Spectral radius and Hamiltonicity of uniform hypergraphs
Pith reviewed 2026-05-22 17:57 UTC · model grok-4.3
The pith
Any r-uniform hypergraph with spectral radius above binom(n-2, r-1) has a Hamiltonian Berge cycle unless it is the near-complete exception.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We prove that any r-uniform hypergraph H on n vertices with spectral radius λ(H) > binom(n-2, r-1) must contain a Hamiltonian Berge cycle unless H is the complete graph K_{n-1}^r with one additional edge. This generalizes a result proved by Fiedler and Nikiforov for graphs. As part of our proof, we show that if |H| > binom(n-1, r), then H contains a Hamiltonian Berge cycle unless H is the complete graph K_{n-1}^r with one additional edge, generalizing a classical theorem for graphs.
What carries the argument
Spectral radius of the adjacency tensor of the r-uniform hypergraph, which exceeds the value binom(n-2, r-1) to force the presence of a Hamiltonian Berge cycle.
If this is right
- If the number of edges exceeds binom(n-1, r), the hypergraph contains a Hamiltonian Berge cycle unless it is the stated exception.
- The spectral condition extends the Fiedler-Nikiforov theorem from graphs to uniform hypergraphs.
- The edge-count condition extends a classical graph theorem to hypergraphs.
Where Pith is reading between the lines
- The threshold could support computational checks for Hamiltonicity by calculating the spectral radius rather than searching for cycles directly.
- Analogous spectral conditions might apply to other hypergraph properties such as perfect matchings or connectivity.
- The uniqueness of the exception suggests the bound is tight and could guide further refinements in extremal hypergraph problems.
Load-bearing premise
The spectral radius defined via the adjacency tensor increases with added edges such that the stated extremal hypergraph is the unique maximizer without the cycle below the threshold.
What would settle it
Construct or exhibit an r-uniform hypergraph on n vertices that is not K_{n-1}^r plus one edge, has no Hamiltonian Berge cycle, yet has spectral radius strictly larger than binom(n-2, r-1).
read the original abstract
Let $n$ and $r$ be integers with $n-2\ge r\ge 3$. We prove that any $r$-uniform hypergraph $\mathcal{H}$ on $n$ vertices with spectral radius $\lambda(\mathcal{H}) > \binom{n-2}{r-1}$ must contain a Hamiltonian Berge cycle unless $\mathcal{H}$ is the complete graph $K_{n-1}^r$ with one additional edge. This generalizes a result proved by Fiedler and Nikiforov for graphs. As part of our proof, we show that if $|\mathcal{H}| > \binom{n-1}{r}$, then $\mathcal{H}$ contains a Hamiltonian Berge cycle unless $\mathcal{H}$ is the complete graph $K_{n-1}^r$ with one additional edge, generalizing a classical theorem for graphs.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves that for integers n and r with n-2 ≥ r ≥ 3, any r-uniform hypergraph H on n vertices satisfying λ(H) > binom(n-2, r-1) contains a Hamiltonian Berge cycle unless H is exactly K_{n-1}^r with one additional edge. As an auxiliary result used in the proof, it also shows that the same conclusion holds whenever the number of edges exceeds binom(n-1, r). Both statements generalize the corresponding theorems of Fiedler and Nikiforov from the graph case (r=2).
Significance. If the technical details hold, the results supply a clean spectral sufficient condition for Hamiltonicity in uniform hypergraphs together with an explicit extremal example, extending classical graph-theoretic work into the hypergraph setting and linking spectral radius techniques with Berge-cycle problems in extremal combinatorics.
major comments (2)
- [Proof of the spectral-radius statement (likely §4)] The central reduction from the spectral-radius hypothesis to the edge-count hypothesis (used to invoke the auxiliary theorem) requires proving that binom(n-2, r-1) is the maximum possible spectral radius among all r-uniform hypergraphs with at most binom(n-1, r) edges and that this maximum is achieved uniquely by K_{n-1}^r. The manuscript invokes monotonicity of λ under edge addition for the adjacency tensor, but does not supply a self-contained argument ruling out equality or larger values for other (possibly disconnected) hypergraphs with the same number of edges; this step is load-bearing for both main theorems.
