Let F be a totally real number field F of degree n>1. Let v1,…,vn:F↪R be the real embeddings of F, and for a∈F abbreviate ai:=vi(a) and extend this elementwise to matrices. Write α∈F>0× if α is totally positive and ZF,>0×:=ZF×∩F>0×.
Let k1,…,kn≥2 be positive integers of the same parity, and let N be a nonzero ideal of the ring of integers ZF of F.
For an ideal b⊆ZF, let
γ∈Γ0(N)b:={γ=(acbd)∈GL2(F):c∈Nb, b∈b−1, and det(γ)∈ZF,>0×}.
A Hilbert modular form of weight (k1…,kn) and level N is a tuple (fb)b of holomorphic functions f:Hn→C, indexed by ideals b representing the narrow class group of ZF, such that for all z=(z1…,zn)∈Hn and all γ=(acbd)∈Γ0(N)b we have
fb(γz)=fb(c1z1+d1a1z1+b1,…,cnzn+dnanzn+bn)=i=1∏n((aidi−bici)ki/2(cizi+di)ki)fb(z).
A Hilbert cusp form is a Hilbert modular form that vanishes at the cusps.