse and spanning multiple biological scales. Recent pathology foundation models have substantially improved representation quality by scaling training data and model capacity, but largely retain uniform tokenization. We instead investigate whether pathology representations can be improved by learning where to allocate spatial resolution

Read original source ↗ Open in workspace

recordType

paper

region

Global