DQAOA-GPT: AI-Accelerated Distributed Quantum Optimization for Combinatorial Problems

This paper proposes DQAOA-GPT, a hybrid framework that combines distributed quantum approximate optimization with GPT-generated circuits. DQAOA decomposes a large combinatorial optimization problem into smaller subproblems, while the generative model directly produces circuits instead of relying on repeated variational parameter optimization. The authors report evaluations on dense HUBO problems with up to 100 decision variables, claiming lower computational cost and competitive solution quality versus conventional DQAOA. The abstract does not provide concrete speedups, hardware details, baselines beyond DQAOA, or statistical uncertainty.