We introduce a hybrid quantum-classical deep learning architecture for large language model fine-tuning. The classical portion of the architecture is a sentence transformer that is powerful enough to display significant accuracy for complex tasks such as sentiment prediction. The quantum portion of the architecture consists of parameterized quantum circuits that utilize long-range connections between qubits. We analyze the performance of the hybrid models for various settings of hyperparameters, including the number of qubits, the depth of the quantum circuits, learning rate, number of reuploading steps, etc. Based on a screening study of main effects, we show an overall improvement in prediction accuracy over a comparable classical baseline, with a trend of increasing accuracy with number of qubits. We observe up to 3.14% improvements in accuracy over classical architectures of comparable model size, within the set of hyperparameters probed in this study. We demonstrate the contribution of each module in our architecture through ablation studies. Our studies are based on finite shot-counts and include simulations based on noisy quantum gates.
We introduce a hybrid quantum-classical deep learning architecture for large language model fine-tuning. The classical portion of the architecture is a sentence transformer that is powerful enough to display significant accuracy for complex tasks such as sentiment prediction. The quantum portion of the architecture consists of parameterized quantum circuits that utilize long-range connections between qubits. We analyze the performance of the hybrid models for various settings of hyperparameters, including the number of qubits, the depth of the quantum circuits, learning rate, number of reuploading steps, etc. Based on a screening study of main effects, we show an overall improvement in prediction accuracy over a comparable classical baseline, with a trend of increasing accuracy with number of qubits. We observe up to 3.14% improvements in accuracy over classical architectures of comparable model size, within the set of hyperparameters probed in this study. We demonstrate the contribution of each module in our architecture through ablation studies. Our studies are based on finite shot-counts and include simulations based on noisy quantum gates.
The rapid progress of large language models (LLMs) has transformed natural language processing, yet the challenge of efficient adaptation remains unresolved. Full fine-tuning achieves strong performance but imposes prohibitive computational and memory costs. Parameter-efficient fine-tuning (PEFT) strategies, such as low-rank adaptation (LoRA), Prefix tuning, and sparse low-rank adaptation (SoRA), address this issue by reducing trainable parameters while maintaining competitive accuracy. However, these methods often encounter limitations in scalability, stability, and generalization across diverse tasks. Recent advances in quantum deep learning introduce novel opportunities through quantum-inspired encoding and parameterized quantum circuits (PQCs). In particular, the quantum-amplitude embedded adaptation (QAA) framework demonstrates expressive model updates with minimal overhead. This paper presents a systematic survey and comparative analysis of conventional PEFT methods and QAA. The analysis demonstrates trade-offs in convergence, efficiency, and representational capacity, while providing insight into the potential of quantum approaches for future LLM adaptation.
Quantum machine learning has recently emerged as a promising paradigm that leverages the expressive power of quantum circuits to address complex learning tasks. In this work, we investigate the applicability of hybrid quantum-classical neural networks to sentiment analysis, a central problem in natural language processing. We focus on a dataset of tweets related to COVID-19, where the textual content is vectorized using TF-IDF and fed into both classical feedforward networks and hybrid architectures incorporating parameterized quantum circuits. Our results show that hybrid models can achieve accuracy comparable to the classical baseline, while exhibiting distinct learning dynamics, especially in terms of validation loss and accuracy, that suggest a richer representational capacity. Moreover, when applying transfer learning to an SMS spam classification task, the hybrid models consistently outperform the classical counterpart, achieving an accuracy increase of 15 percentage points (from 66% to 81%) on the spam class, demonstrating enhanced generalization. These findings highlight the feasibility of employing QML for natural language processing and point toward the potential advantages of hybrid models as quantum hardware continues to advance.
Large language models (LLMs) exhibit abilities beyond natural language modelling and text generation. Recent advances in their reasoning capabilities have spurred interest in applying LLMs to complex scientific tasks requiring deep domain expertise and sophisticated reasoning. Quantum computing, as a highly specialised field with significant knowledge barriers and hardware constraints, could greatly benefit from such advancements. However, a key open question that first must be answered is: How can we develop fine-tuning pipelines that instil genuine quantum reasoning in LLMs, rather than task-specific pattern matching? We study this question through quantum circuit simulation as a training objective, where the model must predict the measurement probability distribution resulting from a sequence of quantum gate operations. We propose and compare two fine-tuning pipelines: (1) Supervised Fine-Tuning (SFT) on explicit gate-by-gate state-vector simulation traces, and (2) a two-stage SFT+Group Relative Policy Optimisation (GRPO) approach that sequentially applies SFT followed by GRPO with verifiable rewards. Our findings show that SFT achieves near-perfect in-distribution and gate-count extrapolation accuracy, significantly outperforming both the base model and the GPT-OSS-120B baseline. SFT+GRPO trades some in-distribution precision for better generalisation to larger qubit systems that SFT alone cannot handle. Both pipelines significantly outperform the baselines, demonstrating that targeted fine-tuning on explicit reasoning traces is an effective strategy for advancing quantum reasoning in LLMs.
