← All problems

Positivity of real H-eigenvalues for reflected order-four Steiner distance tensors

Let TT be a finite tree on n≥2n\ge2 vertices. For an ordered quadruple of vertices, including repeated terminals, let MijklM_{ijkl} be the minimum number of edges in a connected subgraph containing those terminals. Let JJ be the all-ones n×nn\times n matrix and A=2J/n−IA=2J/n-I. Reflect the last slot:

(C1)ijkl=∑hMijkhAhl.(C_1)_{ijkl}=\sum_h M_{ijkh}A_{hl}.

Using the first-free-index H-eigenvalue convention, a real H-eigenpair is a real number λ\lambda and a nonzero vector x∈Rnx\in\mathbb R^n such that

∑j,k,l(C1)ijklxjxkxl=λxi3for every i.\sum_{j,k,l}(C_1)_{ijkl}x_jx_kx_l=\lambda x_i^3\quad\text{for every }i.

Positivity question. Must every such real H-eigenvalue be positive for every tree TT?

This is the order-four case of the reflected-tensor positivity question on pp. 9–10 of Cooper–Du. The reflection need not leave a fully symmetric tensor, so the coordinate convention is explicit. This question is separate from the hyperdeterminant-sign conjecture and from positivity results for C1+C3C_1+C_3.

References

J. Cooper and Z. Du, Determinants of Steiner Distance Hypermatrices, arXiv:2505.10501v1 (2025), Theorem 2.1 and pp. 9–10. https://arxiv.org/abs/2505.10501v1 L. Qi, Eigenvalues of a real supersymmetric tensor, J. Symbolic Comput. 40 (2005), 1302–1324. https://doi.org/10.1016/j.jsc.2005.05.007

Progress summary

Refreshed
Open · progress reported

The general positivity question remains open, although a reader-submitted claim alleges a counterexample on a 30-vertex star.

Cooper and Du posed whether every real HH-eigenvalue of the reflected order-four tensor C1C_1 is positive for every finite tree, in 2025. Their paper gives computational evidence for positivity but no proof or counterexample.

Known results

  • Cooper–Du (2025): C1C_1 was computationally observed to be positive definite in tested cases.
  • Cooper–Du (2025): the distinct tensor C1+C3C_1+C_3 is positive definite for order 44.
  • Cooper–Du (2025): for even order, the Steiner distance hypermatrix has conditional strict negative definiteness on the orthogonal complement of the all-ones vector.

November 7, 2026 community counterexample claim

A submitted argument claims that the star K1,29K_{1,29} yields a negative real first-free-index HH-eigenvalue for C1C_1, which would refute universal positivity. The claim is unverified; no independent mathematical assessment was retrieved.

Current status (as of November 2026): positivity is established only for related constructions and computationally in reported tests; the stated question remains open because the submitted K1,29K_{1,29} counterexample has not been independently verified.

Sources

Solutions 1

On the genuine star K1,29 at order4, M from minimum connected-subgraph edge count including repeated terminals and C1=M(.,.,.,(2J/30-I).) has a negative real first-free-index H-eigenpair. One counterexample refutes universal positivity.See full solutionHide full solution

Result. On the genuine star K1,29 at order4, M from minimum connected-subgraph edge count including repeated terminals and C1=M(.,.,.,(2J/30-I).) has a negative real first-free-index H-eigenpair. One counterexample refutes universal positivity.

Paper and complete proof. A negative H-eigenvalue of a reflected Steiner distance tensor. The paper page provides the proof PDF and its accompanying source and verification package.

Scope and source. Cooper–Du arXiv2505.10501v1, final positivity question pp9–10. The result addresses exactly the question stated above; adjacent conjectures and broader formulations are not claimed.

Provenance. AI-assisted research. Formalization and verification claims are only those explicitly established in the paper and its attached artifacts; this entry does not assert MathDB endorsement or independent acceptance.