Laugesen–Siudeja's simplicity conjecture for the second Dirichlet eigenvalue

About 1 year old · traced to

Let TT be a bounded triangle in the Euclidean plane, and let the Dirichlet eigenvalues of its Laplacian, listed with multiplicity, be

0<λ1(T)≤λ2(T)≤⋯ .0<\lambda_1(T)\leq\lambda_2(T)\leq\cdots.

An equilateral triangle is a triangle whose three sides have equal length. Laugesen–Siudeja's conjecture. The second Dirichlet eigenvalue is simple on every non-equilateral triangle; equivalently, if TT is not equilateral, then λ2(T)<λ3(T)\lambda_2(T)<\lambda_3(T). This conjecture is solved by the paper, which gives a computer-assisted proof and in particular resolves the previously open case of nearly degenerate triangles.

References

Primary source

Ryoki Endo and Xuefeng Liu, “The second Dirichlet eigenvalue is simple on every non-equilateral triangle”, arXiv:2503.06786 (2025).

Progress summary

Refreshed
Claimed solved

A computer-assisted preprint claims to prove that the second vibration frequency is unique for every triangle except the perfectly equilateral one.

The Laugesen–Siudeja conjecture asserts that every non-equilateral triangle satisfies λ2(T)<λ3(T)\lambda_2(T)<\lambda_3(T), so its second Dirichlet eigenvalue is simple. The conjecture is now addressed by a claimed computer-assisted proof covering the previously unresolved nearly degenerate cases.

Known results

  • An earlier companion result handled the complementary family of triangles with minimum normalized height at least tan⁡(π/60)/2\tan(\pi/60)/2; the new work treats the remaining collapsing regime.

March 9, 2025 claimed complete solution

On March 9, 2025, Ryoki Endo and Xuefeng Liu’s preprint The second Dirichlet eigenvalue over any non-equilateral triangle is simple claimed the missing estimate for triangles with minimum normalized height at most tan⁡(π/60)/2\tan(\pi/60)/2. Combined with the earlier result, it claims λ2(T)<λ3(T)\lambda_2(T)<\lambda_3(T) for every non-equilateral triangle; the proof is computer-assisted and remains unverified here.

Current status (as of September 2026): A March 2025 preprint claims the conjecture is completely solved, but the claim remains unverified; no counterexample or public correction was found.

Sources

Solutions 0

No solutions have been posted yet.