Sentence classification is a core task in Natural Language Processing, and sentiment analysis in particular remains difficult because of contextual ambiguity, semantic complexity, and the overlap between closely related sentiment classes. This paper presents a hybrid quantum-classical architecture for sentence classification that combines classical deep learning with quantum feature extraction. The classical branch employs a Feedforward Neural Network (FNN) and a Long Short-Term Memory (LSTM) network to learn complementary semantic and contextual representations, while the quantum branch reduces TF-IDF features through Truncated Singular Value Decomposition (SVD) and encodes them into a parameterized quantum circuit built with Qiskit. The classical and quantum representations are combined through feature-level fusion and classified using a MultiLayer Perceptron (MLP). The framework is evaluated on the SST-2 dataset for binary sentiment classification and on a merged SST-2 + SST-5 dataset for multi-class classification. The proposed hybrid model achieves an accuracy of 90.5%, precision of 89.9%, recall of 90.1%, and F1-score of 90.1% on SST-2, with a ROC-AUC close to 0.97, exceeding the individual FNN and LSTM baselines. On the merged multi-class dataset, the hybrid model attains 66.65% accuracy against 65.15% for FNN and 62.65% for LSTM. These results indicate that feature-level integration of classical and quantum representations can improve sentence classification performance within currently available quantum simulation environments.
Although recent advances in transfer learning have simplified training, fine-tuning big language models remains an energy-intensive and expensive process, with high hardware and energy costs that make it difficult to use and scale. While quantum machine learning (QML) presents a promising theoretical tool to overcome these bottlenecks, the practical implementation is hampered by the barren plateau problem, a phenomenon well established in literature that manifests through exponentially vanishing gradients of deep parameterized quantum circuits (PQCs) with increasing circuit depth, preventing informative parameter updates by rendering the optimization landscape flat. This study presents a hybrid quantum-classical architecture in which a PQC layer is integrated into a pretrained classical LLM backbone and fine-tuned throughout this work. This iterative refinement loop is controlled by an Expected Improvement acquisition function, which actively guides the search process through parameter regimes with obvious non-vanishing gradient variance. The study proposes that this theory-driven initialization scheme reduces the onset of barren plateaus, thereby enabling efficient and stable convergence in hybrid optimization. Experiments show that the proposed method consistently outperforms purely classical fine-tuning baselines and randomly initialized quantum baselines on representative downstream natural language processing tasks. The experimental results show that principled parameter initialization leads to concrete improvements in convergence stability, gradient trainability, and task-level performance. In summary, this research provides a scalable method for integrating near-term noisy intermediate-scale Quantum (NISQ) devices into state-of-the-art deep learning pipelines, fostering the further real-world adoption of hybrid quantum-classical systems for demanding artificial intelligence tasks.
In recent years, the field of machine learning has witnessed a paradigm shift with the emergence of hybrid quantum-classical neural networks. These networks combine the power of classical neural networks with the computational advantages offered by quantum variational circuits. This study presents a comprehensive survey of hybrid quantum-classical machine learning algorithms, focusing on their architecture, performance, and hyperparameter optimization techniques. Introducing the concept of hybrid quantum-classical machine learning and its significance in contemporary image classification tasks. It explores the theoretical foundations of quantum variational circuits and their integration with classical neural network architectures. Special attention is given to prominent algorithms such as Quantum Convolutional Neural Networks, Quanvolutional Neural Networks, and Quantum ResNet, each offering unique advantages in image classification tasks. A major aspect covered in this study is the hyperparameter optimization techniques employed in hybrid quantum-classical machine learning. The chapter discusses quantum-inspired optimization methods tailored for efficiently tuning the hyperparameters of complex neural network models. Through comparative analysis, the effectiveness of these optimization techniques is demonstrated in reducing computational overhead and improving model performance. Moreover, the study also delves into experimental results obtained from benchmarking hybrid quantum-classical models against traditional classical approaches. It highlights the accuracy variations observed across different quantum architectures and explores the impact of architectural permutations on model convergence. Insights gained from these experiments provide valuable guidance for researchers and practitioners seeking to leverage hybrid quantum-classical neural networks in real-world applications.