With 42cb7e3, the NN function (especially for $F_2$) saturates in the small $x$ extrapolation region.
There are several possibilities to address this but the easiest check would be to use a linear activation throughout the network and use $\ln (x)$ in the input instead of $x$.
With 42cb7e3, the NN function (especially for $F_2$) saturates in the small $x$ extrapolation region.
There are several possibilities to address this but the easiest check would be to use a linear activation throughout the network and use $\ln (x)$ in the input instead of $x$.