String theory combined with inflationary cosmology has led to the picture of inflationary multiverse, populated by a multitude of vacua with diverse properties. Our Universe begins with a quantum tunneling from an eternally-inflating parent vacuum, followed by a period of slow-roll inflation. The details of the high-energy vacuum landscape are not well understood, and its effective field theory is often modeled as random Gaussian fields. In this talk, I will focus on small-field inflation in the landscape of the random Gaussian fields and explain that inflation is not multi-field, the probability for future detection of spatial curvature is rather low (P ~ 0.001), and the distribution for the amplitude of density fluctuations has a unique shape and a preferred domain.