/-
Copyright 2026 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.
-/modulepublicimportFormalConjecturesUtil
Bugeaud Collection of Conjectures and Open Questions: Automatic Expansions in Two Bases
Problem 10.32 asks whether some irrational real number has automatic expansions in two
multiplicatively independent bases. The answer is no. Adamczewski and Faverjon [AF20] proved
it with the multivariate Mahler method, in a form that is stronger than the question asked:
the base-$b$ digit sequence may be $k$-automatic for an arbitrary $k \ge 2$, not only for
$k = b$.
The definitions used here are in FormalConjecturesForMathlib:
Nat.MultiplicativelyIndependent and Nat.MultiplicativelyIndependentFamily,
AutomaticSequence.IsAutomatic, AutomaticSequence.IsAutomaticInBase and
AutomaticSequence.IsBroadlyAutomaticInBase.
References:
AF20 Adamczewski, Boris, and Colin Faverjon.
"Mahler's method in several variables and finite automata." arXiv:2012.08283 (2020).
Appendix A, Conjectures A.2 and A.3, both proved there from Part (i) of Theorem 1.1.
[AS03] Allouche, Jean-Paul, and Jeffrey Shallit. "Automatic Sequences: Theory,
Applications, Generalizations." Cambridge University Press, 2003. Chapter 13,
Open Problems 7, p. 403, is the problem that Problem 10.32 modifies.
[Bug12] Bugeaud, Yann. "Distribution modulo one and Diophantine approximation."
Vol. 193. Cambridge University Press, 2012. Chapter 10.
[Cob69] Cobham, Alan. "On the base-dependence of sets of numbers recognizable by finite
automata." Mathematical Systems Theory 3 (1969): 186-192.
[LvdP82] Loxton, John H., and Alfred J. van der Poorten. "Arithmetic properties of the
solutions of a class of functional equations." Journal für die reine und angewandte
Mathematik 330 (1982): 159-172. Its transcendence method was the first attack on the
problem; Nishioka later reported a gap in its main proof.
Adamczewski–Faverjon [AF20, Conjecture A.3], proved there from Part (i) of Theorem 1.1.
Let $b_1, \dots, b_r$ be multiplicatively independent positive integers, and for every $i$ let
$\xi_i$ be a real number that is automatic in base $b_i$. Then $\xi_1, \dots, \xi_r$ are
algebraically independent over $\mathbb{Q}$, unless one of them is rational.
Two points of fidelity. The hypothesis is the joint multiplicative independence of
$b_1, \dots, b_r$ that [AF20] defines on p. 3 (no nonzero integer tuple $(n_i)$ has
$\prod_i b_i^{n_i} = 1$), not the weaker pairwise condition; it is what the proof feeds to
Part (i) of Theorem 1.1, via the multiplicative independence of $1/b_1, \dots, 1/b_r$.
Automaticity is the broad notion that the proof of Conjecture A.3 uses: the base-$b_i$ digits
of $\xi_i$ are $k$-automatic for some $k \ge 2$, which need not be $b_i$.
Problem 10.32, strong form. No irrational real number is automatic in two multiplicatively
independent bases, in the broad sense that its digit sequence in each of the two bases is
k-automatic for some k. This is [AF20, Conjecture A.2], with the broad notion of
automaticity that its proof provides.
Problem 10.32. Is there an irrational real number whose expansions in two multiplicatively
independent bases can both be generated by a finite automaton? The answer is no, by [AF20].