This tells you how many Surrogate Variables have been found in your data. SV1 will usually correspond to your batch effect. If you run SVA these SVs will all be removed from your data. This may be more cleaning than you want: it can remove real signal from your dataset. It is better to run SVA outside this program so you have more control over the effects.
This tab is empty until you run SVA using the button in the left panel. Afterwards, it refers you to the other tabs, where you will now be able to select post-SVA plots by selecting the radio button.