Explicabilidad en algoritmos de búsqueda cuántica en hipercubo con valores de Shapley
Abstract
This work explores the explainability of the coined quantum walk search algorithm on the hypercube by incorporating the SMEF-E (Shapley–Matrix Explainability Framework – Energy) methodology. In this approach, cooperative game theory is combined with Hamiltonian-based value functions to attribute, step by step, the energetic and functional impact of the oracle, the Grover coin, and the flip-flop shift. The resulting Shapley value decomposition offers an interpretable quantification of operator influence throughout the amplification process, revealing how constructive interference emerges and is redistributed as the algorithm approaches its optimal success probability. The experimental analysis confirms theoretical predictions while providing transparent insights into the mechanisms driving quantum advantage in spatial search.
