MST intrinsic dimension of frozen SSL features predicts probe accuracy

Mordacq et al. estimate intrinsic dimension of frozen penultimate-layer features from 33 pretrained self-supervised vision checkpoints across 14 methods (VICReg, DINO, BarlowTwins, DINOv2, DINOv3, I-JEPA, iBOT, CLIP, EVA-CLIP, SigLIP, PE-Core, Franca) using a minimum-spanning-tree estimator (IdEst) [mordacq-etal-2026-idest-assessing-ssl-representations-via-intrinsic-dimension] MST intrinsic dimension is strongly negatively correlated with downstream linear-probe accuracy across four datasets (Spearman rho approximately -0.8, e.g. ImageNet -0.74, iNat-18 -0.84; Kendall tau approximately -0.6) [mordacq-etal-2026-idest-assessing-ssl-representations-via-intrinsic-dimension] Lower-dimensional frozen representations transfer better, making intrinsic dimension a label-free proxy for representation quality without a linear-probe run [mordacq-etal-2026-idest-assessing-ssl-representations-via-intrinsic-dimension] The relationship is correlational, over frozen features and linear-probe evaluation only [mordacq-etal-2026-idest-assessing-ssl-representations-via-intrinsic-dimension]