Q-GenZ-AIQuanta: A Quantum-Enhanced Evaluation Model for Next-Gen Text-to-Text Generative AI in Knowledge Management System
Mohammad Imroz Malik, Navin Kumar Agrawal
The mesmerising evolution of Generative Artificial Intelligence (Gen.AI) has transformed knowledge management, yet evaluating its reliability and contextual precision remains a critical challenge. This paper proposes GenZ-AIQuanta, a hybrid quantum–classical (Q–C) evaluation model designed for text-to-text generative AI systems. The model integrates classical linguistic metrics—BLEU (Biilingual Evaluations Study) and BERT.Score (Bi-direction Encoder Representation from Transformer Score)—with quantum-derived semantic measures such as QSS (Quantum Semantic Similarity) and QSF (Quantum State Fidelity). Implemented using Python, Qiskit, and HuggingFace Transformers, the system enables multidimensional analysis via a hybrid aggregation engine. Experimental outcomes exhibit significant improvements in accuracy, contextual coherence, and explainability. Moreover, the EC (Explainability Coefficient) highlights enhanced interpretive reliability across evaluation layers. The proposed Q–C hybrid architecture establishes a scalable, explainable, and quantum-enhanced foundation for assessing next-generation GenAI models within knowledge-centric environments.
International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond (NQComp)•Mar 5, 2026•0 citations