Code + derived-data companion for three RMT preprints: effective-rank collapse (Firma B note), a random-matrix account of structure formation in neural networks (main paper), and the plate-network unification note.

A quantitative finance engine that uses Random Matrix Theory (Marchenko-Pastur) to denoise correlation matrices and Hierarchical Risk Parity (HRP) for robust portfolio allocation.

PCA on financial returns with Marchenko-Pastur RMT signal/noise separation