[2504.07396] Automating quantum feature map design via large language models

Quantum Physics arXiv:2504.07396 (quant-ph) [Submitted on 10 Apr 2025 (v1), last revised 24 Mar 2026 (this version, v2)] Title:Automating quantum feature map design via large language models Authors:Kenya Sakka, Kosuke Mitarai, Keisuke Fujii View a PDF of the paper titled Automating quantum feature map design via large language models, by Kenya Sakka and 1 other authors View PDF Abstract:Quantum feature maps are a key component of quantum machine learning, encoding classical data into quantum states to exploit the expressive power of high-dimensional Hilbert spaces. Despite their theoretical promise, designing quantum feature maps that offer practical advantages over classical methods remains an open challenge. In this work, we propose an agentic system that autonomously generates, evaluates, and refines quantum feature maps using large language models. The system consists of five components: Generation, Storage, Validation, Evaluation, and Review. Using these components, it iteratively improves quantum feature maps. Through numerical evaluations on widely used benchmark datasets, the system discovers and improves quantum feature maps without human intervention. On MNIST, the best generated feature map achieves 97.3% classification accuracy, outperforming existing quantum feature maps and achieving competitive performance with classical kernels, remaining within 0.3 percentage points of the radial basis function kernel. Similar improvements are observed on Fashion-MNIST an...