Quantum transfer learning (QTL) is often evaluated by replacing a classical classifier with a fixed variational quantum head, but this hides a key question: when is the quantum branch actually useful? We propose QSTAR: Quantum Selective Transfer with Adaptive Routing, a selective QTL framework that keeps high-confidence classical predictions and routes only low-confidence samples to a fallback branch. Using a frozen ResNet18 backbone on Fashion-MNIST, we compare manually designed QTL heads, KetGPT-designed quantum heads, and parameter-matched classical baselines under a common data split and optimization schedule. Standard QTL heads reach at most 57.0% accuracy, while the strongest KetGPT head in the main filtered sweep reaches 78.5% accuracy and 0.785 F1-score. Although the strongest fixed classical head remains higher at 81.6%, selective routing gives the quantum branch a clearer role. On low-confidence samples, KetGPT #180 improves accuracy over a parameter-matched MLP fallback by 6.82, 4.31, and 3.03 percentage points at thresholds of 0.70, 0.80, and 0.90. At the full-system level, Adaptive KetGPT-QTL reaches 80.9% accuracy and 0.807 F1-score, outperforming the adaptive classical baseline. A separate compact-circuit ablation identifies KetGPT #160 as a stronger fixed-head candidate, reaching 81.9% accuracy with only 10 quantum parameters and 9 gates. These results suggest that architecture-searched quantum heads are most useful as targeted fallback branches for uncertain inputs rather than uniform replacements for classical classifiers.
Quantum transfer learning (QTL) is often evaluated by replacing a classical classifier with a fixed variational quantum head, but this hides a key question: when is the quantum branch actually useful? We propose QSTAR: Quantum Selective Transfer with Adaptive Routing, a selective QTL framework that keeps high-confidence classical predictions and routes only low-confidence samples to a fallback branch. Using a frozen ResNet18 backbone on Fashion-MNIST, we compare manually designed QTL heads, KetGPT-designed quantum heads, and parameter-matched classical baselines under a common data split and optimization schedule. Standard QTL heads reach at most 57.0% accuracy, while the strongest KetGPT head in the main filtered sweep reaches 78.5% accuracy and 0.785 F1-score. Although the strongest fixed classical head remains higher at 81.6%, selective routing gives the quantum branch a clearer role. On low-confidence samples, KetGPT #180 improves accuracy over a parameter-matched MLP fallback by 6.82, 4.31, and 3.03 percentage points at thresholds of 0.70, 0.80, and 0.90. At the full-system level, Adaptive KetGPT-QTL reaches 80.9% accuracy and 0.807 F1-score, outperforming the adaptive classical baseline. A separate compact-circuit ablation identifies KetGPT #160 as a stronger fixed-head candidate, reaching 81.9% accuracy with only 10 quantum parameters and 9 gates. These results suggest that architecture-searched quantum heads are most useful as targeted fallback branches for uncertain inputs rather than uniform replacements for classical classifiers.
Quantum Transfer Learning (QTL) offers a promising approach for visual quantum machine learning under near-term constraints, where limited qubit counts, shallow circuit depths, and costly hybrid optimization restrict end-to-end quantum training. In this setting, pretrained classical backbones can extract high-level visual features, while compact quantum modules operate as trainable classification heads. However, existing QTL results are difficult to compare because they often differ in datasets, preprocessing, backbone settings, qubit budgets, circuit designs, optimization choices, and reporting protocols. This work presents a controlled benchmarking methodology for evaluating representative QTL methods under a unified transfer-learning pipeline. The benchmark compares DQN-QTL, QPIE-QTL, AE-CQTL, PVCQTL, and ED-QTL under shared preprocessing rules, frozen-backbone settings, training conditions, and reporting metrics. The evaluation focuses on Fashion-MNIST and Hymenoptera Ants vs. Bees as the two main datasets, while CIFAR-10 is used to provide additional configuration-level evidence on a harder natural-image task. Beyond predictive performance, the benchmark analyzes circuit size, trainable parameters, quantum parameters, training time, and architectural sensitivity to qubit count and circuit depth. The results show that no single QTL family dominates across all settings: performance depends on the dataset, encoding strategy, circuit design, and computational cost. These findings highlight the need for resource-aware QTL evaluation and provide guidance for selecting hybrid quantum-classical transfer models under near-term resource constraints.
We address the limitations of variational quantum circuits (VQCs) in hybrid classical-quantum transfer learning by introducing post-variational strategies, which reduce training overhead and mitigate optimization issues. Our approach Post Variational Classical Quantum Transfer Learning (PVCQTL) includes three designs: (1) modified observable construction, (2) a hybrid approach, and (3) a variational-post-variational combination. We evaluate these on pre-trained models (VGG19, ResNet50, ResNet18, MobileNet) for 4 and 8 qubits, with ResNet50 performing best in deepfake detection. Compared to classical models (MLP, ResNet50) and quantum baselines hybrid quantum classical neural network (HQCNN), classical-quantum transfer learning (CQTL). PVCQTL consistently achieves better accuracy. The modified observable variant reaches 85% accuracy for Deepfake dataset with lower computational cost. To evaluate generalizability, we tested PVCQTL on three additional binary classification datasets, observing improved accuracy on each. We conducted ablation studies to assess the effects of architectural choices on quantum component variations, including the choice of quantum gates, use of fixed ansatz circuits, and observable measurements. Robustness to input noise and sensitivity of the PVCQTL models were examined through ablation studies on learning rate, batch size, and number of qubits. These results demonstrate that PVCQTL offers a measurable improvement over traditional hybrid classical-quantum approaches.
