AI RESEARCH

Learnability and Competition in High-Dimensional Multi-Component ICA

arXiv CS.LG

ArXi:2605.08552v1 Announce Type: cross Independent Component Analysis (ICA) is a foundational tool for unsupervised representation learning, yet its high-dimensional theory remains largely limited to single-component recovery. We develop an asymptotically exact mean-field theory for multi-component online ICA, capturing the coupling induced by simultaneous learning and orthogonalization.