Abstract
The usual power-law description of corrections to dynamic scaling is shown to be inapplicable to the superfluid transition in He4. A nonlinear dynamic renormalization-group analysis is presented which treats both large departures from scaling and the possible breakdown of scaling. Recent experimental results on second-sound damping are explained. The temperature dependence for (TλT)Tλ<104 is predicted to deviate only weakly from the dynamic scaling behavior.