Optimization
Stochastic and convex optimization, training dynamics, sampling algorithms.
2 items · All topics
-
Diffusion Removes Langevin's Conditioning Dependence: A Sharp Gaussian Analysis
For Gaussian targets, tuned diffusion sampling avoids the condition-number penalty that Langevin methods pay; both rates are sharp.
-
A stochastic subgradient method with optimal failure exponent
Averaged subgradient descent attains the sharp large-deviation exponent under sub-Gaussian noise, and a matching lower bound shows the constant is optimal.