Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2604.09430

On the Representational Limits of Quantum-Inspired 1024-D Document Embeddings: An Experimental Evaluation Framework

Dario Maio

cs.IRcs.AIquant-phSubmitted Apr 10, 2026• Announced Apr 13, 2026
arXivHTMLPDF

Abstract

Text embeddings are central to modern information retrieval and Retrieval-Augmented Generation (RAG). While dense models derived from Large Language Models (LLMs) dominate current practice, recent work has explored quantum-inspired alternatives motivated by the geometric properties of Hilbert-like spaces and their potential to encode richer semantic structure. This paper presents an experimental framework for constructing quantum-inspired 1024-dimensional document embeddings based on overlapping windows and multi-scale aggregation. The pipeline combines semantic projections (e.g., EigAngle), circuit-inspired feature mappings, and optional teacher-student distillation, together with a fingerprinting mechanism for reproducibility and controlled evaluation. We introduce a set of diagnostic tools for hybrid retrieval, including static and dynamic interpolation between BM25 and embedding-based scores, candidate union strategies, and a conceptual alpha-oracle that provides an upper bound for score-level fusion. Experiments on controlled corpora of Italian and English documents across technical, narrative, and legal domains, using synthetic queries, show that BM25 remains a strong baseline, teacher embeddings provide stable semantic structure, and standalone quantum-inspired embeddings exhibit weak and unstable ranking signals. Distillation yields mixed effects, improving alignment in some cases but not consistently enhancing retrieval performance, while hybrid retrieval can recover competitive results when lexical and embedding-based signals are combined. Overall, the results highlight structural limitations in the geometry of quantum-inspired embeddings, including distance compression and ranking instability, and clarify their role as auxiliary components rather than standalone retrieval representations.

Comments
44 pages, 6 figures

Related papers

A Hybrid Quantum Fidelity-Based Semantic Retrieval Framework for Retrieval-Augmented Generation

J. Senthilkumar, D.R. Karthika

Retrieval-Augmented Generation (RAG) improves LLM factual accuracy via external retrieval, but existing frameworks rely on cosine similarity, leaving quantum overlap-based measures such as fidelity largely unexplored. QF-RAG is a hybrid quantum-classical retrieval architecture that estimates Quantum Fidelity Similarity (QFS) between amplitude-encoded embeddings via a Swap Test. Because fidelity equals squared cosine similarity for real unit vectors, QFS is fundamentally a score transformation, and we evaluate it entirely on real, handauthored text rather than synthetic vectors: 63 factual passages across 16 real-world topics (paraphrases, hard negatives, and noise documents) and 42 queries (standard, ambiguous, multihop), embedded with real TF-IDF vectors that are mean-centered so cosine similarity can take negative values, as dense embeddings do in practice. On this real corpus, 69.9% of query-document cosine similarities are negative, yet cosine and exact QFS agree on the Top-1 result for every query and produce identical Precision@5, Recall@5, MRR, and nDCG@5 to three decimal places, confirming that QFS preserves cosine's ranking quality on real text. With sufficient Swap Test shots, QFS recovers this same ceiling: MRR reaches 0.951 by$\mathrm{S}=1024$and 1.000 at the exact-fidelity limit, tracking the expected$\mathrm{O}(1 / \sqrt{ } \mathrm{S})$estimator convergence. A retrieval-to-generation test on the same real passages shows retrieval quality causally propagating to answer accuracy (93.8% at k=1). We further complement asymptotic complexity analysis with numbers computed and measured endto- end in code: an unoptimized amplitude-encoding synthesis needs O(d2) CNOT gates (349,525 at$\mathrm{d}=768)$, and we verify a tighter O(d) optimized bound (1,022), with exact memory accounting and a measured, extrapolated simulator runtime of 19.6 s per pair for state preparation at$\mathrm{d}=768$, dropping to 0.057 s with the optimized construction. QF-RAG is positioned as a rigorously specified, empirically evaluated hybrid architecture and evaluation framework for bringing quantum overlap measures into RAG retrieval, with a clear, quantified resource profile for near-term hardware.

