Ji-Ha's Blog

Retention and radius parametrization for Lion-K with Corrected Cautious Weight Decay.

Autoregression and diffusion look like opposites, but both are solving the same transport problem - how to turn simple noise into structured data.

Why coordinate-level optimizers can behave differently on weights that represent the same model, and how quotient-aware updates remove the hidden gauge.

A GPU-friendly algorithm for tight certified singular-value endpoint bounds using scaled Gram matrices, SYRKs, Frobenius reductions, and a small scalar moment problem.

A hardware-aware hybrid polar decomposition for ML: one Dynamic Weighted Halley (rational) step to handle the hard early regime, then two Polar Express (polynomial) cleanup steps once the spectrum is easy. The result is exactly two rectangular GEMMs, no eigendecomposition or power iteration, and robust convergence from condition numbers up to 1000.