Fetching the paper…

Accelerating Ill-Conditioned Low-Rank Matrix Estimation via Scaled Gradient Descent · Around