Research questionHow can frozen vision foundation models localize industrial defects from only a few examples?Pretrained attention often prioritizes semantic relationships, which may obscure the localized texture and structural deviations that define industrial defects. With few examples, learning representations tailored to these deviations is also constrained.