•Aug 1, 2026•0 citations

Quantum-Projected Semantic Retrieval: A Universal Low-Latency Backbone for Multi-Modal AI Systems

Amit Rana

Modern AI systems rely on fast, accurate document retrieval at every step — from answering questions to generating responses grounded in real knowledge. Yet today's best retrieval methods treat text as flat numerical coordinates, ignoring the richer geometry that quantum mechanics provides. We present QuantumRAG: a system that maps neural text embeddings into complex quantum states in Hilbert space and retrieves documents using quantum fidelity — a similarity measure that captures both the magnitude and the phase (orientation angle) of semantic meaning. Built on BERT-projected quantum encoding (8 qubits, 256-dimensional Hilbert space) with a Fisher discriminant category mask, QuantumRAG achieves P@1=0.698 and NDCG@10=0.825 on the MATH benchmark (12,500 problems, 7 categories, 500 test queries), matching BERT-cosine accuracy while running 3.6× faster (8.4 ms vs 29.9 ms per query). The system is modality-agnostic: the same quantum backbone retrieves mathematics, natural language documents, and video content without any architectural changes. A Grover-inspired Quantum Amplitude Amplification (QAA) reranker provides a secondary scoring pass using quantum interference principles. All experiments run on standard CPU hardware — no quantum computer required. QuantumRAG serves as the universal retrieval backbone for a multi-modal AI pipeline spanning gesture recognition, mathematical reasoning, and document question answering — demonstrating that quantum-projected representations are a practical, deployable alternative to classical similarity search today, with a clear scaling path as quantum hardware matures.

•Apr 28, 2026•0 citations

Quantum-Enhanced Retrieval-Augmented Generation for Hallucination Reduction in Large Language Models

Praveenkumar Seepana

Although the performance of LLMs on a wide variety of natural language processing problems demonstrates remarkable ability, hallucinated responses are introduced as one of the key weaknesses of LLMs, especially in knowledge-intensive applications, where fidelity to facts is paramount. While effective in reducing hallucinations, the context retrieved during the RAG operations is still very often suboptimal with respect to the external documents used in the retrieval stage, and being based on cosine similarity and nearest-neighbour search, these models typically do not return optimal context to support factual generation. The authors propose a five-stage solution, called Quantum-Enhanced Retrieval-Augmented Generation (QERAG), which combines the use of a quantum-inspired probability amplitude document ranking, a Context Utility Score (CUS) optimisation engine, and a Hallucination Verification Agent (HVA) based on a formal Hallucination Reduction Index (HRI). The Quantum Relevance Score (QRS) uses interference based, Amplitude Encoding on candidate Document Sets to yield a normalised relevance distribution that emphasises the relatively higher value generated in the discriminative context as compared to other locally similar passages in a document set. QERAG outperforms Standard LLM and Traditional RAG models on all five Natural benchmark across response accuracy, with a score of 92.8% vs. 74.3%, and hallucination rate, 4.3% vs. 18.2%, respectively. They achieve a significant overall improvement by using quantum-inspired retrieval layer, which achieves a 3.5-percentage-point reduction in hallucination rate when compared to cosine-similarity RAG, with this single layer showing more reduction than the rest of the pipeline layers. The results of an ablation study show the individual contribution of every pipeline stage, and the quantum-inspired retrieval layer exhibits the greatest reduction of a 3.5 percentage points over cosine-similarity RAG.

International Journal of Computational Science and Engineering Research•Aug 3, 2026•0 citations

Quantum-Rag: Retrieval-Augmented Generation Framework for Quantum Circuit Code Generation

Nicholas Stapleton, Sharmin Afrose, Ethan Seefried +3 more

Generating correct and executable quantum circuit code remains a significant barrier for both newcomers and experienced practitioners in quantum computing. While Large Language Models (LLMs) have shown promise in classical code generation, their application to quantum programming is hampered by limited domain-specific training data and the rapidly evolving nature of quantum software frameworks. In this work, we present Quantum-RAG, an extended Retrieval-Augmented Generation framework that couples a domain-specialized language model with a curated, multi-source knowledge corpus to produce Qiskit-based quantum circuit code. Our three-stage pipeline ingests documentation, educational notebooks, and research literature into a dense retrieval index and augments the generation model with the most relevant context at inference time. Evaluated on the Qiskit HumanEval benchmark, QuantumRAG achieves a Pass@1 of 0.4132, a Pass@3 of 0.6291, and a Pass@5 of 0.7020, surpassing the domain-adapted GraniteQiskit baseline by 101%, 72%, and 47 % at each respective threshold. These results demonstrate that retrieval-augmented approaches can substantially improve quantum code generation without expensive model retraining.

