“…In fact, in [1,2,3] non-steep (valencelike) initial distributions xf (x, µ 2 ) at some low scale µ ≈ 0.6 GeV were suggested in order to predict, purely dynamically, the rise of xq and xg in the above range of x and Q 2 . This prediction was subsequently confirmed by the measurements of F p 2 (x, Q 2 ) and xg(x, Q 2 ) at HERA [4,5]. As was stated in [2,3], the available pre-HERA data at x > 10 −2 utilized to fix the valence-like input distributions still allowed for a slight, typically O(10%), variation (increase) of µ which did not affect F 2 (x > 10 −2 , Q 2 ) but resulted in an about 10% (20%) uncertainty of the radiative predictions at x = 10 −3 (10 −4 ).…”