How far a shared linear map suffices to predict feature changes from image edits: probe comparisons across ConvNeXt, SwinV2, and DINOv3
The work trains probes of increasing capacity, from a spatially shared linear map to nonlinear per-vector, receptive-field, and global transformer models, to predict feature-space changes induced by geometric transforms, photometric edits, occlusions, and diffusion-generated semantic edits, finding that across ConvNeXt, SwinV2, and DINOv3 a single shared linear map often predicts held-out manipulation outcomes nearly as well as substantially more expressive probes for the supervised backbones, with sufficiency generally increasing with depth, while this pattern is less consistent for DINOv3.