The variational quantum algorithm (VQA), executable in near-term devices, has significantly attracted attention when proposed a decade ago, but its scalability issues, such as the barren plateau issue, hinder its practical applications.

I strongly believe that utilizing the power of classical machine learning is the key to circumvent this situation. To show the significance of this possibility, we have recently posted the paper: https://arxiv.org/abs/2401.09253. In the paper, we propose the Generative Quantum Eigensolver (GQE) utilizing the generative model of quantum circuits and its pre-trainable Transformer implementation (GPT-QE). Please check our paper for the details.