- [Extremal spectral-radius lemma preceding the main theorems] The uniqueness claim for the maximizer K_{n-1}^r and the strict increase of λ upon adding an edge to it rest on properties of the Perron vector of the adjacency tensor. No explicit verification is given that the vector remains positive or that the Rayleigh quotient strictly increases in the exceptional case; without this, the strict inequality λ(H) > binom(n-2, r-1) may not force |E(H)| > binom(n-1, r) as asserted.
minor comments (2)
- [§2] The definition of the adjacency tensor and the precise normalization used for the spectral radius should be restated in the preliminaries so that the comparison λ(H) > binom(n-2, r-1) is immediately interpretable without consulting external references.
- [Introduction] A short remark clarifying how the Berge-cycle notion reduces to the ordinary Hamilton cycle when r=2 would help readers see the exact scope of the generalization.
Simulated Author's Rebuttal
We thank the referee for the thorough review and valuable feedback on our paper. The comments point to areas where the proof can be clarified and strengthened. We address each major comment below and will make the necessary revisions to the manuscript.
read point-by-point responses
-
Referee: The central reduction from the spectral-radius hypothesis to the edge-count hypothesis (used to invoke the auxiliary theorem) requires proving that binom(n-2, r-1) is the maximum possible spectral radius among all r-uniform hypergraphs with at most binom(n-1, r) edges and that this maximum is achieved uniquely by K_{n-1}^r. The manuscript invokes monotonicity of λ under edge addition for the adjacency tensor, but does not supply a self-contained argument ruling out equality or larger values for other (possibly disconnected) hypergraphs with the same number of edges; this step is load-bearing for both main theorems.
Authors: We appreciate the referee's observation regarding the need for a more explicit justification in the reduction step. While the monotonicity property of the spectral radius with respect to edge addition is known for hypergraph adjacency tensors, we acknowledge that a self-contained argument is desirable. In the revised version of the manuscript, we will insert a new lemma that proves λ(H) ≤ binom(n-2, r-1) for any r-uniform hypergraph H with |E(H)| ≤ binom(n-1, r), with equality holding if and only if H is isomorphic to K_{n-1}^r. This will include arguments showing that any other hypergraph, connected or disconnected, cannot exceed this bound, thereby supporting the strict inequality in the main theorems. revision: yes
-
Referee: The uniqueness claim for the maximizer K_{n-1}^r and the strict increase of λ upon adding an edge to it rest on properties of the Perron vector of the adjacency tensor. No explicit verification is given that the vector remains positive or that the Rayleigh quotient strictly increases in the exceptional case; without this, the strict inequality λ(H) > binom(n-2, r-1) may not force |E(H)| > binom(n-1, r) as asserted.
Authors: We agree that the details concerning the Perron vector and the strict monotonicity should be verified explicitly in the text. In the revision, we will expand the extremal spectral-radius lemma to include a direct argument showing that the Perron vector of K_{n-1}^r can be chosen with all positive entries on the n-1 vertices, and that adding any edge to this hypergraph results in a strict increase in the Rayleigh quotient. This will rigorously establish that λ(H) > binom(n-2, r-1) implies |E(H)| > binom(n-1, r). revision: yes
Circularity Check
Direct combinatorial proof against external spectral threshold with no self-referential reduction
full rationale
The paper derives the Hamiltonicity threshold by direct comparison of λ(H) to the fixed binomial coefficient binom(n-2,r-1), which is the spectral radius of the extremal complete hypergraph K_{n-1}^r. This comparison is external to the paper's own constructions and generalizes the Fiedler-Nikiforov graph result without fitting parameters or renaming known patterns. The uniqueness of the exception case K_{n-1}^r plus one edge is established inside the proof via edge-addition monotonicity of the adjacency tensor, rather than imported via self-citation. No step reduces by construction to its own inputs; the argument remains self-contained against standard tensor definitions and classical extremal combinatorics.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math The spectral radius of an r-uniform hypergraph is defined via the largest eigenvalue of its adjacency tensor.
- domain assumption A Berge cycle is a sequence of distinct vertices and edges where each consecutive pair is contained in the corresponding edge.
Lean theorems connected to this paper
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
λ(H) = max_{||x||_r=1} P_H(x) where P_H(x) = r ∑_{e∈E} ∏ x_i; Bai-Lu: λ(H) ≤ f_r(m) with equality only for complete + isolates
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.