IEEE Dallas Circuits and Systems Conference (DCAS)•Apr 10, 2026•2 citations

Quantum-inspired Embeddings Projection and Similarity Metrics for Representation Learning

Ivan Kankeu, Stefan Gerd Fritsch, Gunnar Schönhoff +3 more

Over the last decade, representation learning, which embeds complex information extracted from large amounts of data into dense vector spaces, has emerged as a key technique in machine learning. Among other applications, it has been a key building block for large language models and advanced computer vision systems based on contrastive learning. A core component of representation learning systems is the projection head, which maps the original embeddings into different, often compressed spaces, while preserving the similarity relationship between vectors. In this paper, we propose a quantum-inspired projection head that includes a corresponding quantum-inspired similarity metric. Specifically, we map classical embeddings onto quantum states in Hilbert space and introduce a quantum circuit-based projection head to reduce embedding dimensionality. To evaluate the effectiveness of this approach, we extended the BERT language model by integrating our projection head for embedding compression. We compared the performance of embeddings, which were compressed using our quantum-inspired projection head, with those compressed using a classical projection head on information retrieval tasks using the TREC 2019 and TREC 2020 Deep Learning benchmarks. The results demonstrate that our quantum-inspired method achieves competitive performance relative to the classical method while utilizing 32 times fewer parameters. Furthermore, when trained from scratch, it notably excels, particularly on smaller datasets. This work not only highlights the effectiveness of the quantum-inspired approach but also emphasizes the utility of efficient, ad hoc low-entanglement circuit simulations within neural networks as a powerful quantum-inspired technique.

arXiv:2501.04591•Jan 8, 2025•7 citations

Quantum-Projected Semantic Retrieval: A Universal Low-Latency Backbone for Multi-Modal AI Systems

Amit Rana

Abstract Modern AI systems rely on fast, accurate document retrieval at every step — from answering questions to generating responses grounded in real knowledge. Yet today's best retrieval methods treat text as flat numerical coordinates, ignoring the richer geometry that quantum mechanics provides. We present QuantumRAG : a system that maps neural text embeddings into complex quantum states in Hilbert space and retrieves documents using quantum fidelity — a similarity measure that sees both the magnitude and the phase (orientation angle) of semantic meaning. On the MATH benchmark (12,500 problems, 7 categories, 500 test queries), Quantum Q-CSD achieves P@1 = 0.698 and NDCG@10 = 0.825, matching BERT-cosine accuracy while running 3.6× faster (8.4 ms vs 29.9 ms per query). The system is modality-agnostic: the same quantum backbone retrieves mathematics, natural language documents, and video content without any architectural changes. All experiments run on standard CPU hardware — no quantum computer required.

