LLM Agent Automates Quantum Sensing with NV Centers

A team of researchers has created an agentic AI workflow that utilizes a large language model (LLM) to independently perform experiments involving nitrogen-vacancy (NV) centers in diamond. This innovative system, described in arXiv:2607.25145, integrates ongoing project documentation, tools for quantitative analysis, and precise control over experiments. In one instance, the agent identified a specific NV center, adjusted its resonant frequency, assessed T2* using Ramsey sequences, and incorporated a CPMG measurement to explore a subtle feature possibly associated with nearby 13C. Additionally, two offline benchmarks were established to assess the agent's reasoning without relying on physical equipment. This research propels autonomous quantum sensing forward by facilitating LLM-based decision-making in experimental physics.