Properties of the Tropical Characteristic Polynomial of Symmetric Matrices

Summary
This paper investigates the coefficient sequence (deltam) of the tropical characteristic polynomial the corresponding equation in the paper for symmetric matrices A in mathbbRmax^ntimes n. Building on the combinatorial interpretation of delta_m as the tropical permanent of the best mtimes m principal submatrix (Theorem 1), the authors derive new inequalities for m = 3 and m = 4—namely, at least one of two concavity-type conditions must hold in each case (Theorems 3 and 4). These yield necessary conditions for a sequence to arise from a symmetric matrix, and they constrain the Newton polygon’s local shape, ruling out certain “impossible” saturation patterns (Figures 5–6).

Mathematical/empirical assessment
The proofs rely cleanly on symmetry-driven cycle decompositions: Lemma 2 (even cycles dominate transposition products) and Lemma 3 (odd-cycle weight bounds via delta_2) underpin the case analysis in Theorem 4. The inequalities are tight in the sense that counterexamples exist for nonsymmetric matrices (the final example), confirming their structural specificity. While the paper does not provide empirical validation or numerical experiments, the theoretical derivation is self-contained and internally consistent; all key steps reference only the given definitions and prior lemmas (e.g., Eq. (3)–(5), Lemma 4, Theorem 1).

Strengths
The work delivers exactly what its abstract promises: a deeper understanding of how symmetry shapes the coefficient sequence of tropical characteristic polynomials. The concavity-style inequalities are novel, well-motivated, and elegantly connected to geometric intuition via the Newton polygon. The exposition is clear and pedagogically effective—especially the illustrative figures (e.g., Figure 2 for Lemma 2) and the logical sketch in Figure 4. Framing the results as necessary conditions (rather than full characterizations) reflects appropriate scholarly restraint.

Concerns
The analysis is currently limited to delta0 through delta4, and the conjecture extending the pattern to all m (Conjecture 1) remains open. While the paper rightly identifies this as future work, strengthening the manuscript would benefit from even a heuristic discussion of why the m=5 case resists the current techniques—e.g., whether new cycle types (like a 5-cycle or mixed 2+3 decompositions) introduce combinatorial obstructions not present for mle4. Also, while Figure 5 visually supports the “at most one non-saturated index between saturated ones”, a brief verbal summary of that implication—tying it directly to the inequalities—would help readers bridge the algebraic and geometric perspectives more smoothly.

Final decision
Weak accept