An LLM System for Autonomous Variational Quantum Circuit Design

The article presents an autonomous framework that utilizes large language models (LLMs) for iterative quantum circuit design, integrating components such as Exploration, Generation, and Evaluation. The framework was evaluated on tasks including quantum feature map construction and ansatz generation, achieving superior performance against traditional quantum feature maps and competitive accuracy in molecular ground state estimation. This approach demonstrates the potential of LLMs in automating complex design processes in quantum computing, thereby streamlining scientific optimization workflows.