How Embeddings Shape Graph Neural Networks: Classical vs Quantum-Oriented Node Representations

2026-04-16

AI Summary

This study addresses the lack of systematic comparison between classical and quantum-inspired node embedding approaches under unified experimental conditions. For the first time, it systematically evaluates three classes of embedding methods—classical baselines, variational quantum circuit embeddings, and quantum-inspired embeddings—in graph classification tasks, while strictly controlling backbone architectures, data splits, optimization strategies, and evaluation metrics. The experiments reveal that quantum-inspired embeddings achieve superior performance and greater training stability on structurally dense datasets, such as discretized QM9 and certain TU benchmarks, yet underperform classical methods on attribute-sparse social graphs. These findings highlight a critical trade-off between inductive bias and training stability, providing an empirical foundation for advancing quantum graph representation learning.

Technology Category

Machine Learning: Quantum Machine LearningData Mining & Knowledge Management: Graph Mining, Social Network Analysis & CommunityKnowledge Representation and Reasoning: Qualitative Reasoning

Application Category

Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphsSearch and Retrieval-Augmented AI: Web evaluation methodologies and metricsSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Node embeddings act as the information interface for graph neural networks, yet their empirical impact is often reported under mismatched backbones, splits, and training budgets. This paper provides a controlled benchmark of embedding choices for graph classification, comparing classical baselines with quantum-oriented node representations under a unified pipeline. We evaluate two classical baselines alongside quantum-oriented alternatives, including a circuit-defined variational embedding and quantum-inspired embeddings computed via graph operators and linear-algebraic constructions. All variants are trained and tested with the same backbone, stratified splits, identical optimization and early stopping, and consistent metrics. Experiments on five different TU datasets and on QM9 converted to classification via target binning show clear dataset dependence: quantum-oriented embeddings yield the most consistent gains on structure-driven benchmarks, while social graphs with limited node attributes remain well served by classical baselines. The study highlights practical trade-offs between inductive bias, trainability, and stability under a fixed training budget, and offers a reproducible reference point for selecting quantum-oriented embeddings in graph learning.

Problem

• node embeddings
• graph neural networks
• graph classification
• quantum-oriented representations
• controlled benchmark

Innovation

• graph neural networks
• node embeddings
• quantum-oriented representations
• controlled benchmarking
• graph classification