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

Quantum-Enhanced Retrieval-Augmented Generation for Hallucination Reduction in Large Language Models Seepana Praveenkumar 3 3 (July - September) 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 knowledgeintensive 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 nearestneighbour search these models typically do not return optimal context to support factual generation The authors propose a fivestage solution called QuantumEnhanced RetrievalAugmented Generation QERAG which combines the use of a quantuminspired 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 928 vs 743 and hallucination rate 43 vs 182 respectively They achieve a significant overall improvement by using quantuminspired retrieval layer which achieves a 35percentagepoint reduction in hallucination rate when compared to cosinesimilarity 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 quantuminspired retrieval layer exhibits the greatest reduction of a 35 percentage points over cosinesimilarity RAG

2026