Maximum Manifold Capacity Representations (MMCR) objective

methods / Causal Validation / Maximum Manifold Capacity Representations (MMCR) objective

Maximum Manifold Capacity Representations (MMCR) objective

Techniqueadvanced

A self-supervised training objective that directly maximizes the nuclear norm of the batch's augmentation-centroid matrix, which under an elliptical-manifold approximation is provably equivalent to maximizing manifold classification capacity (jointly minimizing manifold radius and dimensionality while decorrelating class centroids) - turning a passive geometric measurement (MFTMA) into an active training-time causal manipulation of representation geometry.

Used in (1 observation)

MMCR trains competitive SSL via a distinct low-dim, large-radius geometrymeasured

structure: Linear Separability · models: ResNet-50 (Maximum Manifold Capacity Representations, self-supervised, ImageNet), ResNet-50 (MoCo v2, unsupervised contrastive pretraining, ImageNet), ResNet-50 (SimCLR contrastive pretraining, ImageNet) · paper: Learning Efficient Coding of Natural Images with Maximum Manifold Capacity Representations