I think Flux might reverse the convolutional filter when applying it, tested you code but with out2 = Flux.sigmoid.(sum(reverse(w) .* inp) .+ b) which seemed to give the same output then.
I think Flux might reverse the convolutional filter when applying it, tested you code but with out2 = Flux.sigmoid.(sum(reverse(w) .* inp) .+ b) which seemed to give the same output then.