because a lot of it is meant to correspond to math equations so variables names like w, u, v , b ,g match the equations in the papers ? I actually think it's pretty readable, as long as you know what it i supposed to implement (i don't; but i imagine they are implementing a complex graph), and short names help figure out where things go in and out in one screenfull.

Complex graphs are literally a spaggeti of arrows, and this format actually is pretty readable (even though in pytorch it would be more readable). I guess they leave comments out because it's not really possible to understand each line on its own (unless it's an implementation detail); you have to read the paper to know what s going on