Universal Spectral-Curvature Phase Diagrams of Graphs: Predicting Non-Euclidean Embedding Distortion and Optimal GNN Geometry via Gram-Eigenvalue Decomposition

Universal two-dimensional spectral-curvature phase diagram (g3, d_eff+) derived from the positive and negative eigenspaces of double-centered squared-distance Gram matrices. Proves the nuclear-norm range identity g3 in [0, 1), the exact diameter-2 complement-spectrum and random-graph asymptotic laws, and structural failure modes of Gromov delta-hyperbolicity, validated across 20 canonical manifolds and 3,246 real-world GNN benchmark graphs.

Authors

Publication Details

Journal
Figshare
Published
2026-11-20
DOI
https://doi.org/10.6084/m9.figshare.35069164
Primary Topic
Advanced Graph Neural Networks
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Universal Spectral-Curvature Phase Diagrams of Graphs: Predicting Non-Euclidean Embedding Distortion and Optimal GNN Geometry via Gram-Eigenvalue Decomposition

Shoji Fujita
2 links
Figshare
Advanced Graph Neural Networks
preprint

Universal Spectral-Curvature Phase Diagrams of Graphs: Predicting Non-Euclidean Embedding Distortion and Optimal GNN Geometry via Gram-Eigenvalue Decomposition

Shoji Fujita
preprint en

Abstract

Universal two-dimensional spectral-curvature phase diagram (g3, d_eff+) derived from the positive and negative eigenspaces of double-centered squared-distance Gram matrices. Proves the nuclear-norm range identity g3 in [0, 1), the exact diameter-2 complement-spectrum and random-graph asymptotic laws, and structural failure modes of Gromov delta-hyperbolicity, validated across 20 canonical manifolds and 3,246 real-world GNN benchmark graphs.

Figshare
Advanced Graph Neural Networks
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
Universal Spectral-Curvature Phase Diagrams of Graphs: Predicting Non-Euclidean Embedding Distortion and Optimal GNN Geometry via Gram-Eigenvalue Decomposition — Shoji Fujita · Figshare (2026) | TGRS Research Map | TGRS