Quantivex: an LLM-guided framework for noise-aware variational quantum simulation on superconducting processors

We present the first version of Quantivex, a Python package that answers, which qubits of a noisy quantum chip to use and whether the result can be trusted. Language models guide the analysis, while code performs the calculations. Initial validation includes 10 runs on IBM quantum hardware. Software rights: Quantivex software is governed by its applicable software license. Publication of this research paper does not independently grant rights to use, modify, distribute, or commercialize the underlying proprietary software.