The Epstein zeta function for a matrix
of a positive definite real quadratic
form and
a complex variable with
(where
denotes the real part) is
defined by
|
(1)
|
where the sum is over all column vectors with integer coordinates and the prime means the summation excludes the origin (Terras 1973). Epstein (1903) derived the analytic continuation, functional equation, and so-called Kronecker limit formula for this function.
Epstein (1903) defined this function in the course of an effort to find the most general possible function satisfying a functional
equation similar to that satisfied by the Riemann
zeta function
(Glasser and Zucker 1980, p. 68).
A slightly different notation is used in theoretical chemistry, where the Epstein zeta function arises in connection with lattice sums.
Let
be a positive definite quadratic form
|
(2)
|
where
with
,
...
is a symmetric matrix. Then the Epstein zeta
function can be defined as
|
(3)
|
where
and
are arbitrary vectors, the sum
runs over a
-dimensional
lattice, and
is omitted if
is a lattice vector (Glasser and
Zucker 1980, p. 69).
For a point lattice with covolume 1, Luo
and Wei (2026) proved that the Epstein zeta function
is minimized for every
uniquely, up to an orthogonal
transformation, by the rescaled D4
lattice
.
Their computer-assisted proof uses certified interval bounds. Luo and Wei (2026) state that ChatGPT assisted with
review of proof arguments and citations and with preparation
of the computational verification materials.