/-
Copyright 2025 The Formal Conjectures Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-/modulepublicimportFormalConjecturesUtilpublicimportFormalConjectures.Millennium.RiemannHypothesis
Artin's conjecture on primitive roots
Artin's conjecture predicts, given an integer $a$, densities of primes $p$ for which
$a$ is a primitive root modulo $p$. Under certain conditions (when $a$ is not a
power and its squarefree part is not $1\pmod{4}$) the density is given by Artin's constant
$$\prod_{p\ \text{prime}} \left(1 - \frac{1}{p(p - 1)}\right).$$
For more general values of $a$, this constant must be corrected by certain factors.
When $a = b^m$, $m$ is a maximal odd power, the squarefree part of $b$ satisfies
$b_0 \not\equiv 1\pmod{4}$. Then Artin's constant should be multiplied by
$$\prod_{p \mid m} \frac{p(p - 2)}{p^2 - p - 1}.$$
When $a = b^m$, $m$ is a maximal power, the squarefree part of $b$ satisfies
$b_0\equiv 1\pmod{4}$. Then Artin's constant should be multiplied by the factor in
the above bullet, as well as an additional entanglement factor from the primes dividing
$\gcd(b_0, m)$ and primes dividing $b_0$:
$$1 - \prod_{p \mid \gcd(b_0, m)} \frac{1}{2 - p}
\prod_{p \mid b_0, p\nmid m} \frac{1}{1 + p - p^2}.$$
When $a = -1$ or $a$ is a square, then the density is $0$.
Note that Artin's conjecture has been proved by Hooley [Ho67] subject to the Riemann hypothesis
for the Dedekind zeta functions of the Kummer fields $\mathbb{Q}(\zeta_k, a^{1/k})$, $k$
squarefree. These fields are in general non-abelian over $\mathbb{Q}$, so this hypothesis is not
covered by the Generalized Riemann Hypothesis for Dirichlet $L$-functions.
Z is a Dedekind zeta function of the number field K: it is holomorphic away from $s = 1$
and agrees with the Dedekind zeta series of K on $\operatorname{Re} s > 1$, where that series
converges. Such a Z is uniquely determined on $\mathbb{C} \setminus {1}$.
The Riemann hypothesis for the Dedekind zeta function $\zeta_K$ of a number field $K$:
every zero of (the analytic continuation of) $\zeta_K$ in the critical strip
$0 < \operatorname{Re} s < 1$ has real part $\frac{1}{2}$.
Hooley's hypothesis for $a$: the Riemann hypothesis holds for the Dedekind zeta function
of the Kummer field $\mathbb{Q}(\zeta_k, a^{1/k})$, the splitting field of $X^k - a$ over
$\mathbb{Q}$, for every squarefree $k$. This is the hypothesis under which Hooley [Ho67] proved
Artin's conjecture.
Artin's conjecture on $S(a)$ when $a = b^m$ is a power, where $m$ is odd and maximal,
requires a correction factor to multiply ArtinConstant and is given by
$$\prod_{p \mid m} \frac{p(p - 2)}{p^2 - p - 1}.$$
-- Eq. (1.2) of https://arxiv.org/pdf/1112.4816noncomputabledefpowCorrectionFactor(m:ℕ):ℝ:=∏p∈m.primeFactors,p*(p-2:ℝ)/(p^2-p-1)
Artin's conjecture on $S(a)$ when $a = b^m$ is a power, and the squarefree part
of $b_0\equiv 1\pmod{4}$, requires a further correct factor to
ArtinConstant * powCorrectionFactor m, which modifies primes which divide
$\gcd(b_0, m)$ and primes which do not divide $m$ separately as
$$ 1 - \prod_{p \mid \gcd(b_0, m)} \frac{1}{2 - p}
\prod_{p \mid b_0, p\nmid m} \frac{1}{1 + p - p^2}.$$
-- Eq. (1.4) of https://arxiv.org/pdf/1112.4816noncomputabledefentanglementFactor(b:ℕ)(m:ℕ):ℝ:=(1-(∏p∈(b.squarefreePart.gcdm).primeFactors,1/(2-p:ℝ))*(∏p∈b.squarefreePart.primeFactors\m.primeFactors,1/(1+p-p^2:ℝ)))
Artin's Conjecture on Primitive Roots, first half.
Let $a$ be an integer that is not a square number and not $−1$. Then the set $S(a)$
of primes $p$ such that $a$ is a primitive root modulo $p$ has a positive asymptotic
density inside the set of primes. In particular, $S(a)$ is infinite.
Artin's Conjecture on Primitive Roots, first half, conditional on the Riemann hypothesis
for the Dedekind zeta functions of the fields $\mathbb{Q}(\zeta_k, a^{1/k})$, $k$ squarefree
[Ho67].
Artin's Conjecture on Primitive Roots, second half.
Write $a = a_0 b^2$ where $a_0$ is squarefree. Under the conditions that $a$ is not a perfect
power and $a_0\not\equiv 1\pmod{4}$ (sequence A85397 in the OEIS), the density of the set
$S(a)$ of primes $p$ such that $a$ is a primitive root modulo $p$ is independent of $a$ and
equals Artin's constant.
Artin's Conjecture on Primitive Roots, second half, conditional on the Riemann hypothesis
for the Dedekind zeta functions of the fields $\mathbb{Q}(\zeta_k, a^{1/k})$, $k$ squarefree
[Ho67].
Artin's Conjecture on Primitive Roots, second half, different residue version
If $a$ is a square number or $a = −1$, then the density of the set $S(a)$ of primes
$p$ such that $a$ is a primitive root modulo $p$ is $0$.
Artin's Conjecture on Primitive Roots, second half, power version
If $a = b^m$ is a perfect odd power of a number $b$ whose squarefree part
$b_0\not\equiv 1 \pmod{4}$, then the density of the set $S(a)$ of primes $p$ such that
$a$ is a primitive root modulo $p$ is given by
$$C\prod_{p \mid m} \frac{p(p - 2)}{p^2 - p - 1}$$,
where $C$ is Artin's constant.
Artin's Conjecture on Primitive Roots, second half, power version, conditional on the
Riemann hypothesis for the Dedekind zeta functions of the fields $\mathbb{Q}(\zeta_k, a^{1/k})$,
$k$ squarefree [Ho67].
Artin's Conjecture on Primitive Roots, second half, power version
If $a = b^m$ is a perfect power of a number $b$ whose squarefree part $b_0\equiv 1 \pmod{4}$,
then the density of the set $S(a)$ of primes $p$ such that $a$ is a primitive root modulo $p$
is given by
$$C \left(\prod_{p \mid m} \frac{p(p-2)}{(p ^ 2 - p - 1)}\right)
\left(1 - \prod_{p \mid \gcd(b_0, m)} \frac{1}{2 - p}
\prod_{p \mid b_0, p\nmid m} \frac{1}{(1 + p - p ^ 2)}\right),$$
where $C$ is Artin's constant.
Artin's Conjecture on Primitive Roots, second half, power version, conditional on the
Riemann hypothesis for the Dedekind zeta functions of the fields $\mathbb{Q}(\zeta_k, a^{1/k})$,
$k$ squarefree [Ho67].