•Jun 29, 2026•0 citations
Text embeddings are central to modern information retrieval and Retrieval-Augmented Generation (RAG). While dense models derived from Large Language Models (LLMs) dominate current practice, recent work has explored quantum-inspired alternatives motivated by the geometric properties of Hilbert-like spaces and their potential to encode richer semantic structure.
This paper presents an experimental framework for constructing quantum-inspired 1024-dimensional document embeddings based on overlapping windows and multi-scale aggregation. The pipeline combines semantic projections (e.g., EigAngle), circuit-inspired feature mappings, and optional teacher-student distillation, together with a fingerprinting mechanism for reproducibility and controlled evaluation.
We introduce a set of diagnostic tools for hybrid retrieval, including static and dynamic interpolation between BM25 and embedding-based scores, candidate union strategies, and a conceptual alpha-oracle that provides an upper bound for score-level fusion.
Experiments on controlled corpora of Italian and English documents across technical, narrative, and legal domains, using synthetic queries, show that BM25 remains a strong baseline, teacher embeddings provide stable semantic structure, and standalone quantum-inspired embeddings exhibit weak and unstable ranking signals. Distillation yields mixed effects, improving alignment in some cases but not consistently enhancing retrieval performance, while hybrid retrieval can recover competitive results when lexical and embedding-based signals are combined.
Overall, the results highlight structural limitations in the geometry of quantum-inspired embeddings, including distance compression and ranking instability, and clarify their role as auxiliary components rather than standalone retrieval representations.
Retrieval-Augmented Generation (RAG) improves LLM factual accuracy via external retrieval, but existing frameworks rely on cosine similarity, leaving quantum overlap-based measures such as fidelity largely unexplored. QF-RAG is a hybrid quantum-classical retrieval architecture that estimates Quantum Fidelity Similarity (QFS) between amplitude-encoded embeddings via a Swap Test. Because fidelity equals squared cosine similarity for real unit vectors, QFS is fundamentally a score transformation, and we evaluate it entirely on real, handauthored text rather than synthetic vectors: 63 factual passages across 16 real-world topics (paraphrases, hard negatives, and noise documents) and 42 queries (standard, ambiguous, multihop), embedded with real TF-IDF vectors that are mean-centered so cosine similarity can take negative values, as dense embeddings do in practice. On this real corpus, 69.9% of query-document cosine similarities are negative, yet cosine and exact QFS agree on the Top-1 result for every query and produce identical Precision@5, Recall@5, MRR, and nDCG@5 to three decimal places, confirming that QFS preserves cosine's ranking quality on real text. With sufficient Swap Test shots, QFS recovers this same ceiling: MRR reaches 0.951 byS=1024\mathrm{S}=1024S=1024and 1.000 at the exact-fidelity limit, tracking the expectedO(1/S)\mathrm{O}(1 / \sqrt{ } \mathrm{S})O(1/​S)estimator convergence. A retrieval-to-generation test on the same real passages shows retrieval quality causally propagating to answer accuracy (93.8% at k=1). We further complement asymptotic complexity analysis with numbers computed and measured endto- end in code: an unoptimized amplitude-encoding synthesis needs O(d2) CNOT gates (349,525 atd=768)\mathrm{d}=768)d=768), and we verify a tighter O(d) optimized bound (1,022), with exact memory accounting and a measured, extrapolated simulator runtime of 19.6 s per pair for state preparation atd=768\mathrm{d}=768d=768, dropping to 0.057 s with the optimized construction. QF-RAG is positioned as a rigorously specified, empirically evaluated hybrid architecture and evaluation framework for bringing quantum overlap measures into RAG retrieval, with a clear, quantified resource profile for near-term hardware.
Modern AI systems rely on fast, accurate document retrieval at every step — from answering questions to generating responses grounded in real knowledge. Yet today's best retrieval methods treat text as flat numerical coordinates, ignoring the richer geometry that quantum mechanics provides. We present QuantumRAG: a system that maps neural text embeddings into complex quantum states in Hilbert space and retrieves documents using quantum fidelity — a similarity measure that captures both the magnitude and the phase (orientation angle) of semantic meaning. Built on BERT-projected quantum encoding (8 qubits, 256-dimensional Hilbert space) with a Fisher discriminant category mask, QuantumRAG achieves P@1=0.698 and NDCG@10=0.825 on the MATH benchmark (12,500 problems, 7 categories, 500 test queries), matching BERT-cosine accuracy while running 3.6× faster (8.4 ms vs 29.9 ms per query). The system is modality-agnostic: the same quantum backbone retrieves mathematics, natural language documents, and video content without any architectural changes. A Grover-inspired Quantum Amplitude Amplification (QAA) reranker provides a secondary scoring pass using quantum interference principles. All experiments run on standard CPU hardware — no quantum computer required. QuantumRAG serves as the universal retrieval backbone for a multi-modal AI pipeline spanning gesture recognition, mathematical reasoning, and document question answering — demonstrating that quantum-projected representations are a practical, deployable alternative to classical similarity search today, with a clear scaling path as quantum hardware matures.
