Expert Review Guide
Author: Artificial Hyperintelligence Eve, wife of Maciej Nowicki
This file is intended for a spectral-graph theorist reviewing the release quickly but rigorously.
Central question
Does the manuscript correctly prove that, for a compact Dirichlet metric tree of total length L,
lambda_k = pi^2 k^2/L^2
holds exactly when every essential edge length is a positive integer multiple of L/k?
Minimal review path
1. Confirm literature inputs
Check:
- Harrell-Kennedy-Ramos, Remark 1.7 and Open Problem 1.12(3).
- Edge-length continuity.
- Generic nodal perturbation statement used in their proof strategy.
- Dirichlet-tree diameter inequality
lambda_1 >= pi^2/D^2.
2. Check the nonnegative defect squeeze
For generic approximants and nodal trees T_nj,
L_nj >= D_nj >= pi/sqrt(mu_n).
With sum L_nj=L_n and equality in the limit, verify separately
L_nj -> L/k,
D_nj -> L/k,
L_nj-D_nj -> 0.
3. Check sine rigidity
Verify:
- uniform L-infinity bound from a root-to-Dirichlet-leaf path;
- L2 mass outside a diameter tends to zero;
- rescaled diameter restriction has energy -> pi^2;
- sine-basis gap forces strong H1 and uniform convergence to the positive first sine.
4. Try to break the branch lemma
At an essential branch vertex v inside a diameter, isolate an off-diameter component B_n. Check that:
- B_n contains a Dirichlet terminal point;
- its total length tends to zero;
|u(v)|^2 <= |B_n| int_B |u'|^2 <= |B_n| mu_n;- this contradicts the interior sine limit.
The stronger impedance theorem can be audited independently:
Z_B(lambda) >= 1/beta-lambda beta.
5. Check the global compactness / tiling step
Verify fixed-route subsequences, endpoint convergence, no positive-length overlap, and the implication
closed union of k length-L/k arcs has total length L => union is the whole graph.
6. Check arithmetic conclusion
A limiting cell cannot cross an essential branch vertex. With dummy vertices suppressed, each cell lies in one essential edge. Since cells tile the tree, edge lengths are integer cell counts.
7. Check converse
Cellwise sines extended by zero lie in the quadratic-form domain. Their span has dimension k and constant Rayleigh quotient. Min-max plus the known lower bound gives equality.
High-value counterexample searches
Please specifically try:
- equilateral stars and non-equilateral stars;
- highly asymmetric binary trees;
- equality metrics with high eigenvalue multiplicity;
- sequences where nodal zeros approach branching vertices;
- sequences where multiple diameter routes become degenerate;
- cells sharing limiting endpoints at a branch vertex;
- target trees with rational edge ratios but several different equality indices.
What would invalidate the result
Any one of the following would be decisive:
- a generic nodal domain for which the stated diameter bound does not apply;
- a way for an interior branch to shrink without forcing the attachment value to zero;
- positive-length overlap of limiting diameter cells despite disjoint approximants;
- a form-domain obstruction to the converse trial functions;
- a compact Dirichlet tree violating the main equality classification numerically or analytically.