Some Tribonacci Conjectures

Jeffrey Shallit

DOI: https://doi.org/10.48550/arXiv.2210.03996

2022-10-08

Abstract:In a recent talk of Robbert Fokkink, some conjectures related to the infinite Tribonacci word were stated by the speaker and the audience. In this note we show how to prove (or disprove) the claims easily in a "purely mechanical" fashion, using the Walnut theorem-prover.

Combinatorics,Discrete Mathematics,Formal Languages and Automata Theory

What problem does this paper attempt to address?

This paper attempts to solve several conjectures related to the infinite Tribonacci word. Specifically, the author Jeffrey Shallit shows how to use the Walnut theorem prover to prove or disprove these conjectures in a "purely mechanical" way. These conjectures mainly involve the properties of the Tribonacci sequence and its related sequences, especially the properties of two sequences A140100 and A140101 related to combinatorial games. The paper verifies whether these sequences are Tribonacci - synchronous by constructing an automaton method and uses the Walnut tool for strict mathematical proof.

The main problems mentioned in the paper include:
1. Verify whether the sequences \(X=(X(k))_{k\geq0}\) and \(Y =(Y(k))_{k\geq0}\) are Tribonacci - synchronous.
2. Verify the conjecture proposed by Robbert Fokkink in his speech, that is, for all \(n\geq1\), either \(X(Y(n))=X(n)+Y(n)\) or \(X(Y(n))=X(n)+Y(n)- 1\).
3. Verify two conjectures proposed by Julien Cassaigne in the same speech:
- For all \(n\geq1\), either \(X(Y(n))=X(n)+Y(n)\) or \(Y(X(n))=X(n)+Y(n)\).
- For all \(n\geq1\), we have \(X(n)+Y(n)-Y(X(n))\in\{0,1,2\}\).
4. Disprove the conjecture proposed by Dan Rust in the same speech, that is, for all \(n\geq1\), we have \(X(Y(n))+Y(X(n))=2X(n)+2Y(n)-1\).

By using the Walnut theorem prover, the author successfully verifies the correctness or incorrectness of the above conjectures and shows the effectiveness of this method in dealing with such problems.