Although the performance of LLMs on a wide variety of natural language processing problems demonstrates remarkable ability, hallucinated responses are introduced as one of the key weaknesses of LLMs, especially in knowledge-intensive applications, where fidelity to facts is paramount. While effective in reducing hallucinations, the context retrieved during the RAG operations is still very often suboptimal with respect to the external documents used in the retrieval stage, and being based on cosine similarity and nearest-neighbour search, these models typically do not return optimal context to support factual generation. The authors propose a five-stage solution, called Quantum-Enhanced Retrieval-Augmented Generation (QERAG), which combines the use of a quantum-inspired probability amplitude document ranking, a Context Utility Score (CUS) optimisation engine, and a Hallucination Verification Agent (HVA) based on a formal Hallucination Reduction Index (HRI). The Quantum Relevance Score (QRS) uses interference based, Amplitude Encoding on candidate Document Sets to yield a normalised relevance distribution that emphasises the relatively higher value generated in the discriminative context as compared to other locally similar passages in a document set. QERAG outperforms Standard LLM and Traditional RAG models on all five Natural benchmark across response accuracy, with a score of 92.8% vs. 74.3%, and hallucination rate, 4.3% vs. 18.2%, respectively. They achieve a significant overall improvement by using quantum-inspired retrieval layer, which achieves a 3.5-percentage-point reduction in hallucination rate when compared to cosine-similarity RAG, with this single layer showing more reduction than the rest of the pipeline layers. The results of an ablation study show the individual contribution of every pipeline stage, and the quantum-inspired retrieval layer exhibits the greatest reduction of a 3.5 percentage points over cosine-similarity RAG.
Generating correct and executable quantum circuit code remains a significant barrier for both newcomers and experienced practitioners in quantum computing. While Large Language Models (LLMs) have shown promise in classical code generation, their application to quantum programming is hampered by limited domain-specific training data and the rapidly evolving nature of quantum software frameworks. In this work, we present Quantum-RAG, an extended Retrieval-Augmented Generation framework that couples a domain-specialized language model with a curated, multi-source knowledge corpus to produce Qiskit-based quantum circuit code. Our three-stage pipeline ingests documentation, educational notebooks, and research literature into a dense retrieval index and augments the generation model with the most relevant context at inference time. Evaluated on the Qiskit HumanEval benchmark, QuantumRAG achieves a Pass@1 of 0.4132, a Pass@3 of 0.6291, and a Pass@5 of 0.7020, surpassing the domain-adapted GraniteQiskit baseline by 101%, 72%, and 47 % at each respective threshold. These results demonstrate that retrieval-augmented approaches can substantially improve quantum code generation without expensive model retraining.
Over the last decade, representation learning, which embeds complex information extracted from large amounts of data into dense vector spaces, has emerged as a key technique in machine learning. Among other applications, it has been a key building block for large language models and advanced computer vision systems based on contrastive learning. A core component of representation learning systems is the projection head, which maps the original embeddings into different, often compressed spaces, while preserving the similarity relationship between vectors. In this paper, we propose a quantum-inspired projection head that includes a corresponding quantum-inspired similarity metric. Specifically, we map classical embeddings onto quantum states in Hilbert space and introduce a quantum circuit-based projection head to reduce embedding dimensionality. To evaluate the effectiveness of this approach, we extended the BERT language model by integrating our projection head for embedding compression. We compared the performance of embeddings, which were compressed using our quantum-inspired projection head, with those compressed using a classical projection head on information retrieval tasks using the TREC 2019 and TREC 2020 Deep Learning benchmarks. The results demonstrate that our quantum-inspired method achieves competitive performance relative to the classical method while utilizing 32 times fewer parameters. Furthermore, when trained from scratch, it notably excels, particularly on smaller datasets. This work not only highlights the effectiveness of the quantum-inspired approach but also emphasizes the utility of efficient, ad hoc low-entanglement circuit simulations within neural networks as a powerful quantum-inspired technique.
Abstract Modern AI systems rely on fast, accurate document retrieval at every step — from answering questions to generating responses grounded in real knowledge. Yet today's best retrieval methods treat text as flat numerical coordinates, ignoring the richer geometry that quantum mechanics provides. We present QuantumRAG : a system that maps neural text embeddings into complex quantum states in Hilbert space and retrieves documents using quantum fidelity — a similarity measure that sees both the magnitude and the phase (orientation angle) of semantic meaning. On the MATH benchmark (12,500 problems, 7 categories, 500 test queries), Quantum Q-CSD achieves P@1 = 0.698 and NDCG@10 = 0.825, matching BERT-cosine accuracy while running 3.6× faster (8.4 ms vs 29.9 ms per query). The system is modality-agnostic: the same quantum backbone retrieves mathematics, natural language documents, and video content without any architectural changes. All experiments run on standard CPU hardware — no quantum computer required.