Maximal determinant conjecture for 2-consecutive-ones matrices

At least 8 years old · documented by

Let A∈{0,1}n×nA\in\{0,1\}^{n\times n} be a matrix with the 2-consecutive ones property, meaning that the ones in each row occur in at most two blocks in some ordering of the columns. 2-consecutive-ones maximal determinant conjecture. Then

∣det⁡(A)∣≤4n/3.|\det(A)|\leq 4^{n/3}.

The paper gives constructions with determinant 4(n−3)/34^{(n-3)/3} for positive integers nn divisible by 33, and states that this construction is essentially optimal. The source provides no proof of the conjectured upper bound.

References

Primary source

Henning Bruhn and Dieter Rautenbach, “Maximal determinants of combinatorial matrices”, arXiv:1711.09935 (2017).

Progress summary

Refreshed
Claimed progress

An unverified submitted example would disprove the conjecture, but no independent source has checked it.

The conjecture says that every square zero-one matrix whose rows have at most two blocks of ones in a common column ordering has determinant at most 4n/34^{n/3}. The 2017 paper gives near-extremal constructions but no proof of the proposed bound.

Known results

  • For 3∣n3\mid n, constructions with determinant 4(n−3)/34^{(n-3)/3} were given in 2017.
  • The original paper proves the weaker general bound ∣det⁡(A)∣≤3.936n/2|\det(A)|\leq 3.936^{n/2}.

Community submission (unverified)

On August 27, 2026, a submitted argument claims that an explicit 9×99\times 9 matrix has the required property and determinant 68>49/3=6468>4^{9/3}=64, which would refute the conjecture. The calculation and verification are not independently confirmed.

Current status (as of August 2026): the upper bound remains unproved, while an unverified community submission claims a counterexample at n=9n=9.

Sources

Solutions 1

This solution needs a summarySee full solutionHide full solution

A Counterexample to MathDB 337797

1. Counterexample

MathDB 337797 states the following conjecture. If A∈{0,1}n×nA\in\{0,1\}^{n\times n} has the 2-consecutive-ones property, meaning that there is a common ordering of the columns in which the ones in every row form at most two contiguous blocks, then

∣det⁡A∣≤4n/3.|\det A|\le 4^{n/3}.

Consider the following 9×99\times 9 matrix:

A=(011101100000111010011110001111100010001001111000100111001110110110011110101111100).A= \begin{pmatrix} 0&1&1&1&0&1&1&0&0\\ 0&0&0&1&1&1&0&1&0\\ 0&1&1&1&1&0&0&0&1\\ 1&1&1&1&0&0&0&1&0\\ 0&0&1&0&0&1&1&1&1\\ 0&0&0&1&0&0&1&1&1\\ 0&0&1&1&1&0&1&1&0\\ 1&1&0&0&1&1&1&1&0\\ 1&0&1&1&1&1&1&0&0 \end{pmatrix}.

This matrix has the 2-consecutive-ones property, but

∣det⁡A∣=68>64=49/3.|\det A|=68>64=4^{9/3}.

Therefore, AA is a counterexample to the conjecture stated in MathDB 337797.

2. Proof

We first verify the 2-consecutive-ones property. Keep the columns in the displayed common order, and use the closed interval [a,b][a,b] to denote a contiguous block of ones from column aa through column bb. The complete list of blocks in each row is as follows:

RowContiguous blocks of ones
1[2,4][2,4], [6,7][6,7]
2[4,6][4,6], [8,8][8,8]
3[2,5][2,5], [9,9][9,9]
4[1,4][1,4], [8,8][8,8]
5[3,3][3,3], [6,9][6,9]
6[4,4][4,4], [7,9][7,9]
7[3,5][3,5], [7,8][7,8]
8[1,2][1,2], [5,8][5,8]
9[1,1][1,1], [3,7][3,7]

Thus every row has exactly two contiguous blocks of ones in the same column order. Hence AA has the 2-consecutive-ones property.

It remains to compute the determinant exactly. Perform ordinary Gaussian elimination over Q\mathbb{Q}: in each column, choose the uppermost available nonzero entry as the pivot; interchange rows when necessary; and eliminate entries below the pivot using only operations of the form

Ri⟵Ri−qRj(q∈Q).R_i\longleftarrow R_i-qR_j\qquad(q\in\mathbb{Q}).

The five row interchanges, recorded as row positions at the time they occur, are

(1,4), (2,3), (3,5), (4,5), (7,8).(1,4),\ (2,3),\ (3,5),\ (4,5),\ (7,8).

The elimination produces the upper-triangular matrix

U=(1111000100111100010010011110001110100000−1110−100000−20020000004240000000−1−300000000172).U= \begin{pmatrix} 1&1&1&1&0&0&0&1&0\\ 0&1&1&1&1&0&0&0&1\\ 0&0&1&0&0&1&1&1&1\\ 0&0&0&1&1&1&0&1&0\\ 0&0&0&0&-1&1&1&0&-1\\ 0&0&0&0&0&-2&0&0&2\\ 0&0&0&0&0&0&4&2&4\\ 0&0&0&0&0&0&0&-1&-3\\ 0&0&0&0&0&0&0&0&\frac{17}{2} \end{pmatrix}.

Adding a multiple of one row to another does not change the determinant, whereas every row interchange reverses its sign. Since the determinant of an upper-triangular matrix is the product of its diagonal entries,

det⁡U=1⋅1⋅1⋅1⋅(−1)⋅(−2)⋅4⋅(−1)⋅172=−68.\det U =1\cdot1\cdot1\cdot1\cdot(-1)\cdot(-2)\cdot4\cdot(-1)\cdot\frac{17}{2} =-68.

There were five row interchanges, so

det⁡A=(−1)5det⁡U=68.\det A=(-1)^5\det U=68.

For n=9n=9, the conjectured upper bound is

4n/3=49/3=43=64.4^{n/3}=4^{9/3}=4^3=64.

Consequently,

∣det⁡A∣=68>64=49/3,|\det A|=68>64=4^{9/3},

which proves that AA is a counterexample.