For
h(d)=∏p∣d1−g(p)g(p),G(D,z)=∑ℓ<Dℓ∣P(z)h(ℓ),
defineGd(y,z)=∑ℓ<yℓ∣P(z)(ℓ,d)=1h(ℓ).
The optimizing Selberg
weights are
λd=μ(d)∏p∣d(1−g(p))−1G(D,z)Gd(D/d,z)
for
d<D dividing
P(z), and zero otherwise. They satisfy
λ1=1 and diagonalize the main
quadratic form to
1/G(D,z).
New to topics? Read the docs here!