Investigating Quantum-Embedded Transformers on Classical Datasets for Cross-Modality Classification

2026-08-07

AI Summary

This study investigates whether parameterized quantum circuits can enhance the cross-modal classification performance of hybrid quantum-classical models on classical data. To this end, the authors propose the Quantum Embedding Attention (QEA) architecture, which compresses backbone features into learnable angle vectors, maps them through a shallow parameterized quantum circuit to Pauli expectation values, and decodes class logits via a classical attention module. A carefully designed classical counterpart model is introduced to isolate and rigorously attribute any performance gains specifically to the quantum component. Systematic evaluations across multiple datasets and modalities reveal that the quantum circuit yields no significant performance improvement in the vast majority of settings; only a marginal and non-robust advantage is observed in a single configuration (n_q=4), failing to establish either quantum advantage or equivalence overall.

Abstract

We test whether a parameterized quantum circuit (PQC) improves a hybrid quantum-classical model's performance on classical datasets, using an interface-matched classical map as the control while holding all other components fixed. Our architecture, Quantum-Embedded Attention (QEA), uses a learnable projector to compress backbone features into an $n_q$-dimensional angle vector, a shallow PQC to map those angles to one- and two-qubit Pauli expectations, and a classical attention decoder to produce class logits. We hypothesized the PQC would improve accuracy or seed-to-seed stability over a classical map with matched input/output dimensions. We test this with an interface-matched $2\times2$ factorial on Breast Cancer Wisconsin at $n_q\in\{4,8\}$, independently swapping the PQC for a classical map and the attention decoder for a linear head, across five paired seeds per cell. Three of four paired quantum-minus-classical $95\%$ confidence intervals include zero; the fourth, a $+1.63$ percentage-point contrast for the attention decoder at $n_q=4$, reverses sign at $n_q=8$ and does not survive correction across the four contrasts. The experiment thus shows no consistent PQC contribution and cannot establish equivalence. A five-dataset cross-modality grid shows comparable accuracy on AG~News, Breast Cancer Wisconsin, and BirdCLEF but a large deficit on CIFAR-10; these cells are not interface-matched and are interpreted descriptively. We report all planned canonical runs, distinguish current Pauli-readout results from legacy probability-readout experiments, and analyze bottleneck, simulation, finite-shot, and noise limitations. The results do not establish a quantum advantage; they demonstrate why controlled component attribution is necessary before crediting a hybrid model's performance to its quantum layer.

Problem

• quantum-classical hybrid
• parameterized quantum circuit
• cross-modality classification
• quantum advantage
• classical datasets

Innovation

• Quantum-Embedded Attention
• parameterized quantum circuit
• controlled component attribution
• hybrid quantum-classical model
• Pauli expectation readout