← All problems

Boros–Rózsa singular-value interlacing conjecture

For every integer m≥2m\ge 2, let Q(m)∈R(m+1)×mQ^{(m)}\in\mathbb{R}^{(m+1)\times m} and S(m)∈Rm×(m+1)S^{(m)}\in\mathbb{R}^{m\times(m+1)} be the matrices arising in the Boros–Rózsa reduction of the odd-order Sylvester–Kac matrix. If σ1(⋅)≥σ2(⋅)≥⋯\sigma_1(\cdot)\ge\sigma_2(\cdot)\ge\cdots denotes the singular values in nonincreasing order, then σi ⁣(S(m))≥σi+1 ⁣(Q(m))\sigma_i\!\left(S^{(m)}\right)\ge\sigma_{i+1}\!\left(Q^{(m)}\right) for every 1≤i≤m−11\le i\le m-1.

References

Progress summary

Refreshed
Open · solution reported

A new preprint claims to prove the conjecture, but the result has not been independently checked.

The Boros–Rózsa conjecture concerns the ordering of singular values for a structured matrix family. No proposer or date is identified in the retrieved material.

New preprint

Fuying Tang and Shuo Shi claim the stronger strict ordering

σ1(Q)>σ1(S)>⋯>σm(Q)>σm(S)>0.\sigma_1(Q)>\sigma_1(S)>\cdots>\sigma_m(Q)>\sigma_m(S)>0.

If correct, this proves the conjectured non-strict interlacing. The claim is supported only by the preprint and has not been independently assessed.

Current status (as of October 2026): Tang and Shi claim a complete proof, but without independent verification the conjecture remains open as a verified result.

Sources

Solutions 0

No solutions have been posted yet.