We demonstrate empirically on both toy and real-world examples the benefit of using the Power Descent and going beyond the Entropic Mirror Descent framework, which fails as the dimension grows.
DAUDEL, K., DOUC, R. et PORTIER, F. (2021). Infinite-dimensional gradient-based descent for alpha-divergence minimisation. Annals of Statistics, 49(4), pp. 2250 – 2270.