Image-based quantum machine learning usually follows a hybrid pipeline: a classical encoder compresses an image, and a quantum circuit then acts as a kernel, feature map, trainable classifier, or measurement-based readout. Yet it remains unclear which of these roles contributes useful predictive structure because prior studies often change the input representation, classical baseline, and measurement budget simultaneously. We present a controlled benchmark spanning five roles: quantum-kernel geometry, fixed product and entangled features, trainable data-reuploading representations, shot-aware probability readout, and task-adaptive Pauli measurement. Six quantum-derived methods, six matched classical controls, and two unmatched neural references are evaluated on MNIST, Fashion-MNIST, CIFAR-10, and BreastMNIST using a shared frozen ResNet-18 and train-only PCA interface. We assess exact and finite-shot inference, predictive performance, calibration, selective risk, resource requirements, and statistical reliability. PdrQC-Matched was the strongest quantum-derived method on all four datasets. C-RBF was better on Fashion-MNIST and CIFAR-10 by 2.19 and 3.50 accuracy points, while BreastMNIST showed no detectable difference. Despite lower exact accuracy, PdrQC-Matched achieved significantly lower Brier score and AURC than C-RBF on Fashion-MNIST and CIFAR-10. Finite-shot results showed no automatic benefit from entanglement or shot-aware training. Overall, image-QML conclusions depend on quantum role, measurement budget, classical control, and evaluation metric.
This chapter presents quantum transfer learning (QTL) as a practical and conceptually grounded framework for integrating quantum computation into modern machine learning. Building on the classical transfer learning paradigm, this chapter shows how pretrained classical models can be reused for large-scale representation learning while variational quantum circuits (VQCs) are deployed as compact downstream adaptation modules. Emphasis is placed on hybrid classical-quantum architectures in which quantum feature extractors operate on low-dimensional embeddings produced by convolutional networks, transformers, or graph neural networks, and their measurement-based outputs are processed by classical heads for prediction. This chapter develops the mathematical and architectural foundations of this approach, explains the role of variational quantum feature maps, and examines why quantum feature extractors may be useful in data-limited and structure-rich settings. Applications to text, image, and molecular data are discussed to illustrate the domain-independent nature of the framework. Practical implementation is addressed through PennyLane and PyTorch integration, highlighting differentiable quantum layers and end-to-end hybrid optimization. A hands-on example using encoded embeddings and VQCs demonstrates methodology, benchmarking, and model evaluation in a realistic supervised-learning setting. This chapter concludes with a critical discussion of limitations, including task dependence, noise, measurement cost, benchmarking challenges, and the continued dominance of large pretrained classical encoders.
Token generation in Large Language Models (LLMs) are designed on multi-head attention transformers. They capture long-range sequences inherently, which might span transformer full attention pass, and results in escalation of the model cost quadratically. This cost inflation trigger higher inference latency, limiting its deployment in resource-constrained scenarios. Quantum computing (QC) can address these limitations, owing to its advantages of entanglement in high-dimensional Hilbert spaces, and allow parallel sequence generation by LLMs. However, pure quantum transformers design are challenged by high noise fidelity, coherence, and lack of sufficient qubits for operation. Motivated by this, in this paper, we design a framework, QuantLLM, a hybrid classical-quantum transformer architecture that reduces latency via a token-wise, and semantically-aware adaptive routing mechanism. In the routing design, tokens are assigned to either classical or quantum attention paths computed by a differentiable complexity score, that minimizes the LLM inference latency. Results are evaluated on the benchmark WikiText-2 dataset. The framework reduces inference latency by a maximum of 30.06% compared to quantum-enhanced attention baselines. At a balanced setting, it achieves 84.1% accuracy with 290 ms latency, versus 85.7% with 418 ms against baseline scheme. For long sequences, up to 48.7% of tokens are routed through the quantum path, with token-level fidelity scores (0.872 for complex nouns and 0.122 for punctuation), which asserts optimized semantic-aware context routing. These results indicate the potential viability of the framework design in practical model settings.
Quantum transfer learning combines pretrained classical deep learning models with quantum circuits to reuse expressive feature representations while limiting the number of trainable parameters. In this work, we introduce a family of compact quantum transfer learning architectures that attach variational quantum classifiers to frozen convolutional backbones for image classification. We instantiate and evaluate several classical-quantum hybrid models implemented in PennyLane and Qiskit, and systematically compare them with a classical transfer-learning baseline across heterogeneous image datasets. To ensure a realistic assessment, we evaluate all approaches under both ideal simulation and noisy emulation using noise models calibrated from IBM quantum hardware specifications, as well as on real IBM quantum hardware. Experimental results show that the proposed quantum transfer learning architectures achieve competitive and, in several cases, superior accuracy while consistently reducing training time and energy consumption relative to the classical baseline. Among the evaluated approaches, PennyLane-based implementations provide the most favorable trade-off between accuracy and computational efficiency, suggesting that hybrid quantum transfer learning can offer practical benefits in realistic NISQ era settings when feature extraction remains classical.
Muhammad Shafique