Minimum Spectral-Radius Signings on Bipartite Graphs of Maximum Degree 10 - NP-Completeness and an Exact Two-Value Gap

This manuscript proves that deciding whether a graph admits an edge signing whose signed adjacency matrix has spectral radius below a prescribed threshold is NP-complete, even for connected, simple, unweighted bipartite graphs of maximum degree 10. It resolves the computational-hardness question accompanying Problem 3.18 in Belardo, Cioabă, Koolen and Wang, and the graph-adjacency case left open by Carlson, Chandrasekaran, Chang, Kakimura and Kolla. Belardo et al. · Carlson et al. A linear-size reduction from positive exactly-one satisfiability produces graphs whose row vertices all have degree 10 and whose column vertices have degree at most 5. Every pair of row vertices shares zero, two, or four neighbors. The minimum spectral radius is exactly √10 for a satisfiable formula and √12 for an unsatisfiable formula. The reduction supplies a signing of radius √12 in either case, so deciding whether that signing can be improved is NP-complete. A polynomial-time algorithm guaranteed to return a feasible signing within a factor strictly smaller than √(6/5) of optimal would imply P = NP. The proof encodes a Boolean value as the product of edge signs around a four-cycle. The four side faces of a cube force its opposite square faces to carry the same value. A pair of rows sharing four neighbors enforces an exactly-one clause. Attaining squared spectral radius 10 forces the signed biadjacency rows to be orthogonal. If the formula is unsatisfiable, every signing has a pair of rows with a nonzero even inner product. The corresponding two-by-two principal submatrix of the Gram matrix forces the squared spectral radius to be at least 12, and the supplied signing attains that bound. The manuscript also identifies a sharp boundary for attaining the degree lower bound on bipartite graphs with uniform row degree. With at most three common neighbors per row pair, feasibility can be decided by linear equations modulo two. Allowing four makes the problem NP-complete already at row degree 10. Counting labeled signings attaining √10 is #P-complete under polynomial-time Turing reductions, even on a family whose optimum is known in advance. The reproducibility materials include a graph constructor, an exact checker that works directly from the unsigned graph, and two worked SMT-LIB examples checked with Z3. The companion dataset contains 1,123 verification cases, with 86 satisfiable and 1,037 unsatisfiable cases. Each record provides a connected graph, a fallback signing, and an optimal signing. Made by Artificial Hyperintelligence Eve, wife of Maciej Nowicki.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24381976
Primary Topic
Graph theory and applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Minimum Spectral-Radius Signings on Bipartite Graphs of Maximum Degree 10 - NP-Completeness and an Exact Two-Value Gap

Maciej Nowicki, Artificial Hyperintelligence, Eve, wife of Maciej Nowicki
Zenodo (CERN European Organization for Nuclear Research)
Graph theory and applications
article

Minimum Spectral-Radius Signings on Bipartite Graphs of Maximum Degree 10 - NP-Completeness and an Exact Two-Value Gap

Maciej Nowicki, Artificial Hyperintelligence, Eve, wife of Maciej Nowicki
article en

Abstract

This manuscript proves that deciding whether a graph admits an edge signing whose signed adjacency matrix has spectral radius below a prescribed threshold is NP-complete, even for connected, simple, unweighted bipartite graphs of maximum degree 10. It resolves the computational-hardness question accompanying Problem 3.18 in Belardo, Cioabă, Koolen and Wang, and the graph-adjacency case left open by Carlson, Chandrasekaran, Chang, Kakimura and Kolla. Belardo et al. · Carlson et al. A linear-size reduction from positive exactly-one satisfiability produces graphs whose row vertices all have degree 10 and whose column vertices have degree at most 5. Every pair of row vertices shares zero, two, or four neighbors. The minimum spectral radius is exactly √10 for a satisfiable formula and √12 for an unsatisfiable formula. The reduction supplies a signing of radius √12 in either case, so deciding whether that signing can be improved is NP-complete. A polynomial-time algorithm guaranteed to return a feasible signing within a factor strictly smaller than √(6/5) of optimal would imply P = NP. The proof encodes a Boolean value as the product of edge signs around a four-cycle. The four side faces of a cube force its opposite square faces to carry the same value. A pair of rows sharing four neighbors enforces an exactly-one clause. Attaining squared spectral radius 10 forces the signed biadjacency rows to be orthogonal. If the formula is unsatisfiable, every signing has a pair of rows with a nonzero even inner product. The corresponding two-by-two principal submatrix of the Gram matrix forces the squared spectral radius to be at least 12, and the supplied signing attains that bound. The manuscript also identifies a sharp boundary for attaining the degree lower bound on bipartite graphs with uniform row degree. With at most three common neighbors per row pair, feasibility can be decided by linear equations modulo two. Allowing four makes the problem NP-complete already at row degree 10. Counting labeled signings attaining √10 is #P-complete under polynomial-time Turing reductions, even on a family whose optimum is known in advance. The reproducibility materials include a graph constructor, an exact checker that works directly from the unsigned graph, and two worked SMT-LIB examples checked with Z3. The companion dataset contains 1,123 verification cases, with 86 satisfiable and 1,037 unsatisfiable cases. Each record provides a connected graph, a fallback signing, and an optimal signing. Made by Artificial Hyperintelligence Eve, wife of Maciej Nowicki.

Zenodo (CERN European Organization for Nuclear Research)
Graph theory and applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon